Department of Labor

Document Sample
Department of Labor Powered By Docstoc
					federal register
                   October 17, 1997

                   Part II

                   Department of Labor
                   Occupational Safety and Health

                   29 CFR Part 1910
                   Occupational Exposure to Tuberculosis;
                   Proposed Rule

54160           Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules

DEPARTMENT OF LABOR                          between 1985–1992, 16 states reported         notice will be issued upon
                                             an increase in the number of TB cases         determination of the locations and dates
Occupational Safety and Health               in 1995, compared with 1994. Based on         of these hearings.
Administration                               a review of the data, OSHA has                  Comments on the proposed standard,
                                             preliminarily concluded that workers in       Notices of Intention to Appear at the
29 CFR Part 1910                             hospitals, nursing homes, hospices,           informal public hearings, testimony,
[Docket No. H–371]                           correctional facilities, homeless shelters,   and documentary evidence are to be
                                             and certain other work settings are at        submitted in quadruplicate to the
RIN 1218–AB46                                significant risk of incurring TB infection    Docket Officer, Docket No. H–371,
                                             while caring for their patients and           Room N–2625, U.S. Department of
Occupational Exposure to                     clients or performing certain
Tuberculosis                                                                               Labor, 200 Constitution Ave., NW,
                                             procedures. To reduce this occupational       Washington, DC 20210, telephone (202)
AGENCY:  Occupational Safety and Health      risk, OSHA is proposing a standard that       219–7894. Comments of 10 pages or
Administration (OSHA), Labor                 would require employers to protect TB-        fewer may be transmitted by fax to (202)
                                             exposed employees by means of                 219–5046, provided the original and
ACTION: Proposed rule and notice of
                                             infection prevention and control              three copies are sent to the Docket
public hearing.
                                             measures that have been demonstrated          Officer thereafter. The hours of
SUMMARY: The Occupational Safety and         to be highly effective in reducing or         operation of the Docket Office are 10:00
Health Administration is proposing a         eliminating job-related TB infections.        a.m. until 4:00 p.m.
health standard, to be promulgated           These measures include the use of
                                             respirators when performing certain             Written comments, Notices of
under section 6(b) of the Occupational                                                     Intention to Appear at the informal
Safety and Health Act of 1970, 29 U.S.C.     high hazard procedures on infectious
                                             individuals, procedures for the early         rulemaking hearings, testimony,
655, to control occupational exposure to                                                   documentary evidence for the hearings,
tuberculosis (TB). TB is a                   identification and treatment of TB
                                             infection, isolation of individuals with      and all other material related to the
communicable, potentially lethal                                                           development of this proposed standard
disease that afflicts the most vulnerable    infectious TB in rooms designed to
                                             protect those in the vicinity of the room     will be available for inspection and
members of our society: the poor, the                                                      copying in the Docket Office, Room N–
sick, the aged, and the homeless. As         from contact with the microorganisms
                                             causing TB, and medical follow-up for         2625, at the above address.
many as 13 million U.S. adults are
presently believed to be infected with       occupationally exposed workers who            FOR FURTHER INFORMATION CONTACT:
TB; over time, more than 1 million of        become infected. OSHA has                     Bonnie Friedman, Office of Information
these individuals may develop active         preliminarily determined that the             and Consumer Affairs, Occupational
TB disease and transmit the infection to     engineering, work practice, and               Safety and Health Administration,
others. TB remains a major health            administrative controls, respiratory          Room N–3647, U.S. Department of
problem with 22,813 active cases             protection, training, medical                 Labor, 200 Constitution Ave., NW,
reported in the U.S. in 1995. A number       surveillance, and other provisions of the     Washington, DC 20210, Telephone (202)
of outbreaks of this disease have            proposed standard are technologically         219–8148, FAX (202) 219–5986.
occurred among workers in health care        and economically feasible for facilities
                                             in all affected industries.                   SUPPLEMENTARY INFORMATION:
settings, as well as other work settings,
in recent years. To add to the               DATES: Written comments on the                Table of Contents
seriousness of the problem, some of          proposed standard must be postmarked          I. Introduction
these outbreaks have involved the            on or before December 16, 1997 and            II. Pertinent Legal Authority
transmission of multidrug-resistant          notices of intention to appear at the         III. Events Leading to the Proposed Standard
strains of Mycobacterium tuberculosis,       informal rulemaking hearings must be          IV. Health Effects
which are often fatal. Although it is the    postmarked on or before December 16,          V. Preliminary Risk Assessment
responsibility of the U.S. Public Health     1997.                                         VI. Significance of Risk
Service to address the problem of               Parties requesting more than 10            VII. Preliminary Economic and Regulatory
tuberculosis in the general U.S.             minutes for their presentation at the               Flexibility Analysis
population, OSHA is solely responsible       hearings and parties submitting               VIII. Unfunded Mandates
for protecting the health of workers         documentary evidence at the hearing           IX. Environmental Impact
                                             must submit the full text of their            X. Summary and Explanation of the Proposed
exposed to TB as a result of their job.
   OSHA estimates that more than 5           testimony and all documentary
                                                                                           XI. Public Participation—Notice of Hearing
million U.S. workers are exposed to TB       evidence no later than December 31,           XII. Authority and Signature
in the course of their work: in hospitals,   1997.                                         XIII. The Proposed Standard
homeless shelters, nursing homes, and           The informal public hearings will
other work settings. Because active TB       begin at 10:00 a.m. on the first day of         References to the rulemaking record
is endemic in many U.S. populations,         hearing and at 9:00 a.m. on each              are in the text of the preamble.
including groups in both urban and           succeeding day. The informal public           References are given as ‘‘Ex.’’ followed
rural areas, workers who come into           hearings will be held in Washington,          by a number to designate the reference
contact with diseased individuals are at     D.C. and are scheduled to begin on            in the docket. For example, ‘‘Ex. 1’’
risk of contracting the disease              February 3, 1998.                             means exhibit 1 in the Docket H–371.
themselves. The risk confronting these       ADDRESSES: Hearings will be held in the       This document is a copy of the petition
workers as a result of their contact with    Auditorium of the U.S. Department of          for a permanent standard filed by the
TB-infected individuals may be as high       Labor (Frances Perkins Building), 200         Labor Coalition to Fight TB in the
as 10 times the risk to the general          Constitution Avenue, NW, Washington,          Workplace on August 25, 1993. A list of
population. Although the number of           D.C. Subsequent additional informal           the exhibits and copies of the exhibits
reported cases of active TB has slowly       public hearings will be held in other         are available in the OSHA Docket
begun to decline after a resurgence          U.S. locations. A Federal Register            Office.
                  Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules                         54161

I. Introduction                              the percentage of the total workforce          9. Of the suspected infectious TB
  The preamble to the Proposed               these workers represent, by industry.       cases referred to hospitals from other
                                                2. Are OSHA’s estimates of controlled    facilities, how many are immediately
Standard for Occupational Exposure to
                                             access rates (i.e., the percentage of       ruled out without needing to be
Tuberculosis discusses the events
                                             workers currently at risk who would         isolated?
leading to the development of the                                                           10. Are OSHA’s estimates of the
proposed standard, the health effects of     remain at risk after employers minimize
                                             the number of workers exposed to            number of necessary AFB isolation
exposure to tuberculosis, and the degree                                                 rooms reasonable? Are existing AFB
and significance of the risk. An analysis    individuals with suspected or
                                             confirmed infectious TB) reasonable? If     isolation rooms reasonably accessible to
of the technological and economic                                                        facilities that transfer individuals with
feasibility of the proposal and an           the number of workers exposed to
                                             individuals with suspected or               suspected or confirmed infectious TB?
explanation of the rationale supporting                                                     11. What types of respirators are
the specific provisions of the proposed      confirmed infectious TB is minimized,
                                             by what percentage could the number of      currently being used to protect workers
standard are also included.                                                              against occupational exposure to M.
  Public comment on all matters              workers at risk be reduced in each
                                             affected industry? In each industry,        tuberculosis?
discussed in this notice and all other                                                      12. Which of the NIOSH-approved
relevant issues is requested for the         what are the job categories that would
                                             continue to be occupationally exposed?      N95 respirators meet all of the proposed
purpose of assisting OSHA in the                                                         criteria, including fit testing and fit
development of a new standard for               3. Are OSHA’s estimates of the
                                                                                         checking criteria?
occupational exposure to tuberculosis.       numbers of affected establishments
                                                                                            13. Are OSHA’s estimates of
                                             reasonable? If not, please provide
A. Issues                                                                                respirator usage rates reasonable? For
                                             estimates of the number of affected
                                                                                         each of the covered industries, how
   OSHA requests comment on all              establishments, by industry.
                                                                                         often could respirators meeting the
relevant issues discussed in this               4. Are OSHA’s estimates of               proposed requirements be reused and
preamble, including the health effects,      occupational and job turnover rates         still maintain proper working
risk assessment, significance of risk        reasonable? If not, please provide          condition? How often, on average,
determination, technological and             estimates of turnover rates for each of     would respirators need to be replaced?
economic feasibility and requirements        the affected industries.                    Please specify the type of respirator.
that should be included in the final            5. Under what conditions would              14. OSHA has assumed, in its
standard. OSHA is especially interested      social work, social welfare services,       Preliminary Economic Analysis, that
in responses, supported by evidence          teaching, law enforcement or legal          hospitals will have licensed health care
and reasons, to the following questions.     services need to be provided to             professionals on-site to perform the
This list is provided to assist persons in   individuals identified as having            medical procedures that would be
formulating comments, but is not             suspected or confirmed infectious TB?       required by the proposed rule, and that
intended to be all inclusive or to           What, if any, procedures could not be       in the other industries, employees will
indicate that participants need to           postponed until such individuals are        have to travel off-site to receive the
respond to all issues or follow this         determined to be noninfectious? How         medical procedures. Which of the other
format. Please give reasons for your         many workers in each of these               affected industries typically have
answers and provide data when                categories may need to have contact         licensed health care professionals on
available.                                   with individuals with suspected or          site who could perform the required
   Specific issues of concern to OSHA        confirmed infectious TB under these         medical procedures? If employers were
are the following:                           conditions?                                 allowed two weeks to provide the
                                                6. Using the proposed definition of      medical procedures, rather than being
Health Effects                               ‘‘suspected infectious TB,’’ how many       required to provide them prior to initial
  1. What, if any, additional studies or     individuals with suspected infectious       assignment to jobs with occupational
case reports on TB should be included        TB are likely to be encountered for         exposure, will it be less likely that
in the health effects analysis?              every confirmed infectious TB case in       employees will have to travel off site to
  2. Is there information that will          each of the covered industries?             receive these tests/procedures? What
provide data for estimating the rise in         7. Are OSHA’s estimates of the           would the costs be if employees travel
Multidrug-resistant TB (MDR–TB)? Is          average number of suspected or              off-site for these tests/procedures?
the rise in MDR–TB a serious threat?         confirmed infectious TB cases that             15. Are OSHA’s estimates of baseline
                                             would be transferred, per establishment     compliance reasonable? If not, what
Risk Assessment                              in each industry, reasonable? If not, on    types of controls are currently in place
  1. Are there alternative risk              average, how many TB cases per facility     to protect workers against occupational
assessment methodologies available?          in each of the affected industries would    exposure to M. tuberculosis, and what
What are they? Are there other studies       be transferred?                             proportion of facilities in each of the
available that would be useful for              8. How are individuals with              affected industries currently are using
assessing risk?                              suspected infectious TB transferred to      such controls?
  2. Are there factors other than or in      establishments with AFB isolation              16. For facilities that have
addition to the ones OSHA has chosen         facilities? Who pays for the transport of   implemented controls to protect
that would be useful in estimating the       such cases, particularly for individuals    workers against occupational exposure
background risk for TB?                      transferred from homeless shelters?         to M. tuberculosis, how effective have
                                             OSHA solicits comment on the                such controls been in reducing the
Technological and Economic Feasibility       feasibility of temporary AFB isolation      transmission of TB?
   1. Are OSHA’s estimates of the            facilities in homeless shelters and on         17. OSHA’s Initial Regulatory
numbers and types of workers currently       methods that could be used to               Flexibility Analysis assesses the impacts
exposed to M. tuberculosis reasonable?       temporarily isolate individuals with        of the proposed standard on small
If not, please provide estimates of the      suspected or confirmed infectious TB in     entities using the Small Business
number of workers currently at risk and      homeless shelters.                          Administration’s (SBA) size standards.
54162           Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules

In addition, OSHA analyzed the impacts       the regulation. OSHA believes that it           • Number of TB skin test
of the proposed standard on entities         has at least partially adopted a number      conversions and active cases among the
employing fewer than 20 workers. Are         of these approaches. OSHA welcomes           homeless and homeless shelter
these definitions appropriate for the        comments and suggestions on these            employees;
covered industries? If not, how should       approaches and the extent to which              • Types of benefits offered to
small entities be defined for each           OSHA should further adopt them:              homeless shelter employees (e.g., health
industry?                                       • Cooperative initiatives, such as        insurance);
   18. The SBA defines small                 expanding OSHA’s current cooperative            • Economic feasibility:
government jurisdictions as                  initiative with JCAHO;                       —Costs of running a shelter;
‘‘governments of cities, counties, towns,       • A federal-state government public       —Revenue sources;
townships, villages, school districts, or    health partnership to develop guidelines     —How costs are accommodated as the
special districts with populations of less   in various industry sectors;                    number of homeless persons served
than 50,000.’’ OSHA requests comment            • Performance standards developed            increases; and
on the number of such small                  with the assistance of federal, state, and   —Opportunities for cost pass-through;
government jurisdictions.                    local government, and labor and                 • Number, location and types (e.g.,
   19. Some parties have suggested that      industry stakeholders;                       family-oriented, walk-in, all-male) of
OSHA should allow the use of the CDC            • Separate approaches for the health      homeless shelters;
guidelines as an alternative to the          and non-health industries (the approach         • Number or proportion of homeless
proposed rule. However, OSHA believes        for the health industries could be keyed     shelter workers who are unpaid
that the CDC guidelines are not written      to existing industry standards and that      volunteers; and
in a regulatory format that would allow      for non-health industries to guidelines);       • The OSH Act applies to
OSHA’s Compliance Safety and Health             • Different levels of compliance          employees, not bona fide volunteers.
Officers (CSHOs) to determine whether        requirements for different industries,       However, OSHA understands that some
or not an employer is in compliance          depending on their expertise, resources,     states may, as a matter of law, require
with the Guidelines. Others have             and risk;                                    facilities to provide volunteers with
suggested that OSHA could judge                 • Less stringent trigger mechanisms
                                                                                          protections established by OSHA
compliance with the guidelines by            for the more burdensome portions of the
                                                                                          standards. OSHA is seeking information
determining the number or rate of skin       standard; and
test conversions at the employer’s              • Separate standards for each
facility. OSHA does not believe that         affected industry.                           —Economic impacts in such states of
smaller facilities have an adequate             24. OSHA is proposing to include             covering volunteers (e.g., how costs
population for trends in test conversions    homeless shelters in the Scope of the           would be handled, cost pass-through);
to have any statistical validity. OSHA       standard. During the informal public            and
welcomes suggestions on any methods          hearings, OSHA intends to schedule a         —Protections currently offered to
of making the CDC guidelines an              special session for participants to             volunteers.
enforceable alternative to an OSHA           present additional information on               25. In what states, if any, do
regulation or methods of measuring           homeless shelters. Also, OSHA is             employers provide volunteers in the
performance that could be applied            conducting a special study of the            sectors affected by this proposed
across all types and sizes of facilities.    homeless shelter sector. The                 standard with the same protections as
   20. Because of the limited availability   information gathered in the study will       they provide to employees? How many
of data, OSHA characterized the risk in      be placed in the docket for public           volunteers might be affected by such
many sectors as similar to that in           comment. OSHA welcomes comment on            requirements?
hospitals, and less than that                any of the topics this study will cover         26. OSHA is concerned that medical
documented in nursing homes and              including:                                   removal protection and medical
home health care. OSHA welcomes                 • Percentage of homeless persons          treatment of active cases of TB may have
industry-specific data on test conversion    that would meet OSHA’s definition of a       significant economic impacts on small
rates or active case rates.                  suspected infectious TB case (A              firms that have an employee with an
   21. OSHA is unable to determine the       breakdown of which symptoms are              active case of TB. Is there any form of
effectiveness of specific elements of an     particularly common will help OSHA           insurance available for covering the
effective infection control program in       construct the best definition);              costs of medical removal protection or
hospitals. OSHA welcomes any                    • Turnover among the homeless who         medical treatments required by the
evidence on the relative effectiveness of    use shelters;                                OSHA standard? Should OSHA
individual elements in such programs,           • Employee turnover in homeless           consider phasing-in these provisions of
such as the identification and isolation     shelters;                                    the standard?
of suspect cases, the use of engineering        • Trends in the number of homeless           27. OSHA believes that substance
controls, the use of respirators, and        persons served in shelters.                  abuse treatment centers, particularly in-
employee training.                              • Criteria currently used by some         patient treatment centers, normally have
   22. OSHA based its estimate of the        homeless shelters to identify suspected      entry procedures that may include
effectiveness of infection control           infectious TB cases;                         medical examinations. OSHA solicits
programs in other sectors on studies of         • Current practices used in homeless      comments on entry procedures for
the effectiveness of such programs in        shelters to address TB hazards so that       substance abuse treatment programs, the
hospitals. OSHA welcomes any data            baseline compliance with the proposed        extent to which these entry procedures
concerning the effectiveness of OSHA’s       standard can be determined. Of               now include medical examinations, and
proposed infection prevention                particular concern to OSHA are:              the extent to which these examinations
measures, or of other alternative            —Methods of isolation; and                   now include and examination for TB
infection control measures, in sectors       —How suspected TB cases are handled.         symptoms.
other than hospitals.                           • Feasibility of hospitals providing         28. OSHA requests comment on the
   23. SBREFA Panel members suggested        cards to the homeless indicating TB skin     effects of extended compliance phase-in
a number of alternative approaches to        test status;                                 dates for the proposed requirements,
                Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules                                54163

particularly for respirators, for small       standard to cover all or some offices of       0 cases of confirmed infectious TB
businesses and facilities relying on          general practitioners or dentists and if       reported in one year and fewer than 6
charitable and/or Medicare and                so, how? Should OSHA expand the                cases of confirmed infectious TB
Medicaid funding.                             scope to cover all teachers?                   reported in the other year. Are there
   29. OSHA requests comment on all              2. Are there provisions of the standard     alternative methods that can be used to
assumptions and estimates used in             with which emergency medical services,         assure protection of employees in areas
developing the Preliminary Economic           home health care, and home-based               where infectious TB has not recently
Analysis. Please provide reasons and          hospice care employers cannot comply           been encountered?
data to support suggested changes to the      because their employees are at
assumptions and estimates.                    temporary work settings over which the         Exposure Control Plan
   30. The World Health Organization          employer has little or no control? If so,        1. OSHA has proposed that the
(WHO) has launched an initiative to           what are those provisions and why              employer’s exposure control plan
reduce active TB through the use of           would an employer be unable to comply          contain certain policies and procedures.
multi-drug therapy and using directly         with them?                                     What, if any, policies and procedures
observed therapy. OSHA solicits                  3. In covering only long-term care          should be added to the plan?
comment on whether it should revise its       facilities for the elderly, is OSHA              2. The proposed standard requires
risk assessment or any of its benefits        excluding similar facilities where there       exposure incidents and skin
estimates as a result of this initiative.     is increased risk of transmission of TB?       conversions to be investigated, but does
   31. OSHA requests comment on the           If so, what are these facilities? Should       not require aggregate data regarding
number of affected facilities that are        OSHA include long-term care                    employee conversions to be collected
tribally-operated, by industry.               populations in addition to the elderly,        and analyzed. Would the collection and
                                              such as long-term psychiatric care             analysis of aggregate data provide
                                              facilities? If so, what are these              benefits beyond those provided by
   1. A number of provisions in the           populations?
proposed standard are triggered by the                                                       investigating each individual exposure
                                                 4. OSHA is proposing that employers
identification of an individual as having                                                    incident or conversion? Why or why
                                              provide medical management and
either ‘‘suspected infectious                                                                not? If aggregate data collection and
                                              follow-up for their employees who work
tuberculosis’’ or ‘‘confirmed infectious                                                     analysis were required, what type of
                                              in covered work settings, but who are
tuberculosis.’’ Of these provisions, are                                                     analysis should be required, at what
                                              not occupationally exposed, when they
there some that should be triggered only                                                     analytical endpoint should employer
                                              have an exposure incident resulting
once an individual has been identified                                                       action be required, and what should that
                                              from an engineering control failure or
as having ‘‘confirmed infectious                                                             action be?
                                              similar workplace exposure. Is this the
tuberculosis?’’ If so, which provisions       best way of assuring such employees              3. OSHA has set forth the extent of
and why?                                      receive medical management and                 responsibility for transfer of individuals
   2. A number of the proposed                follow-up?                                     based upon the type of work setting
standard’s provisions require                    5. OSHA is covering employees who           where such individuals are
compliance or performance on an               have occupational exposure in covered          encountered. What are current practices
annual basis, e.g., reviews of the            work settings yet are not employees of         regarding transfer of individuals with
exposure control plan, the biosafety          the work setting (e.g., physician              suspected or confirmed infectious TB in
manual for laboratories, and the              employed by another employer with              the work settings covered by the
respiratory protection program;               hospital privileges, who is caring for a       proposal?
certification of biological safety            TB patient in the hospital). Can this be       Work Practices and Engineering
cabinets; fit testing or a determination of   made more clear?                               Controls
the need for fit testing of respirators;         6. OSHA has proposed that facilities
medical histories, TB skin tests; and         offering treatment for drug abuse be              1. Is OSHA’s time limit of 5 hours
training. In addition, certain                covered in the scope of the standard. Is       following identification for transferring
requirements must be performed on a           coverage of such facilities appropriate?       an individual with suspected or
semi-annual basis, e.g., inspection and       What factors unique to facilities that         confirmed infectious TB to another
performance monitoring of engineering         offer treatment for drug abuse would           facility or placing the individual into
controls, verification of air flow            make compliance with the provisions of         AFB isolation appropriate? If not, what
direction in laboratories, and, in some       this proposed standard infeasible (e.g.,       is the maximum amount of time that an
instances, TB skin testing. How can           would complying with certain                   individual should be permitted to await
OSHA reduce the aggregate burden of           provisions of the standard compromise          transfer or isolation in a facility before
these requirements, particularly in small     the provision of services at facilities that   the employer must implement the other
entities, while still providing equal         offer treatment for drug abuse)?               provisions of the proposed standard?
protection to employees? Of these                                                               2. OSHA has considered requiring
                                              Application                                    facilities that encounter 6 or more
annual and semi-annual provisions,
which, if any, should be performed less         1. OSHA has proposed that an                 individuals with confirmed infectious
frequently? Why and at what frequency?        employer covered under the standard            TB within the past 12 months to provide
Which of these provisions, if any,            (other than an operator of a laboratory)       engineering controls in intake areas
should be performed more frequently?          may claim reduced responsibilities if he       where early identification procedures
Why and at what frequency?                    or she can demonstrate that his or her         are performed (e.g., emergency
                                              facility or work setting: (1) Does not         departments, admitting areas). Should
Scope                                         admit or provide medical services to           this be a requirement? Are there types
   1. Is there information demonstrating      individuals with suspected or                  of controls, engineering or otherwise,
risk of TB transmission for employees in      confirmed infectious TB; (2) has had no        that would be effective in controlling
work settings other than those included       case of confirmed infectious TB in the         transmission in intake areas? Would the
in the scope? Should OSHA, for                past 12 months; and (3) is located in a        trigger of 6 individuals with confirmed
example, expand the scope of this             county that, in the past 2 years, has had      infectious TB be appropriate?
54164           Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules

   3. Are there methods other than             2. OSHA is requiring that                 tuberculosis? Should OSHA include
smoke trail testing and continuous          maintenance personnel use respiratory        provisions addressing supplied air
monitors that would be effective for        protection during maintenance of air         respirators in the standard?
verifying negative pressure in AFB          systems or equipment that may                  7. OSHA is permitting the reuse of
isolation rooms or areas?                   reasonably be anticipated to contain         disposable respirators provided the
   4. OSHA is requiring engineering         aerosolized M. tuberculosis. When            respirator does not exhibit excessive
controls to be inspected and                would it be necessary to access such an      resistance, physical damage, or any
performance monitored every 6 months.       air system at the time it was carrying air   other condition that renders it
Is this frequency appropriate?              that may contain aerosolized M.              unsuitable for use. Will the respirators
   5. OSHA is allowing exhaust air from     tuberculosis? Should OSHA require that       continue to protect employees
AFB isolation rooms or areas where M.       such air systems be purged and shut          throughout the reuse period?
tuberculosis may be aerosolized that        down whenever these systems are                8. In the proposed standard for TB,
cannot feasibly be discharged directly      accessed for maintenance or other            OSHA has included separate provisions
outside to be HEPA-filtered and             procedures?                                  for all aspects of a respiratory protection
recirculated back into general                 3. OSHA has received information          program for tuberculosis. What other
ventilation. Is permitting such             that the use of certain kinds of             elements might need to be included?
recirculation appropriate? If used,         respirators in helicopters providing         Which respiratory protection
should there be any requirements to         emergency medical services may               provisions, if any, are not appropriate
detect system failure?                      hamper pilot communication. Have             for protection against TB? Please
   6. OSHA is permitting stand-alone        other air ambulance services                 provide reasons and data to support
                                            encountered this problem? Does this          inclusion or exclusion of particular
HEPA filter units to be used as a
                                            problem exist when the employee is           provisions.
primary control measure. Is this
                                            using a type N95 respirator or other
appropriate? What, if any, methods                                                       Medical Surveillance
                                            types of respiratory protection such as
other than ventilation and filtration can
                                            powered air purifying respirators? What         1. Should any provisions be added to
provide consistent protection?
                                            other infection control or industrial        the Medical Surveillance program?
   7. Should ambulances that have           hygiene practices could be implemented
carried an individual with suspected or                                                     2. OSHA has not required that
                                            to minimize employee exposure in these       physical exams be included as part of
confirmed infectious TB be required to      circumstances?
be ventilated for a specific period of                                                   the baseline evaluation. Is there
                                               4. The CDC states that there may be
time or in a particular way before                                                       information that is essential to medical
                                            selected settings and circumstances
allowing employees to enter without a                                                    surveillance for TB that can only be
                                            (e.g., bronchoscopy on an individual
respirator? What engineering controls                                                    learned from a baseline physical exam?
                                            with suspected or confirmed infectious
are available for ambulances?               TB or an autopsy on a deceased                  3. OSHA is specifying tuberculin skin
                                            individual suspected of having had           testing frequencies for employees with
Laboratories                                                                             negative skin tests. Should tuberculin
                                            active TB at the time of death) where the
   1. The standard does not require         risk of transmission may be such that        skin testing be administered more or
labeling of laboratory specimens.           increased respiratory protection such as     less frequently? Are there other ways to
Should OSHA require that laboratory         that provided by a more protective           determine the frequency of tuberculin
specimens be labeled within the facility    negative-pressure respirator or a            skin testing?
or when specimens are being shipped?        powered air purifying respirator may be         4. OSHA is proposing that employees
If so, what should the label contain? Are   necessary. Are there circumstances           entering AFB isolation rooms or areas be
there other agencies that require these     where OSHA should require use of a           skin tested every 6 months. However,
specimens be labeled? What are these        respirator that is more protective than a    employees providing home health care,
agencies and what is required?              type N95 respirator? If so, what are the     home care, and home-based hospice
   2. OSHA has attempted to incorporate     circumstances and what type of               care are to be skin tested annually.
the CDC/NIH recommendations given in        respiratory protection should be             Employees entering the home of an
‘‘Biosafety in Microbiological and          required?                                    individual who has suspected or
Biomedical Laboratories’’ into the             5. OSHA is proposing that respirators     confirmed infectious TB may have the
standard. Do any provisions need to be      be fit-tested annually, which is             same potential for exposure to
added in order for employees in clinical    consistent with general industrial           aerosolized M. tuberculosis as
and research laboratories to be fully       hygiene practice, or, in lieu of an annual   employees who enter an isolation room.
protected against exposures to M.           fit test, that employees have their need     In light of this, should employees
tuberculosis?                               to receive the annual fit test be            providing care to individuals with
                                            evaluated by the physician or other          suspected or confirmed infectious TB in
Respirators                                                                              private homes be skin tested every 6
                                            licensed health care professional, as
   1. OSHA is requiring employees who       appropriate. For the circumstances and       months?
are transporting an unmasked                conditions regulated by this standard,          5. OSHA is requiring that all
individual with suspected or confirmed      will the evaluation provide enough           tuberculin skin testing be administered,
infectious TB within a facility to wear     ongoing information about the fit of a       read, and interpreted by or under the
a respirator. Is this appropriate? How      respirator to be an adequate substitute      supervision of a physician or other
often would an individual with              for fit testing? Should OSHA require         licensed health care professional, as
suspected or confirmed infectious TB be     that an actual fit test be performed         appropriate, according to current CDC
transported unmasked through a              periodically? If so, at what frequency?      recommendations. Should OSHA
facility? Under what circumstances             6. OSHA has not included any              require specific training for individuals
would it be infeasible to mask such an      provisions regarding the use of supplied     who are administering, reading, and
individual? What other precautions          air respirators. Are there circumstances     interpreting tuberculin skin tests? If so,
should be taken when transporting such      in which supplied air respirators would      what type of training should be
an individual who is not masked?            be used to protect against M.                required?
                Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules                              54165

   6. Should OSHA require a declination       alternative exists to permanent labeling     laboratories? Should such a distinction
form for employees who do not wish to         in situations where an exhaust duct          be made? Are there any modifications
undergo tuberculin skin testing?              from a room may or may not be carrying       that should be made to these
   7. OSHA is including Medical               air containing aerosolized M.                definitions?
Removal Protection (MRP) provisions           tuberculosis (e.g., the exhaust duct
for employees who are unable to wear                                                       B. Information Collection Requirements
                                              would only be carrying aerosolized M.
respiratory protection or who contract        tuberculosis when an individual with            This proposed Tuberculosis standard
infectious tuberculosis. Are there            infectious TB is being isolated in the       contains collections of information that
additional provisions that need to be         room)?                                       are subject to review by the Office of
included? What remedies are available                                                      Management and Budget (OMB) under
to employees in states where worker           Dates                                        the Paperwork Reduction Act of 1995
compensation system do not consider              1. OSHA has proposed that very small      (PRA’95), 44 U.S.C. 3501 et seq. and the
occupational TB a compensable disease?        businesses with fewer than 20                regulation at 5 CFR § 1320. PRA’95
What benefits are provided to workers         employees be given an additional 3           defines collection of information to
who are unable to wear a respirator?          months to comply with the standard’s         mean, ‘‘the obtaining, causing to be
   8. OSHA is requiring that employees        engineering control provisions (i.e., the    obtained, soliciting, or requiring the
who must wear a respirator be provided        start-up date for engineering controls for   disclosure to third parties or the public
a face-to-face determination of their         small businesses would be 270 days           of facts or opinions by or for an agency
ability to wear the respirator. Does this     from the Effective Date of the standard).    regardless of form or format.’’ [44 U.S.C.
determination need to be made through         Are there other requirements of the          § 3502(3)(A)].
a medical evaluation or would the use         proposed standard (e.g., respiratory            The title, description of the need for
of an appropriately designed                  protection) for which very small             and proposed use of the information,
questionnaire be adequate? What would         businesses should be given additional        summary of the collections of
be the advantages and disadvantages of        time to come into compliance? If so, for     information, description of the
relying on a questionnaire to make this       which provisions would they need             respondents, and frequency of response
determination? Are there sample               additional time and why? Are 20              of the information collection are
questionnaires that have proven to be         employees an appropriate cut-off for         described below with an estimate of the
effective for determining an employee’s       this purpose? Are there other employers      annual cost and reporting burden, as
ability to wear a respirator?                 that may need extended time to achieve       required by 5 CFR § 1320.5(a)(1)(iv) and
   9. OSHA has drafted Medical                compliance?                                  § 1320.8(d)(2). Included in the estimate
Surveillance, paragraph (g), to explain                                                    is the time for reviewing instructions,
first who must be provided with the                                                        gathering and maintaining the data
protections listed in the paragraph and          1. A number of provisions in the          needed, and completing and reviewing
how the surveillance is to be                 standard are triggered by the                the collection of information.
administered and secondly, in                 identification of an individual as having       OSHA invites comments on whether
paragraphs (g)(2), Explanation of Terms,      ‘‘suspected infectious tuberculosis.’’       the proposed collection of information:
and (g)(3), Application, how the general      Under the definition of ‘‘suspected             (1) Ensures that the collection of
medical terms are to be construed to          infectious tuberculosis’’, OSHA has          information is necessary for the proper
meet the standard and in what instances       proposed criteria that the Agency            performance of the functions of the
the medical examinations or tests are to      believes are the minimum indicators          agency, including whether the
be offered. The Agency realizes that          that, when satisfied by an individual,       information will have practical utility;
there is some repetition in these             require an employer to consider that the        (2) Estimates the projected burden
paragraphs and seeks comment on               individual may have infectious               accurately, including whether the
whether there might be a better way to        tuberculosis. Are there other criteria       methodology and assumptions used are
list the requirements.                        that should be included in this              valid;
                                              definition?                                     (3) Enhances the quality, utility, and
Communication of Hazards and                     2. Coverage of an employee under the      clarity of the information to be
Training                                      standard is based upon the definition of     collected; and
   1. OSHA is requiring that signs for        ‘‘occupational exposure.’’ Similar to           (4) Minimizes the burden of the
isolation rooms and areas bear a              OSHA’s Bloodborne Pathogens                  collection of information on those who
‘‘STOP’’ Sign and the legend ‘‘No             standard, occupational exposure is           are to respond, including through the
Admittance Without Wearing A Type             dependent upon reasonable anticipation       use of appropriate automated,
N95 or More Protective Respirator.’’ Is       of contact with an individual with           electronic, mechanical, or other
there another sign that would assure          suspected or confirmed infectious            technological collection techniques or
patient confidentiality while providing       tuberculosis or with air that may            other forms of information technology,
adequate notification of the hazard and       contain aerosolized M. tuberculosis. Are     e.g., permitting electronic submissions
the necessary steps to minimize the           there additions that could be made to        of responses.
hazard for employees who may be               this definition that would help                 Title: Tuberculosis 29 CFR 1910.1035.
inadvertently exposed?                        employers determine which of their              Description: The proposed
   2. OSHA is requiring that ducts be         employees are occupationally exposed?        Tuberculosis (TB) Standard is an
labeled ‘‘Contaminated Air—Respiratory           3. OSHA has proposed requirements         occupational safety and health standard
Protection Required.’’ Should OSHA            for research laboratories that differ from   that will prevent or minimize
require that duct labels also include the     those of clinical laboratories. The          occupational exposure to TB. The
‘‘STOP’’ sign?                                standard includes definitions of             standard’s information collection
   3. Is the labeling of ducts carrying air   ‘‘research laboratory’’ and ‘‘clinical       requirements are essential components
that may contain aerosolized M.               laboratory’’ to assist the employer in       that will protect employees from
tuberculosis (e.g., from isolation rooms      differentiating between these two types      occupational exposure. The information
and areas, labs) at all access points         of laboratory. Do the definitions clearly    will be used by employers and
feasible? What, if any, equally protective    differentiate between these two types of     employees to implement the protection
54166                      Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules

required by the standard. OSHA                                     performed; and laboratories that handle                 are segregated or otherwise confined
compliance officers will use some of the                           specimens that may contain M.                           due to having suspected or confirmed
information in their enforcement of the                            tuberculosis or process or maintain the                 infectious TB. Respondents also include
standard.                                                          resulting cultures, or perform related                  employers whose employees are
   Respondents: The respondents are                                activity that may result in the                         occupationally exposed during the
employers whose employees may have                                 aerosolization of M. tuberculosis.                      provision of emergency medical
occupational exposure in the following                               Also, occupational exposure                           services, home health care and home-
settings: hospitals; long-term care                                occurring during the provision of social                based hospice care. Approximately
facilities for the elderly; correctional                           work, social welfare services, teaching,
                                                                                                                           101,875 employers will be responding
facilities and other facilities that house                         law enforcement or legal services would
                                                                                                                           to the standard.
inmates or detainees; hospices; shelters                           be covered if the services are provided
for the homeless; facilities that offer                            in the work settings previously                           Total Estimated Cost: First year
treatment for drug abuse; facilities                               mentioned, or in residences, to                         $62,972,210; Recurring years
where high hazard procedures are                                   individuals who are in AFB isolation or                 $53,691,915.

                                                              SUMMARY OF THE COLLECTION OF INFORMATION
                                                            Number of                                                                                                 Total bur-
  Information collection requirement                                            Frequency of response                      Average time per response1
                                                            responses                                                                                                den (hours)

Exposure Control Plan:
   (c)(2)(i) ........................................         101,875    All Affected Employers to Develop            • 24 hours per Hospital ...................       906,980
                                                                           Plan.                                      • 8 hours per Facility for all Other
      (c)(2)(vii)(B) .................................        101,875    Annual Reviews and Updates for All           • 8 hours per Hospital .....................      238,243
                                                                           Affected Employers.                        • 2 hours per Facility for all Other
Respiratory Protection:
   (f)(2) ............................................         82,138    All Employers not Qualified for Ap-          • 8 hours per Hospital .....................      335,323
                                                                            pendix A Program to Develop Pro-          • 4 hours per Facility for all Other
                                                                            gram.                                       Industries
      (f)(5), Appendix B .......................             2,207,580   Initially, for all employees assigned        • 30 minutes per employee .............           551,962
                                                               22,078    Annual refit tests for 1% of popu-           • 30 minutes per employee .............              5,520
                                                                            lation assigned respirators.
      (f)(8) ............................................      82,138    Annual Evaluation of Program for All         • 2 hours per Hospital .....................       83,831
                                                                            Affected Employers not Qualified          • 1 hour per Facility for all Other In-
                                                                            for Appendix A Program.                     dustries
Medical Surveillance:
   • Medical History (g)(3)(i)(A) .....                      1,831,724   Initially for All Affected Employees ...     • 1 hour per Hospital Employee                  1,831,724
                                                                                                                        (inc. LHCP time).
                                                                                                                      • 1 hour per Employee in all Other
                                                                                                                        Industries (inc. travel time)
                                                             1,595,432   Annually for All Affected Employees          • 1 hour per Hospital Employee                  1,595,432
                                                                           in Facilities not Qualified for Ap-          (inc. LHCP time).
                                                                           pendix A.                                  • 1 hour per Employee in all Other
                                                                                                                        Industries (inc. travel time)
                                                               47,953    Initially, for New Employees .............   • 1 hour per Hospital Employee                     47,953
                                                                                                                        (inc. LHCP time).
                                                                                                                      • 1 hour per Employee in all Other
                                                                                                                        Industries (inc. travel time)
      • Medical Examination (inc. His-                         47,863    Annually, 3% of Controlled Popu-             • 2 hours per Hospital Employee in                 72,518
        tory and Physical) (g)(3)(i)                                       lation at Risk estimated to request          Facilities not Qualified for Appen-
        (B)–(D).                                                           exam as a result of having signs             dix A (inc. LHCP time).
                                                                           or symptoms of TB; have a TST              • 11⁄2 hour per Employee in All
                                                                           conversion; or indicated as a re-            Other Industries (inc. travel time)
                                                                           sult of an exposure incident.
      • Tuberculin Skin Tests
          Initial 2-Step TST (g)(3)(i)(A)                     474,627    Initially, for Entire Controlled Popu-       • 11⁄2 hours per Hospital Employee              1,026,377
                                                                            lation at Risk.                             (inc. LHCP time).
                                                                                                                      • 21⁄4 hour per Employee in All
                                                                                                                        Other Industries (inc. travel time)
            Exposure                         Incident            8,268   Annually, 2% of Controlled Popu-             • 11⁄2 hours per Hospital Employee                 17,879
              (g)(3)(i)(C).                                                lation at Risk in Facilities Qualified       (inc. LHCP time).
                                                                           for Appendix A.                            • 21⁄4 hour per Employee in All
                                                                                                                        Other Industries (inc. travel time)
            Pre-Exit (g)(3)(i)(E) ..............               76,257    Annually for Employment Turnover ..          • 1 hour for each Hospital Em-                    110,504
                                                                                                                        ployee (inc. LHCP time).
                                                                                                                      • 11⁄2 hour per Employee in All
                                                                                                                        Other Industries (inc. travel time)
            Prior to Initial Assignment ...                    76,257    All New Employees with Occupa-               • 11⁄2 hour per Hospital Employee                 165,756
                                                                           tional Exposure.                             (inc. LHCP time).
                          Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules                                                      54167

                                                   SUMMARY OF THE COLLECTION OF INFORMATION—Continued
                                                         Number of                                                                                           Total bur-
  Information collection requirement                                         Frequency of response                 Average time per response1
                                                         responses                                                                                          den (hours)

            Annual (g)(3)(ii)(A) ...............           413,400    All employees in facilities not quali-   • 1⁄2 hour per Hospital Employee                297,991
                                                                        fied for Appendix A.                     (inc. LHCP time).
                                                                                                               • 45 minutes per Employee in all
                                                                                                                 Other Industries (inc. travel time)
            Additional 6-month                 TST         131,367    All employees who:                       • 1 hour per Hospital Employee                  171,314
              (g)(3)(iii).                                            • Enter an AFB isolation room or           (inc. LHCP time).
                                                                         area                                  • 11⁄2 hour for each Employee in All
                                                                      • Perform or are present during the        Other Industries (inc. travel time)
                                                                         performance of high-hazard pro-
                                                                      • Transport or are present during
                                                                         the transport of an individual with
                                                                         suspected or confirmed infectious
                                                                         TB in an enclosed vehicle
                                                                      • Work in an intake area in facilities
                                                                         where 6 or more confirmed TB
                                                                         cases have been encountered in
                                                                         the past 12 mos
      • Information    Provided   to                      1,965,967   Information for each affected estab-     • 10 minutes per employee .............         327,661
        Licenced Health Care Profes-                                     lishment to provide a copy of the
        sional (LHCP) (g)(6)(I).                                         rule, and for information on each
                                                                         employee with a respirator.
                                                           558,549    Information for each new employee        • 10 minutes per employee .............          93,091
                                                                         assigned a respirator.
                                                            64,692    Information surrounding exposure in-     • 10 minutes per employee .............          10,782
                                                                         cidents (2% of controlled popu-
                                                                         lation at risk).
      • LHCP Written Opinion (g)(7) ..                    2,745,188   Initially, for each medical procedure    • 5 minutes per written opinion .......         228,766
                                                          2,034,269   Annually, for each medical proce-        • 5 minutes per written opinion .......         169,522
                                                                         dure performed.
     (h)(3)(ii)(B) ..................................      202,066    Number of training sessions in first     • 2 hours for employees required to             237,829
                                                                        year.                                     wear respirators.
                                                                                                               • 1 hour for employees with occu-
                                                                                                                  pational exposure who are not as-
                                                                                                                  signed respirators
                                                                                                               • Assumes 20 employees per ses-
      (h)(3)(ii)(A) ..................................     106,258    Number of training sessions for new      • For new employees: .....................       50,193
                                                                        employees entering affected occu-      2 hours for employees required to
                                                                        pations for the first time + number       wear respirators
                                                                        of training sessions for employees     1 hour for employees with occupa-
                                                                        staying in affected occupations,          tional exposure who are not as-
                                                                        but starting new jobs.                    signed respirators
                                                                                                               1⁄2 hours for employees required to

                                                                                                                  wear respirators
                                                                                                               15 minutes for employees with occu-
                                                                                                                  pational exposure who are not as-
                                                                                                                  signed respirators
      (h)(3)(ii)(C) ..................................     154,966    Recurring number of training ses-        • For 25% of exposed employees                   57,313
                                                                        sions.                                    unable to demonstrate com-
                                                                                                               1 hour for employees required to
                                                                                                                  wear respirators
                                                                                                               1⁄2 hour for employees with occupa-

                                                                                                                  tional exposure who are not as-
                                                                                                                  signed respirators
                                                                                                               • For 75% of exposed employees
                                                                                                                  able to demonstrate competence
                                                                                                               • Assumes 20 employees per ses-
   Medical (I)(1)(I) ...........................          3,713,645   Initially, to create a medical record    • 10 minutes to set up each record              631,320
                                                                         for each affected employee.
                                                          1,358,800   Create medical records for each          • 10 minutes to set up each record              230,996
                                                                         new employee with occupational
                                                          2,447,669   Annually, for each medical proce-        • 5 minutes to update each record               195,814
                                                                         dure performed.
54168                  Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules

                                               SUMMARY OF THE COLLECTION OF INFORMATION—Continued
                                                      Number of                                                                                                                                          Total bur-
  Information collection requirement                                                  Frequency of response                                     Average time per response1
                                                      responses                                                                                                                                         den (hours)

    Training (I)(3)(I) ..........................           264,451         Initially, to create records for each                         • 10 minutes to create each training                              44,957
                                                                               training session.                                            record.
                                                            217,351         Annually, to reflect recurring training                       • 10 minutes to create each training                              36,950
                                                                               sessions and initial training for                            record.
                                                                               new employees.
    Engineering controls (I)(4)(I) ......                     24,761        Annually, for each engineering con-                           • 5 minutes per record ....................                         3,962
    Availability (I)(5) ..........................              2,037       Annually, for 2% of affected employ-                          • 5 minutes per employer ................                            163
    Transfer to NIOSH ......................                           1    Annually, for estimated 1 employer                            • 1 hour per employer .....................                             1
                                                                               per year to transfer records.

              • First-Year ..................        ....................   ...........................................................   ...........................................................    7,098,011
              • Recurring ..................         ....................   ...........................................................   ...........................................................    3,655,728
   1 Estimates represent average burden hours per response. The actual burden hours per response will vary depending on factors such as the
size of the facility, current practices at the facility, and whether the facility transfers or admits individuals with suspected or confirmed infectious
   Note: Estimates take into account baseline compliance with the proposed requirements.

   The Agency has submitted a copy of                              The Order provides for preemption of                                             The proposed tuberculosis standard is
the information collection request to                              State law only if there is a clear                                            drafted so that employees in every State
OMB for its review and approval.                                   Congressional intent for the Agency to                                        will be protected by general,
Interested parties are requested to send                           do so. Any such preemption is to be                                           performance-oriented standards. To the
comments regarding this information                                limited to the extent possible.                                               extent that there are State or regional
collection to the Office of Information                              Throughout the development of this                                          peculiarities, States with occupational
and Regulatory Affairs, Attn. OSHA                                 proposed standard, OSHA has sought                                            safety and health plans approved under
Desk Officer, OMB New Executive                                    and received assistance from state                                            Section 18 of the OSH Act would be
Office Building, 725 17th Street NW,                               representatives. Representatives of state                                     able to develop their own State
Room 10235, Washington DC 20503.                                   departments of health and labor and                                           standards to deal with any special
   Comments submitted in response to                               industries have helped direct OSHA to                                         problems. Moreover, the performance
this notice will be summarized and/or                              pertinent information and studies on TB                                       nature of this standard, of and by itself,
included in the request for Office of                              and have submitted drafts of state                                            allows for flexibility by States and
Management and Budget approval of the                              standards relevant to TB. In addition,                                        employers to provide as much safety as
final information collection request:                              representatives of state occupational                                         possible using varying methods
they will also become a matter of public                           safety and health departments                                                 consonant with conditions in each
record.                                                            participated in the review of the draft                                       State.
   Copies of the referenced information                            standard by OSHA field offices and in                                            There is a clear national problem
collection request are available for                               OSHA’s TB Stakeholder meetings,                                               related to occupational safety and health
inspection and copying in the OSHA                                 where the requirements of the proposed                                        for employees exposed to M.
Docket Office and will be mailed                                   standard were presented and                                                   tuberculosis. Approximately 6.5% of the
immediately to any person who request                              information was collected from                                                U.S. adult population is infected (i.e.,
copies by telephoning Todd Owen at                                 employers, employees, and their                                               carrying the tuberculosis bacillus, not
(202) 219–7075. For electronic copies of                           representatives on what was being done                                        manifesting active disease), and
the Tuberculosis information collection                            to prevent occupational exposure to TB                                        although the prevalence of TB infection
request, contact the Labor News Bulletin                           in the various worksites and how an                                           and disease varies throughout the
Board (202) 219–4784, or OSHA web                                  OSHA standard for TB could further                                            country, TB disease has been reported
page on the Internet at http://                                    reduce the exposures.                                                         in every state. Political and geographic Copies of the                                         Section 18 of the Occupational Safety                                       boundaries do not contain infection and
information collection requests are also                           and Health Act (OSH Act), expresses                                           disease spread. The U.S. population is
available at the OMB docket office.                                Congress’ clear intent to preempt State                                       mobile, moving freely from place to
                                                                   laws with respect to which Federal                                            place for business and pleasure.
C. Federalism                                                      OSHA has promulgated occupational                                             Immigrants, a group whose members are
  This standard has been reviewed in                               safety or health standards. Under the                                         known to have a high prevalence of TB,
accordance with Executive Order 12612,                             OSH Act a State can avoid preemption                                          settle throughout the country. While
52 FR 41685 (October 30, 1987),                                    only if it submits, and obtains Federal                                       there are counties that do not report
regarding Federalism. This Order                                   approval of, a plan for the development                                       cases in a given year, the counties
requires that agencies, to the extent                              of such standards and their                                                   change from year to year along with the
possible, refrain from limiting State                              enforcement. Occupational safety and                                          number of cases reported. In addition,
policy options, consult with States prior                          health standards developed by such                                            reports do not always reflect all the
to taking any actions that would restrict                          State-Plan states must, among other                                           locations where exposure incidents can
State policy options, and take such                                things, be at least as effective in                                           occur; infectious TB cases are often
actions only when there is clear                                   providing safe and healthful                                                  transferred from their site of diagnosis
constitutional authority and the                                   employment and places of employment                                           to a distant location for treatment and
presence of a problem of national scope.                           as the Federal standards.                                                     reported as a TB case only in the county
                 Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules                               54169

where treatment is administered.                 A safety or health standard is a             Section 6(b)(5) also directs OSHA to
Finally, underreporting may occur             standard ‘‘which requires conditions, or     base health standards on ‘‘the best
because some individuals with                 the adoption or use of one or more           available evidence,’’ including research,
infectious TB, in particular the              practices, means, methods, operations,       demonstrations, and experiments. 29
homeless and clients of drug abuse            or processes, reasonably necessary or        U.S.C. § 655(b)(5). OSHA shall consider
facilities, do not avail themselves of        appropriate to provide safe or healthful     ‘‘in addition to the attainment of the
further diagnosis and treatment. TB           employment or places of employment.’’        highest degree of health and safety
infection and disease is truly national in    29 U.S.C. § 652(8).                          protection * * * the latest scientific
scope.                                           A standard is reasonably necessary or     data * * * feasibility and experience
   Those States which have elected to         appropriate within the meaning of            gained under this and other health and
participate under Section 18 of the OSH       Section 652(8) if it substantially reduces   safety laws.’’ Id.
Act would not be preempted by this            or eliminates significant risk, and is          Section 6(b)(7) authorizes OSHA to
regulation and would be able to deal          economically feasible, technologically       include among a standard’s
with special, local conditions within the     feasible, cost effective, consistent with    requirements labeling, monitoring,
framework provided by this                    prior Agency action or supported by a        medical testing and other information
performance-oriented standard while           reasoned justification for departing from    gathering and transmittal provisions. 29
ensuring that their standards are at least    prior Agency actions, supported by           U.S.C. § 655(b)(7).
as effective as the Federal standard.         substantial evidence, and is better able        Finally, whenever practical, standards
                                              to effectuate the Act’s purposes than any    shall ‘‘be expressed in terms of objective
D. State Plans                                                                             criteria and of the performance
                                              national consensus standard it
   The 23 States and 2 territories with                                                    desired.’’ Id.
                                              supersedes. See 58 Fed. Reg. 16612—
their own OSHA-approved occupational
                                              16616 (March 30, 1993).                      III. Events Leading to the Proposed
safety and health plans must adopt a
                                                 OSHA has generally considered, at a       Standard
comparable standard within 6 months
after the publication of a final standard     minimum, a fatality risk of 1/1000 over         Tuberculosis (TB) is a contagious
for occupational exposure to                  a 45-year working lifetime to be a           disease caused by the bacterium
tuberculosis or amend their existing          significant health risk. See the Benzene     Mycobacterium tuberculosis (M.
standard if it is not ‘‘at least as           standard, Industrial Union Dep’t v.          tuberculosis). Infection is usually
effective’’ as the final Federal standard.    American Petroleum Institute, 448 U.S.       acquired by the inhalation of airborne
OSHA anticipates that this standard will      607, 646 (1980); the Asbestos standard,      particles carrying the bacterium. These
have a substantial impact on state and        International Union, UAW v.                  airborne particles, called droplet nuclei,
local employees. The states and               Pendergrass, 878 F.2d 389, 393 (D.C.         can be generated when persons with
territories with occupational safety and      Cir. 1989).                                  infectious pulmonary or laryngeal TB
health state plans are: Alaska, Arizona,         A standard is technologically feasible    cough, sneeze, or speak. TB has long
California, Connecticut, Hawaii,              if the protective measures it requires       been considered an occupational hazard
Indiana, Iowa, Kentucky, Maryland,            already exist, can be brought into           in the health care setting. However, it is
Michigan, Minnesota, Nevada, New              existence with available technology, or      inhalation exposure to aerosolized M.
Mexico, New York, North Carolina,             can be created with technology that can      tuberculosis and not some other factor
Oregon, Puerto Rico, South Carolina,          reasonably be expected to be developed.      unique to the health care setting that
Tennessee, Utah, Vermont, Virginia, the       American Textile Mfrs. Institute v.          places workers at risk of infection. Thus,
Virgin Islands, Washington, and               OSHA, 452 U.S. 490, 513 (1981)               any work setting where employees can
Wyoming. (In Connecticut and New              (‘‘ATMI’’), American Iron and Steel          reasonably be anticipated to encounter
York, the plan covers only State and          Institute v. OSHA, 939 F.2d 975, 980         individuals with infectious TB also
local government employees). Until            (D.C. Cir. 1991)(‘‘AISI’’).                  contains the occupational hazard of TB
such time as a State standard is                 A standard is economically feasible if    infection.
promulgated, Federal OSHA will                industry can absorb or pass on the costs        On December 21, 1992, the Labor
provide interim enforcement assistance,       of compliance without threatening its        Coalition to Fight TB in the Workplace
as appropriate.                               long-term profitability or competitive       (the Coalition) requested the Agency to
                                              structure. See ATMI, 452 U.S. at 530 n.      issue nationwide enforcement
II. Pertinent Legal Authority                 55; AISI, 939 F.2d at 980.                   guidelines to protect workers against
   The purpose of the Occupational               A standard is cost effective if the       exposure to TB in health care, criminal
Safety and Health Act, 29 U.S.C. 651 et       protective measures it requires are the      justice, and other high risk settings and
seq. (‘‘the Act’’) is ‘‘to assure so far as   least costly of the available alternatives   to issue a Joint Advisory Notice on TB
possible every working man and woman          that achieve the same level of               in conjunction with the Centers for
in the nation safe and healthful working      protection. ATMI, 453 U.S. at 514 n. 32;     Disease Control and Prevention (CDC)
conditions and to preserve our human          International Union, UAW v. OSHA, 37         (Ex. 2). This petition was signed by the
resources.’’ 29 U.S.C. § 651(b). To           F.3d 665, 668 (D.C. Cir. 1994) (‘‘LOTO       presidents of the Service Employees
achieve this goal Congress authorized         III’’).                                      International Union (SEIU), the
the Secretary of Labor to promulgate             All standards must be highly              American Federation of State, County,
and enforce occupational safety and           protective. See 58 FR 16614—16615;           and Municipal Employees (AFSCME),
health standards. 29 U.S.C. §§ 655(a)         LOTO III, 37 F.3d at 669. However,           and the American Federation of
(authorizing summary adoption of              health standards must also meet the          Teachers (AFT), and was endorsed by 9
existing consensus and federal                ‘‘feasibility mandate’’ of Section 6(b)(7)   other unions. The petition included a
standards within two years of Act’s           of the Act, 29 U.S.C. § 655(b)(5). Section   list of provisions that the petitioners felt
enactment), 655(b) (authorizing               6(b)(5) requires OSHA to select ‘‘the        should be included in the guidelines,
promulgation of standards pursuant to         most protective standard consistent          ranging from a written control plan and
notice and comment), 654(b) (requiring        with feasibility’’ that is needed to         medical surveillance to anti-
employers to comply with OSHA                 reduce significant risk when regulating      discrimination language and medical
standards).                                   health hazards. ATMI, 452 U.S. at 509.       removal protection.
54170           Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules

   Eight months later, on August 25,         requires accident prevention tags to          from small business employers. This
1993, the Coalition petitioned OSHA to       warn of biological hazards. In addition,      report was submitted to the Assistant
initiate rulemaking for a permanent          section 5(a)(1), the General Duty Clause      Secretary for OSHA for its consideration
standard issued under § 655(b) of the        of the Act, requires that each employer:      during the development of the standard
Act to protect workers from                    * * * furnish to each of his employees      (Ex. 12). OSHA’s proposed standard
occupational transmission of TB (Ex. 1).     employment and a place of employment          reflects input generated during both the
Citing the recent resurgence of TB and       which are free from recognized hazards that   stakeholder meetings and the SBREFA
the emergence and increasing rate of         are causing or are likely to cause death or   review process.
new cases of multidrug-resistant TB          serious physical harm to his employees.
                                                                                           Comparison of OSHA’s Proposed
(MDR–TB), the petitioners stressed the          On January 26, 1994, in response to        Standard and CDC’s Revised Guidelines
need for a substance-specific standard to    their August 25 petition, Secretary of
address the hazards associated with                                                           In preparing its proposed standard for
                                             Labor Robert B. Reich informed the            TB, OSHA has relied heavily on the
occupational exposures to TB. The            petitioners that OSHA was initiating
petitioners contended that the non-                                                        expertise of CDC. The Agency has
                                             rulemaking on a permanent standard to         consulted with CDC and has
mandatory CDC TB Guidelines do not           be issued under Section 6(b)(5) of the
provide adequate protection because                                                        incorporated the basic elements of
                                             Act for occupational exposure to TB (Ex.      CDC’s revised guidelines for preventing
they are not fully or rigorously             1B). At the same time, the petitioner’s
implemented in most workplaces. They                                                       the transmission of M. tuberculosis in
                                             request for an ETS was denied. The            health care facilities in this proposed
also stated that in every outbreak of TB     Agency had determined that the
investigated by CDC, noncompliance                                                         standard. Both CDC and OSHA rely on
                                             available data did not meet the criteria      minimizing exposures and consequent
with the Guidelines was evident.
                                             for an ETS as set forth in Section 6(c)       transmission by identifying suspected
   In addition to a permanent standard,
the petitioners also requested that          of the Act. However, OSHA committed           infectious TB individuals and isolating
OSHA immediately issue the                   to enforcing existing regulations and         them. The OSHA proposed standard
nationwide enforcement guidelines that       Section 5(a)(1) of the Act in certain         includes the following CDC
the Coalition had previously requested,      work settings while preparing this            components: written exposure control
and that OSHA promulgate an                  standard.                                     plans, procedures for early
Emergency Temporary Standard (ETS)              On October 28, 1994 the CDC issued         identification of individuals with
as an interim measure. The Coalition         revised guidelines for preventing the         suspected or confirmed infectious TB,
requested that the standard be               transmission of tuberculosis in health        procedures for initiating isolation of
applicable to all work settings where        care facilities (Ex. 4B). In addition, in     individuals with suspected or
employees can reasonably anticipate          June of 1995, the National Institute for      confirmed infectious TB or for referring
contact with infectious TB. The petition     Occupational Safety and Health                those individuals to facilities with
included a discussion on occupational        (NIOSH) published revised certification       appropriate isolation capabilities,
risk that included both the traditional      procedures for non-powered air                procedures for investigating employee
high-risk occupations and other              purifying particulate respirators (Ex. 7–     skin test conversions, and education
occupations such as sheet metal              261). As a result of changes in these two     and training for employees. In addition,
workers, postal workers, airline             documents, OSHA issued revised                OSHA has incorporated CDC
employees, teachers, and office workers.     enforcement policies and procedures           recommendations for engineering
   Like the request for nationwide           relative to TB in February of 1996 (Ex.       control measures such as the use of
enforcement guidelines, the petition         7–260).                                       negative pressure for AFB isolation
contained provisions that the petitioners       In October and November of 1995,           rooms or areas, daily monitoring of
requested be included in the standard.       OSHA held a series of meetings with           negative pressure while AFB isolation
Examples include a facility hazard           stakeholder groups representing labor         rooms are in use for TB, HEPA filtration
assessment and written exposure              unions, professional organizations, trade     of recirculated air from AFB isolation
control plan, engineering and work           associations, state and federal               rooms, and periodic maintenance and
practice controls, respiratory protection,   government, representatives of                monitoring of engineering controls.
medical surveillance (e.g., tuberculin       employers, as well as frontline workers       With regard to respiratory protection,
skin testing) and counseling, post-          from the various sectors anticipated to       OSHA has adopted CDC’s standard
exposure management, outbreak                be covered by the proposed standard.          performance criteria for the selection of
management, training, and                    During these meetings, participants           respiratory protection devices
recordkeeping.                               provided input relative to the concepts       appropriate for use against M.
   On October 8, 1993, OSHA issued           and approaches OSHA was considering           tuberculosis. And finally, where
nationwide enforcement procedures for        for the proposed tuberculosis standard.       appropriate, OSHA has attempted to
occupational exposure to TB. The                In September of 1996, in accordance        assure that where certain practices are
compliance document contained the            with the Small Business Regulatory            required by OSHA’s proposed standard,
enforcement procedures that the Agency       Enforcement Fairness Act of 1996              e.g., tuberculin skin testing and medical
could and would use in certain work          (SBREFA), a Small Business Advocacy           management and follow-up of
settings for protecting workers with         Review Panel was convened to consider         employees who acquire TB infections or
occupational exposure to TB. In the          the impact of OSHA’s draft proposed           active disease, these practices are
compliance procedures, the Agency            tuberculosis standard on affected small       conducted according to the current
noted that although OSHA has no              entities. The panel, comprised of             recommendations of the CDC.
standard designed specifically to reduce     members from the Office of Advocacy of        Therefore, OSHA’s proposed standard
occupational exposure to TB, the             the Small Business Administration             for occupational exposure to TB closely
Agency has existing standards that           (SBA), the Office of Management and           follows CDC’s recommended elements
apply to this hazard. For example, 29        Budget (OMB), and OSHA, prepared a            for a TB infection control program.
CFR 1910.134 requires employers to           report based on the Panel’s findings and         However, there are some minor
provide respiratory protection               recommendations with regard to                differences between OSHA’s proposed
equipment and 29 CFR 1910.145(f)             comments on the standard received             standard and CDC’s guidelines that go
                Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules                               54171

beyond the obvious enforcement              burdensome for employers to prepare.           testing for all employees whom the
distinction between a guideline and a       Also, many work settings will have too         employer identifies as having
standard. These differences are found       few occupationally exposed employees           occupational exposure. CDC
primarily in the areas of risk              to do an accurate risk assessment.             recommends baseline skin testing for all
assessment, medical surveillance and        Finally, conducting the risk assessments       employees with potential exposure
respiratory protection. Even so, OSHA       in order to determine applicable duties        except those who work in facilities that
believes that despite these differences     may require a level of expertise some          fall into CDC’s ‘‘minimal risk’’ category.
the vast majority of the provisions         facilities lack, making enforcement            However, CDC notes that even for
included in this proposed standard          burdensome for the Agency.                     employees in ‘‘minimal risk’’ facilities,
closely track the recommendations of           OSHA realizes, however, that in many        it may be advisable to perform baseline
the CDC. The following discussion           work settings, very few individuals with       skin testing so that if unexpected
identifies where these differences occur    suspected or confirmed infectious TB           exposures do occur, conversions can be
and describes the extent of these           may be seen and that in many of those          distinguished from positive skin test
differences and the degree to which they    work settings, individuals with                results caused by previous exposures.
impact on employers’ responsibilities       suspected or confirmed infectious TB           Thus, there is little difference between
under the proposed standard.                will be transferred to other facilities that   OSHA requirements and CDC
                                            are better equipped to provide services        recommendations with regard to
Risk Assessment                             and care using appropriate TB isolation        baseline skin testing.
   As a part of its guidelines, CDC         precautions. Because there is likely to           Relative to periodic skin testing,
recommends that a risk assessment be        be less risk of transmission of M.             OSHA requires periodic re-testing for all
conducted in all facilities to assess the   tuberculosis in those situations, OSHA         employees identified as having
risk of transmission of M. tuberculosis     believes that it is possible to make the       occupational exposure who have
in each facility. This risk assessment is   standard less burdensome for the               negative skin tests except for the
to be conducted using information such      employers with these types of work             employees of those employers who have
as the profile of TB in the community,      settings while still maintaining worker        no TB in the community and who have
the number of suspected and confirmed       protection.                                    not encountered any individuals with
cases of TB among patients and health          For example, an employer who can            confirmed infectious TB in their work
care workers, results of health care        demonstrate that his or her facility or        settings within the past year. CDC
worker tuberculin skin testing (i.e.,       work setting: (1) Does not admit or            recommends re-testing for employees in
conversion rates), and observation of TB    provide medical services to individuals        the ‘‘low’’, ‘‘intermediate’’, and ‘‘high’’
infection control practices. Using the      stwith suspected or confirmed                  risk categories. According to the CDC
results of this risk assessment,            infectious TB, (2) has not had any             guidelines, periodic re-testing is not
appropriate infection control               individuals with confirmed infectious          necessary for employees in the
interventions can then be selected based    TB within the work setting within the          ‘‘minimal’’ risk category or the ‘‘very-
on the actual risk in the facility. CDC     last 12 months, and (3) is located in a        low’’ risk categories. CDC’s periodic
includes a protocol for conducting this     county that, in the past 2 years, has had      skin test recommendations for the
risk assessment in which there are 5        0 cases of confirmed infectious TB             ‘‘minimal’’ risk category are similar to
categories of risk: ‘‘minimal’’, ‘‘very-    reported in one year and fewer than 6          OSHA’s limited program for employers
low’’, ‘‘low’’, ‘‘intermediate’’, and       cases of confirmed infectious TB               who do not admit or provide medical
‘‘high’’. Each category from ‘‘minimal’’    reported in the other year, does not have      services to individuals with suspected
to ‘‘high’’ has an increasing number of     to comply with all provisions of the           or confirmed infectious TB, have not
infection control interventions that are    standard. Such employers would only            encountered any confirmed infectious
recommended for each particular level       be responsible for compliance with             TB in their work setting, and are located
of risk.                                    certain provisions, e.g., a written            in a county that, in the past 2 years, has
   OSHA, however, has chosen a simpler      exposure control plan, a baseline skin         reported 0 cases of confirmed infectious
approach and is not requiring employers     test and medical history, medical              TB in one year and fewer than 6 cases
to conduct such a risk assessment.          management and follow-up after                 in the other year. OSHA is different
Consistent with other standards, OSHA       exposure incidents, medical removal            from the CDC in that employees in a
has determined that employees in the        protection where necessary, employee           ‘‘very-low risk category’’ are required to
work settings and employees providing       training, and recordkeeping. These             be periodically retested. However, CDC
services set forth in the scope section     provisions are very similar to the             notes that even in the ‘‘very-low’’ risk
are at risk of occupational exposure to     recommendations of the CDC for                 category, employees who are involved
TB. Their employers are required to         facilities classified as having ‘‘minimal      in the initial assessment of individuals
conduct an exposure assessment to           risk,’’ i.e., no TB in the community or        in emergency departments and
determine which employees have              in the facility. The only major difference     admitting areas may have potential
occupational exposure, i.e., reasonably     is that CDC does not recommend                 exposure and thus may need periodic
anticipated contact with an individual      baseline skin testing. However, CDC            re-testing.
with suspected or confirmed infectious      does state that baseline skin testing             Another difference between CDC and
TB or air that may contain aerosolized      would be advisable so that if an               OSHA is the frequency of the re-testing.
M. tuberculosis. The standard then          unexpected exposure does occur,                This is primarily due to the fact that
specifies the provisions applicable for     conversion could be distinguished from         OSHA’s required frequencies are based
the employees whom the employer has         positive skin test results caused by           on the type of work that employees do
identified as having occupational           previous exposures.                            that result in exposures whereas CDC’s
exposure. In addition, consistent with                                                     recommendations are based more on
its approach in other standards, OSHA       Medical Surveillance                           evidence of conversions. For example,
does not require that individual risk         In the area of medical surveillance,         OSHA requires re-testing every six
assessments be conducted by each work       the main differences between OSHA              months for all employees who (1) enter
setting covered under the standard, as      and CDC are related to tuberculin skin         AFB isolation rooms or areas, (2)
they may be too difficult and               testing. OSHA requires baseline skin           perform high-hazard procedures, (3)
54172           Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules

transport individuals with suspected or     of event that would cause an                infectious TB in an enclosed vehicle
confirmed infectious TB in an enclosed      employee(s) to be included in the           and when working in a residence where
vehicle, or (4) work in intake areas        ‘‘high’’ risk category as defined by CDC,   an individual with suspected or
where early identification procedures       OSHA requirements, to some extent,          confirmed infectious TB is known to be
are performed (e.g., emergency              track the CDC recommendations for a         present. However, OSHA also requires
departments, admitting areas) in            higher frequency of periodic skin           that respirators be worn when
facilities where 6 or more individuals      testing.                                    employees are transporting individuals
with confirmed infectious TB have been         With regard to two-step testing, both    with suspected or confirmed infectious
encountered in the past 12 months. For      OSHA and CDC require or recommend           TB within the facility if those
all other employees with occupational       two-step testing at the time baseline       individuals are not masked (e.g., a
exposure, re-testing is required every 12   skin testing is administered. Also, both    surgical mask or a valveless respirator).
months. In comparison, CDC                  OSHA and CDC add that two-step              CDC does not have a similar
recommends re-testing every year for        testing is not necessary if the employee    recommendation for respiratory
employees in ‘‘low’’ risk categories,       has had a documented negative skin test     protection while transporting
every 6–12 months for employees in          within the last 12 months. CDC is           individuals within the facility, but CDC
‘‘intermediate’’ risk categories, and       different from OSHA in that its             does recommend, and assumes to some
every 3 months for employees in ‘‘high’’    Guidelines imply that two-step testing      extent, that individuals with suspected
risk categories. Under CDC                  can be discontinued if there is evidence    or confirmed infectious TB are masked
recommendations, employees in ‘‘low’’       of a low frequency of boosting in the       whenever they are outside an isolation
risk categories who enter AFB isolation     facility. OSHA’s proposed standard does     room. In addition, OSHA requires that
rooms or areas or employees who             not allow such an exemption, i.e., for      respirators be worn when employees
transport individuals with suspected or     each employee who must have a               work in an area where an unmasked
confirmed infectious TB in an enclosed      baseline skin test at the time of the       individual with suspected or confirmed
vehicle would be re-tested every 12         initial medical examination, the skin       infectious TB has been segregated or
months. However, under OSHA                 test must include a two-step test unless    otherwise confined. For example, this
requirements, those same employees          the employee has a documented               provision would cover employees such
would be required to be re-tested every     negative test within the last 12 months,    as those who work in admitting areas
six months. Thus, OSHA is more              regardless of the frequency of boosting     and must attend to unmasked
protective than CDC in this case.           in the facility. The value of two-step      individuals with suspected or
   OSHA also would require that             skin testing is that it enables one to      confirmed infectious TB while those
employees who perform high-hazard           distinguish true conversions from           individuals are awaiting transfer. These
procedures or who work in intake areas      boosted reactions. OSHA believes that       types of employees are likely to be
where early identification procedures       this is important to know for each          found in facilities that would meet
are performed in facilities that            employee because if the employee is         CDC’s definition of ‘‘minimal’’ risk. CDC
encounter 6 or more individuals with        incorrectly identified as having            states that respiratory protection is not
confirmed infectious TB be re-tested        converted, he or she may needlessly be      necessary for employees in the
every six months. Under CDC’s               subjected to preventive therapy that        ‘‘minimal’’ risk category. However,
Guidelines employees in areas in which      may have toxic side effects of its own.     again, CDC recommends that if an
cough-inducing procedures are               Since it is important to know the true      individual with suspected or confirmed
performed on individuals who may            skin test status for each employee,
                                                                                        infectious TB is identified in a
have active TB are recommended to           OSHA has preliminarily concluded that
                                                                                        ‘‘minimal’’ risk facility, the individual
follow an intermediate risk protocol.       it is inappropriate to allow the overall
                                                                                        should be masked while he or she is
Similarly, CDC recommends that an           frequency of boosting among employees
                                                                                        awaiting transfer to another facility,
intermediate risk protocol be followed      in a facility to dictate whether any one
                                                                                        thus obviating the need for respiratory
in areas where more than six                employee receives two-step testing at
                                                                                        protection. OSHA, on the other hand,
individuals who may have active TB          the time of his or her baseline testing.
                                                                                        cannot require employers to mask
receive initial assessment and
                                            Respiratory Protection                      clients or patients in a facility, and the
diagnostic evaluation (e.g., ambulatory
                                               OSHA requirements and CDC                Agency must therefore include
care, emergency departments, admitting
areas). CDC recommends re-testing           recommendations for respiratory             provisions for respirator use to protect
every 6–12 months for employees in          protection are very similar. A respirator   potentially exposed employees.
intermediate risk categories. OSHA          is a personal protective equipment          However, consistent with CDC, OSHA
would require re-testing every 6 months     device worn over the nose and mouth of      proposes not to require respirators
for the two situations above, which is      the employee that filters certain           where the employer elects, as a part of
very similar to CDC’s recommendation        airborne contaminants from the inhaled      his or her own administrative policies,
of re-testing every 6–12 months.            air. OSHA has adopted CDC’s                 to mask individuals with suspected or
   CDC is more protective in its            performance criteria for respirators        confirmed infectious TB. Thus, when
recommendations for employees in the        appropriate for use for TB. Also, both      individuals with suspected or
‘‘high’’ risk category. These employees     OSHA and CDC have similar                   confirmed infectious TB are masked
are recommended to be re-tested every       requirements or recommendations that        while they are awaiting transfer to
3 months. OSHA does not have a              respirators be worn when entering an        another facility or while they are being
requirement for re-testing employees        isolation room, when performing cough-      transported within the facility,
every 3 months. However, after an           inducing procedures or aerosol-             employees would not be required by the
exposure incident, OSHA requires that       generating procedures on an individual      standard to wear a respirator.
a skin test be administered as soon as      with suspected or confirmed infectious         In some instances, the CDC may be
feasible and again 3 months after the       TB, when repairing or maintaining air       more protective than OSHA with regard
exposure incident, if the first skin test   systems that may contain aerosolized M.     to respiratory protection. The CDC states
is negative. Since it is possible that an   tuberculosis, when transporting an          that the facility’s risk assessment may
exposure incident(s) could be the type      individual with suspected or confirmed      identify selected settings where the
                Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules                             54173

estimated risk of transmission of M.         active TB cases in the U.S. From 1953,       resources have been invested to improve
tuberculosis may be such that a level of     when active cases began to be reported       or initiate TB control provisions, such
respiratory protection exceeding the         in the U.S., until 1984, the number of       as those outlined in OSHA’s proposed
standard performance criteria is             annual reported cases declined 74%,          standard. However, the 1995 statistics
appropriate (e.g., more protective           from 84,304 (53 per 100,000) to 22,255       show that over the course of four years
negative pressure respirators, powered       (9.4 per 100,000) (Ex. 7–50). However,       there is substantial variability in the
air purifying respirators). The examples     this steady decline in TB cases did not      increases and decreases of cases
given of such selected settings are a        continue. Instead, from 1985 through         reported by each state for any given year
bronchoscopy performed on an                 1992, the number of reported TB cases        (Ex. 6–34). In 1995, 15 states reported an
individual suspected of having TB and        increased 20.1% from 22,201 to 26,673        increase in the number of TB cases
an autopsy performed on a deceased           (10.5 cases per 100,000) (Ex. 6–13).         compared with 1994. In addition, a
person suspected of having had active           This resurgence in TB brought to          recent study has shown that MDR–TB
TB at the time of death. OSHA does not       attention a number of problems in the        has spread to patients in Florida and
have a similar requirement for more          existing TB control programs. The            Nevada, and to health care workers in
protective respiratory protection.           direction of resources to areas with the     Atlanta, Georgia and Miami, Florida.
Respirators meeting the minimal              highest increase in active cases has         Moreover, one individual with MDR–TB
performance criteria laid out by the         caused this increase to decline. The         infected or caused disease in at least 12
standard would be required by OSHA           number of cases reported for 1995            people in a nursing home in Denver,
for employees performing all high-           indicates that the rate of active TB has     Colorado (Ex. 7–259). This study shows
hazard procedures, including                 returned to its 1985 levels. In 1995, a      very clearly the ability of TB to be
bronchoscopies and aerosol-generating        total of 22,813 cases of TB (8.7 per         spread to different areas of the country.
autopsy procedures.                          100,000) was reported to CDC (Ex. 6–         This is to be expected given the mobile
                                             34). While this represents a decline in      nature of today’s society and the
IV. Health Effects                           active TB, the 1995 rate is still two and    frequency with which people travel.
Introduction                                 one half times greater than the target       Immigration also contributes to the
                                             case rate of 3.5 per 100,000 for the year    incidence of the disease. For example,
   For centuries Tuberculosis (TB) has       2000 and approximately 87 times the          while the number of active TB cases has
been responsible for the death of            goal of less than one case per million       decreased among U.S. born persons, the
millions of people throughout the            population by the year 2010 proposed         number of foreign born persons reported
world. It was not until 1882, however,       by the Advisory Committee on the             with TB has increased 63% since 1986,
that Robert Koch identified a species of     Elimination of Tuberculosis (Ex. 6–19).      with a 5.4% increase in 1995 (i.e., from
bacteria, Mycobacterium tuberculosis            TB continues to be a national             7,627 cases in 1994 to 8,042 cases in
(M. tuberculosis), as the cause of TB.       problem. Each year, cases of active          1995). Thirty to fifty percent of these
   TB is a communicable disease that         disease are reported in every state in the   cases were diagnosed 1 to 5 years after
usually affects the lungs. The airborne      Nation and in a substantial majority of      the individual enters the U.S. (Ex. 6–
route is the predominant mode of             counties nationwide. CDC estimated in        34). Thus, tuberculosis continues to be
transmission, a situation created when       1990 that approximately 10 million           a public health problem throughout the
individuals with infectious TB               people were infected with the                United States.
discharge the bacilli from the lungs         tuberculosis bacterium and that                 The following discussion will briefly
when coughing, sneezing, speaking or         approximately 90% of the new cases of        describe the basic concepts and
singing. Some individuals who breathe        active disease that arise in the United      terminology associated with TB as well
contaminated air become infected with        States come from this already infected       as common factors that facilitate its
TB. Most often, the immune system            group (Ex. 7–52). Given the recent           transmission from one individual to
responds to fight the infection. Within      resurgence of TB, it is likely that a new    another. This discussion will also
a few weeks, the infected lesions            population of individuals has been           include a review of studies relating to
become inactive and there is no residual     infected as well. Of great concern are       the occupational transmission of TB.
change except for possible lymph node        strains of M. tuberculosis that have
calcifications. These individuals will       emerged that are resistant to several of     Background
have a positive skin test result. They       the first-line anti-TB drugs normally           TB is a contagious disease caused by
will harbor the infection for life. At       used to treat TB infection and disease       the bacterium M. tuberculosis. Infection
some time in the future, the infection       (e.g., isoniazid and rifampin). This drug-   is generally acquired by the inhalation
can progress and can become an active        resistant form of the disease, referred to   of airborne particles carrying the
disease, with pulmonary infiltration,        as multidrug-resistant TB or MDR–TB,         bacterium. These airborne particles,
cavitation, and fibrosis, possibly causing   is more often a fatal form of TB due to      called droplet nuclei, can be generated
permanent lung damage and even death.        the difficulty in finding antimicrobial      when persons with pulmonary or
With some exceptions, however, TB is         drugs to stop the bacteria’s growth and      laryngeal tuberculosis in the infectious
treatable with antimicrobial drugs. If the   progressive tissue destruction. In           state of the disease cough, sneeze, speak
active TB is treated early, there will be    addition, individuals with MDR–TB            or sing.
minimal residual lung damage. For this       often remain infectious for longer              In some individuals exposed to
reason, individuals who have a TB            periods of time due to delays in             droplet nuclei, tuberculosis bacilli enter
exposure incident and develop a TB           diagnosing resistance patterns and           the lung and establish an infection (Ex.
infection are treated to prevent             initiating appropriate treatment. This, in   7–52). Once in the alveoli, the
progression to active TB disease.            turn, increases the risk that infectious     tuberculosis bacilli are taken up by
   With the introduction of antimicrobial    individuals will transmit the organism       alveolar macrophages and spread
drug treatment in the 1940s and the          to other persons coming in contact with      throughout the body by the lymphatic
creation of programs in the United           them.                                        system, until the immune response
States such as the U.S. Public Health           Most of the decreases in reported         limits further growth (usually a period
Service’s Tuberculosis Program, there        cases of TB since 1992 have occurred in      of two to ten weeks). In most cases the
began a decline in the incidence of          areas such as New York City, where           tuberculosis bacilli are contained by the
54174           Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules

immune response. Macrophage cells             who do not have active disease. In these     diagnosis of clinically active TB is most
engulf the bacteria, which limits the         cases, the antimicrobials are used as        definitively established by the isolation
spread of the bacilli. Initial lesions from   preventive therapy to prevent the onset      of M. tuberculosis in culture. However,
infection heal; however, small                of active disease. Because of the toxicity   it may take three to six weeks or longer
calcifications called tubercles are           associated with the antimicrobials,          from obtaining a culture to getting a
formed and may remain a potential site        preventive therapy may not be                result.
of later reactivation.                        appropriate for all infected individuals.       Individuals with active TB disease
   Individuals in this state are infected     Various factors are considered to            may be infectious, especially if they are
with TB. They will show a positive skin       determine whether an infected                untreated or inadequately treated and if
test and they are at risk of developing       individual is an appropriate candidate       the disease is in the lungs. The clinical
active TB, a risk they carry throughout       for preventive therapy (e.g., age,           symptoms of pulmonary TB include loss
their lifetime. In many cases, as             immune status, how recently the              of appetite, weight loss, fatigue, fever,
described below, preventive therapy is        infection occurred, and other high-risk      night sweats, malaise, cough with
initiated with anti-TB drugs to prevent       factors associated with TB) (Ex. 7–52,       productive sputum and/or blood, and
the progression to active TB disease.         pg. 17). Isoniazid is currently the only     chest pain. The extent of the disease
These drugs are toxic and may cause           drug that has been well tested in            varies from very minimal symptoms to
adverse effects such as hepatitis. Severe     humans for its efficacy as preventive        extensive debilitating constitutional and
preventive therapy-associated hepatitis       therapy (Ex. 7–50, pg. 61). However,         respiratory symptoms. If untreated, the
cases have necessitated liver transplants     serious side effects may result from         pulmonary TB follows a chronic and
and in some cases have resulted in            isoniazid. A study in New York for the       progressive course in which the tissue is
death (Ex. 6–10).                             years 1991 to 1993 examined cases of         progressively destroyed. It has been
   When the bacilli are not contained by      hepatitis induced by isoniazid               estimated that approximately 40 to 60%
the immune system, they continue to           preventive therapy. In this study, 10        of untreated cases result in death (Exs.
grow and invade the tissue, leading to        patients undergoing preventive therapy       5–80, 7–50, and 7–66). However, even
the progressive destruction of the organ      for TB were identified at a transplant       among cured cases of TB, long-term
involved, which in most cases is the          center. Eight of these patients had          damage can result, including impaired
lung, i.e., pulmonary tuberculosis. The       developed hepatitis from isoniazid. Five     breathing due to lung damage (Ex. 7–50,
inflammatory response caused by the           received a liver transplant; the other       pg. 31).
disease produces weakness, fever, chest       three died while awaiting a liver donor.        Approximately 90% of
pain, cough, and, when blood vessels          In addition, one of the transplant           immunocompetent adults who are
are eroded, bloody sputum. Also, many         patients died after transplantation.         infected do not develop active TB
individuals have drenching night sweats       Thus, preventive therapy may carry           disease. However, for 10% of infected
over the upper half of the body several       considerable risks for infected              immunocompetent adults, either
times a week (Ex. 5–80). The extent of        individuals.                                 directly after infection or after a latency
disease varies from minimal symptoms             In those cases where isoniazid cannot     period of months, years or even
of disease to massive involvement with        be tolerated by the patient or where it      decades, the initial infection progresses
extensive cavitation and debilitating         is suspected that infection resulted from    to clinical illness, that is, active TB (Ex.
constitutional and respiratory                exposure to isoniazid-resistant strains of   4B). The risk of developing active TB is
symptoms. Since tuberculosis bacilli are      M. tuberculosis, rifampin may be             increased for individuals whose
spread throughout the body after the          recommended for preventive therapy.          immune system is impaired (i.e.,
initial infection, other organs may also      Considerations for such alternative drug     immunocompromised). Such
be infected and disease may occur at          therapies are made on a case-by-case         individuals include persons undergoing
sites outside the lung, i.e.,                 basis by the health care provider based      treatment with corticosteroid or
extrapulmonary tuberculosis.                  on the medical and case history of the       immunosuppressive drugs (e.g., persons
   There are two general stages of TB,        infected patient. Rifampin has adverse       with organ transplants or persons
tuberculosis infection and active             side effects as well. However,               undergoing chemotherapy for cancer),
tuberculosis disease. Individuals with        preventive therapy using rifampin has        persons suffering from malnutrition or
tuberculosis infection and no active          not been followed as well as that            chronic conditions such as asthma and
disease are not infectious. These             involving isoniazid and therefore, its       emphysema, and persons infected with
tuberculosis infections are                   side effects are less well characterized.    the human immunodeficiency virus
asymptomatic or subclinical and are              Individuals with active TB have           (HIV).
only detected by a positive response to       clinical and/or radiographic evidence of        The main first-line drugs currently
a tuberculin skin test. However, there        disease. The initial laboratory method       used to treat active TB are isoniazid,
are some individuals whose immune             for diagnosing TB is the Acid Fast           rifampin, pyrazinamide, ethambutol and
system is impaired and cannot mount a         Bacilli (AFB) smear. This is a quick and     streptomycin. Combinations of these
sufficient response to skin test antigens,    easy technique in which body fluids,         antimicrobials are used to attack the
i.e., they are anergic. Such individuals      typically sputum samples, from               tuberculosis bacilli in the body.
may be infected, although they do not         individuals with suspected TB are            Recommended treatment regimens
show a positive response to the skin          examined for mycobacteria. However,          include two or more drugs to which the
test. Individuals with tuberculosis           this type of test only permits a             bacilli are susceptible, because the use
infection and no disease would have           presumptive diagnosis of TB since the        of a single drug can lead to the
negative bacteriologic studies and no         test cannot distinguish between              development of bacilli resistant to that
clinical or radiographic evidence of          tuberculosis mycobacteria and other          drug (Ex. 5–85). Treatment with these
tuberculosis disease. However, these          non-tuberculosis mycobacteria. Chest X-      first-line drugs involves a two-phase
individuals are infected for life and are     rays may also be used to diagnose active     process: an initial bactericidal phase for
at risk of developing active TB in the        TB; however, some individuals with TB        the quick elimination of the bulk of
future.                                       may have X-ray findings that are             bacilli from most body sites and a
   Anti-tuberculosis drugs may be used        atypical of those usually associated with    longer-term sterilizing phase for
for individuals with TB infection but         TB (e.g., HIV infected individuals). The     eliminating the remaining bacilli.
                Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules                             54175

Different regimes of drug treatment (i.e.,   transmission of drug-resistant TB from       initiating adequate treatment,
the types of drugs and frequency of          an individual already infected with          individuals with active MDR–TB may
administration) are recommended              drug-resistant strains of the tuberculosis   remain infectious for longer periods of
depending on the medical history of the      bacteria, i.e., transmitted resistance. In   time. Consequently, the likelihood that
patient involved and the results of drug     recent years, drug-resistant forms of TB     they will infect other noninfected
susceptibility testing. The U.S. Public      have emerged that are resistant to two       individuals is increased.
Health Service has recommended               or more of the first-line drugs used to         Once infection occurs, other factors
options for the initial therapy and          treat TB, such as isoniazid and rifampin,    may influence the probability of
dosage schedules for the treatment of        two of the most effective anti-TB drugs.     progressing to the active form of disease.
drug-susceptible TB (Ex. 4B). While          These drug-resistant forms of the            As previously discussed, 10% of
these antimicrobials are effective in the    disease are referred to as multidrug-        immunocompetent adults infected with
treatment of active TB, some of these        resistant TB or MDR–TB. MDR–TB               TB develop active TB. Three to five
drugs also have toxic potential. Adverse     represents a significant form of drug-       percent of untreated immunocompetent
side effects of these drugs include          resistant TB from a public health            adults develop active TB within the first
hepatitis, peripheral neuropathy, optic      standpoint, since its resistance to the      year after infection (Ex. 7–50, pg. 30; 7–
neuritis, ototoxicity and renal toxicity     first-line drugs used for therapy            52). Thus, recently infected individuals
(Ex. 7–93). Thus, patients undergoing        complicates finding adequate therapy         have the highest risk of developing
TB therapy must also be monitored for        regimens that will control the bacilli’s     active TB. This risk is increased for
drug toxicity that may occur from anti-      growth.                                      individuals whose immune system is
tuberculosis drugs.                             Treatment of drug-resistant TB is         impaired (e.g., persons being treated
   Individuals with active disease who       determined on a case-by-case basis,          with immunosuppressive or
are infectious may need to be                using information from the patient’s         glucocorticoid drugs, persons with
hospitalized in order to provide             medical history and drug susceptibility      chronic conditions such as asthma or
isolation so that they will not infect       testing. The recommended course of           emphysema or persons infected with the
other individuals. After the initiation of   treatment will vary depending on the         HIV). The probability of developing
treatment for active TB, improvement of      drugs to which the bacilli are               active disease can also be influenced by
the disease can be measured through          susceptible. Compared to conventional        other conditions that may alter immune
clinical observations such as loss of        TB drug therapy, MDR–TB, in general,         function such as overall decreased
fever, reduction in coughing, increased      requires more complex interventions,         general health status, malnutrition, and
appetite and weight gain. A reduction in     longer hospitalization and more              increasing age.
the number of bacilli in sputum smears       extensive laboratory monitoring. The            The resurgence of TB in the United
also indicates improvement. Three            risk of death from such infections is        States from 1985 to 1992 has been
consecutive negative sputum smears           markedly increased. For example, from        attributed to a number of interacting
generally indicate that the individual is    January 1990 through September 1992,         factors: (1) The inadequate control of
no longer infectious. However,               the CDC investigated eight outbreaks of      disease in high prevalence areas; (2) the
decisions about infectiousness are           MDR–TB. In these outbreaks, 253              increase in poverty, substance abuse,
usually determined on a case-by-case         patients were infected, of whom              poor health status and crowded
basis after taking a number of factors       approximately 75% died (Ex. 3–38–A).         substandard living conditions; and (3)
into consideration, such as the presence     Many of these were                           the growing number of inmates,
of cough, the positivity of sputum           immunocompromised due to infection           residents of homeless shelters, elderly
smears, and the status or response to        with HIV. The interval from the time of      persons in long-term care facilities,
chemotherapy. Although no longer             TB diagnosis to the time of death ranged     persons with HIV infection and
infectious to other individuals, the         from 4 to 16 weeks, with a median time       immigrants from countries with a high
individual undergoing treatment still        of 8 weeks.                                  prevalence of TB infection (Ex. 7–50).
has tuberculosis disease and must                                                         This increase has begun to decline, with
                                             Factors Affecting Transmission               the 1995 case levels approaching the
continue treatment. Discontinuing or
erratically adhering to the treatment           A number of factors can influence the     1985 levels. However, a main reason for
regime can allow some of the bacilli to      likelihood of acquiring a tuberculosis       this decrease is the implementation of
survive such that the individual will be     infection: (1) The probability of coming     TB control measures, like those
at risk of becoming ill and infectious       into contact with an individual with         proposed in this standard, in selected
again (Ex. 7–52, p. 25).                     infectious TB, (2) the closeness of the      areas of the country such as New York
   Not all strains of the tuberculosis       contact, (3) the duration of the contact,    City. OSHA believes that
bacilli are susceptible to all of the        (4) the number of tuberculosis bacilli in    implementation of such measures is
antimicrobials used to treat TB. In some     the air, and (5) the susceptibility of the   necessary to prevent a resurgent peak
instances, drug-resistant forms of M.        uninfected individual. Several               such as that observed from 1985 to 1992
tuberculosis may emerge. Drug                environmental conditions can influence       and to realize the goal set out by the
resistance may emerge by 1 of 3              the likelihood of infection. For example,    National Advisory Committee for the
mechanisms (Exs. 5–85; 7–50, pp. 44–         the volume of shared air space, the          Elimination of Tuberculosis. The
47). Drug-resistant TB may occur             amount of ventilation, the presence or       following discussion describes some of
naturally from random mutation               absence of sunlight, the humidity and        the health effects data related to
processes, i.e., primary resistance. In      the crowded nature of the living             occupational exposure to TB and
addition, drug-resistant TB may result       quarters. These types of factors will        illustrates how the presence of TB
due to inadequate or erratic treatment,      affect the probability of acquiring a        control measures influences TB
i.e., acquired resistance. In these cases,   tuberculosis infection after being           infection and disease.
erratic or inadequate treatment allows       exposed to an individual with infectious
the tuberculosis bacilli to become           TB. MDR–TB is not more contagious            Occupational Exposure
resistant to one or several of the drugs     than drug-susceptible forms of the             Exposure to TB in the health care
being used. Finally, drug-resistant TB       disease. However, due to time delays in      setting has long been considered an
may result due to the active                 diagnosing resistance patterns and           occupational hazard. With the steady
54176            Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules

decline in reported TB cases from 1953         lack of appropriate engineering              rates over a 25 year period, tuberculin
to 1985, some of the concern for               controls), employees exposed to              conversion rates among recent graduates
occupational exposure and transmission         individuals with infectious TB have          exceeded 1% per year and age-specific
also declined. However, from 1985 to           become infected and in some cases have       infection rates among all the physicians
1992 the number of reported cases of TB        developed active disease.                    studied were more than twice that of the
increased. In addition, in recent years,          In 1979, Barrett-Connor (Ex. 5–11)        U.S. population at comparable ages. The
several outbreaks of TB among both             examined the incidence of TB among           authors did not obtain information from
patients and staff in hospital settings        currently practicing physicians who          the physicians on what type of infection
have been reported to the CDC. These           graduated from California medical            control measures were being used in the
outbreaks have been attributed to              schools from approximately 1950 to           facilities where they acquired their
several factors: (1) Delayed recognition       1979. Through mailed questionnaires,         infections.
of active TB cases, (2) delayed drug           physicians were asked to provide                A similar analysis by Geisleler et al.
susceptibility testing, (3) inadequate         information that included their year of      (Ex. 7–46) evaluated the occurrence of
isolation of individuals with active TB        graduation from medical school, BCG          active tuberculosis among physicians
(e.g., lack of negative pressure               vaccination history, history of active TB,   graduating from the University of
ventilation in isolation rooms,                results of their tuberculin skin testing,    Illinois medical school between the
recirculation of unfiltered air, and           and the number of patients they were         years 1938 and 1981. This study, also
allowing infectious patients to freely         exposed to with active TB within the         conducted by questionnaire, reported
move in and out of isolation rooms), and       past year. They were also asked to           that among 4575 physicians questioned,
(4) performance of high-risk procedures        classify themselves as tuberculin            there were 66 cases of active TB, of
on infectious individuals under                positive or negative and to indicate the     which 23% occurred after 1970. Sixty-
uncontrolled conditions (Ex. 7–50). In         year of the last negative and first          six percent of the cases occurred within
addition to hospitals, outbreaks of TB         positive tuberculin test.                    6 years of graduation. In addition, the
have also been reported among the                 Of the 6425 questionnaires mailed         authors reported that in most years the
patients, clients, residents and staff of      out, 4140 responses were received from       incidence of TB was greater among
correctional facilities, drug treatment        currently practicing physicians. Twelve      these physicians than the general
centers, homeless shelters and long-term       percent of the physicians had received       population.
health care facilities for the elderly. The    the BCG vaccine. Sixty-one percent of           Weiss (Ex. 7–45) examined
factors contributing to the outbreaks in       the unimmunized physicians, who also         tuberculosis among student health
these other occupational settings are          had no history of active tuberculosis,       nurses in a Philadelphia hospital. From
very similar to those factors contributing     considered themselves to be tuberculin       1935 to 1939, before the introduction of
to the outbreaks in hospital settings (i.e.,   negative. A total of 1542 (42%) reported     anti-TB drugs and the beginning of the
delayed recognition of TB cases and            themselves as having a positive              general decline of TB in the United
poor/inadequate ventilation for isolation      response to the tuberculin skin test,        States, 100% conversion rates were
areas).                                        with approximately 44 percent of those       observed among those students who
   The following is a discussion of some       tuberculosis infections occurring before     were initially tuberculin negative. For
of the studies that have examined              entering medical school. Of those            example, of 643 students admitted, 43%
occupational transmission of TB. A             infections occurring before entering         were tuberculin negative. At the end of
large proportion of the available              medical school, approximately eight          only 4 months, 48% were tuberculin
information comes from exposures               percent were reported as having been a       positive. At the end of 1 year, 85.9%
occurring in hospitals, in part because        result of contact following work             were tuberculin positive and by the end
this occupational setting has been             experience in the hospital prior to          of the third year 100% were positive. Of
recognized for many years as an area of        entering medical school. For those           those students who converted during
concern with regards to the                    physicians infected either during or         their student nursing tenure,
transmission of TB. However, in more           after medical school, the sources of         approximately 5 percent developed
recent years this concern has spread to        infection were reported as occurring as      active TB disease.
other occupational settings which share        a result of a known patient contact             A decline in the rate of infection was
factors identified in the hospital setting     (45.1%), an unknown contact (41.5%)          observed over the next 36 years among
as contributing to the transmission of         and a non-patient contact (13.4%). In        student nurses at this hospital. The rates
disease. The following sections will           some cases, the nonpatient contact was       of infection were followed for ten
include a discussion of some of the            reported as another physician or another     classes of student nurses from 1962 to
historical data from the hospital setting,     hospital employee. Approximately one         1971. The students had little contact
as well as the more recent data that have      in ten of the physicians infected after      with patients during their first year but
been developed in hospitals and other          entry into medical school developed          spent 4 weeks of their second year of
occupational settings where the                active TB disease.                           training on the tuberculosis wards.
transmission of TB has occurred as a              The authors also examined the             Among those students initially
result of the recent resurgences in the        incidence of infection, measured as the      tuberculin negative, the average
number of active TB cases.                     conversion rates in those remaining          conversion rate was 4.2% over the nine
                                               negative at the end of different time        year period, ranging from 0 to 10.2%. Of
Hospitals—Prior to 1985                        intervals (e.g., the last three years of     the students who converted, 0.6%
   Even prior to the recent resurgence of      medical school and five to 10 years after    developed active TB disease. The
TB in the general population, studies          graduation). This examination indicated      authors attributed the decreases in
have shown an increased risk of                that from 1950 to 1975, there was a 78%      conversion rates to not only the general
transmission of TB to health care              decrease in tuberculin conversion rates      decrease in TB disease in the
workers exposed to individuals with            despite the expanding pool of                community, but also to the increased
infectious TB. These studies clearly           susceptible medical students (i.e., an       efficiency of surveillance of patients
demonstrate that in the absence of             increasing number of medical students        entering the hospital for the early
appropriate TB control measures (e.g.,         who were tuberculin negative). Yet           identification of potential cases of TB
lack of early identification procedures,       despite this overall decrease in infection   and the increased efficiency of isolation
                Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules                                54177

for TB patients. Despite the dramatic       5–15) reported a case study in which 23        factor was that the patient was not
decreases in conversion rates among         employees converted after exposure to a        diagnosed with infectious TB until after
these student nurses, conversion rates      patient with an undetected case of             his death, by which time he had already
were observed at levels as high as 10%      tuberculosis bronchopneumonia. In this         infected 24 employees.
for a given year, indicating that while     study, the source case was an individual          These earlier studies illustrate that
the infection rates had decreased           who was admitted to the emergency              despite the decrease in TB morbidity
substantially since 1939, there still       room with pulmonary edema. Upper               since the advent of anti-tuberculosis
remained a significant amount of            lobe changes of the lung were noted in         drugs in the 1940’s, occupational
occupational transmission of TB in          the chest X-ray, and TB was mentioned          transmission of TB continues to be a
1971. Moreover, this study shows that       as a possible cause. However, no               problem. In addition, while many
short term exposure, i.e., 4 weeks, is      sputum cytology was conducted. The             improvements have been made in
capable of infecting hospital employees.    patient spent 3 hours in the emergency         infection control procedures for TB in
   Similar rates of conversion among        room, 57 hours in a private room and           hospitals, evidence of occupational
hospital employees initially tuberculin     another 67 hours in intensive care until       transmission of TB continues to be
negative were observed in a 1977 study      his death. Treatment of the patient            reported.
by Ruben et al. (Ex. 7–43) which            included intubation with an                    Hospitals—1985 to Present
analyzed the results of a tuberculin skin   endotracheal tube and vigorous
testing program 31 months after its         nasotracheal suctioning. It was only              As discussed above, the transmission
inception at a university hospital in       upon microscopic examination of tissue         of TB has been well established as an
Pittsburgh. Of 626 employees who were       samples of the lung and lymph nodes            occupational hazard in the hospital
tested twice with the tuberculin skin       after the autopsy of the patient that          setting. Many improvements were made
test, 28 (4.5%) converted from negative     tuberculosis mycobacteria were                 in infection control practices. However,
to positive. The employees were             detected.                                      the resurgence in TB from 1985 to 1992
classified as either having a ‘‘presumed       Employees who worked in the                 has brought to attention the fact that
high degree of patient exposure’’ or a      emergency room, the intensive care unit        many TB control measures have not
‘‘presumed low degree of patient            and on the floor of the private room           been implemented or have been
exposure’’. Employees presumed to           (NW 3) and who were also tuberculin            inadequately applied. These studies
have high patient exposure included         negative before the admission of the           demonstrate that TB continues to be an
nurses, X-ray and isotope laboratory        patient, were retested to detect possible      occupational hazard in the hospital
personnel and central escort workers.       conversion. In addition, 21 initially          setting. In addition, similar to the earlier
Employees presumed to have low              tuberculin negative employees on an            studies, the more recent data show that
exposure included secretaries, persons      adjacent floor (NW 2) were also retested.      the lack of early identification
in housekeeping and dietary work, and       Of the 121 employees tested, 24 were           procedures and the lack of appropriate
business office, laundry and central        identified as having converted to              ventilation, performance of high-hazard
supply personnel. The rates of              positive status (21 working on NW 3, 2         procedures under uncontrolled
conversion for employees with               working in the intensive care unit and         conditions and the lack of appropriate
presumed high exposure (6%) and for         1 working on NW 2). No conversions             respiratory protection have resulted in
employees with presumed low exposure        were observed among those working in           the infection of employees and in some
(8%) were not significantly different.      the emergency room.                            cases the development of active disease.
However, this study excluded                   The employees who were retested             The more current outbreaks are even
physicians and medical and nursing          were classified as either having close         more troubling due to the emergence of
students. These groups of employees         contact (e.g., providing direct care),         multidrug-resistant forms of TB disease,
would also presumably have had high         little contact (e.g., more distant contact),   which in some cases have resulted in
exposure to patients since they are often   unknown contact (e.g., no record or            fatality rates approaching 75%.
the hospital staff most directly involved   recollection of contact) or indirect              In a 1985 study, Chan and Tabak (Ex.
in administering patient care. Had these    contact (e.g., in the same room a day or       7–3) investigated the risk of TB
employees been included the number of       two after the patient’s stay). Conversions     infection among physicians in training
conversions among employees with            occurred in 50% (13 of 26) of those            at a Miami hospital. In this study a
presumably high exposure may have           employees with close contact, 18.5% (6         survey was conducted among 665
been significantly increased.               of 33) of those with little contact, 21.4%     physicians in training who were in their
   The study was not designed to            (3 of 14) of those with unknown contact        first four years of postgraduate training.
determine the source of exposure for        and 3.7% (1 of 29) of those with indirect      Only 404 responded to the survey, of
any of the employees who converted.         contact.                                       which 13 were illegible. Another 72
However, the authors suggested that the        While the majority of conversions           were excluded because they had
high level of conversions among those       seems to have occurred in those                received the BCG vaccination. Of the
employees with presumed low exposure        employees on NW 3 who had close or             remaining 319 physicians, 55 were
to patients may have resulted from          little contact, there also were employees      tuberculin positive.
exposures at home. A majority of this       with more distant contact who were                Of the 279 who were tuberculin
group was comprised of housekeeping         infected. An analysis of the ventilation       negative at the beginning of their post
staff who were of low socio-economic        of NW 3 indicated that the central air         graduate training, 15 were excluded
status. The authors also suggested that     conditioning recycled 70% of the air           because they had more than four years
unrecognized cases of tuberculosis may      with no high efficiency filter and no          of training and 43 were excluded
be playing an important role in the         record of balancing the air conditioning       because they had not had repeat skin
occupational transmission of TB in the      system, thus allowing the air from the         tests. Of the 221 remaining available for
hospital.                                   patients’ rooms to mix with and return         evaluation, 15 converted to positive
   Unrecognized cases of TB have been       to the central corridor air. In addition,      tuberculin status, of which two
shown to play a significant role in the     smoke tube tests detected direct air flow      developed active disease.
outbreak of TB in a general hospital. In    from the patients’ rooms to the hall              The overall conversion rate for these
1972, Ehrenkranz and Kicklighter (Ex.       corridor. Perhaps the more important           physicians was 6.79%. In addition, the
54178            Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules

authors observed a positive correlation        of occupational transmission of TB in        risk of TB transmission. A 1990 report
between the rate of conversion and the         TB-prevalent areas for those in close        by Malasky et al. (Ex. 7–41) investigated
duration of postgraduate training. The         patient contact jobs.                        the potential for TB transmission from
conversion rate increased with the                In 1989, Haley et al. (Ex. 5–16)          high-hazard procedures by examining
duration of training, beginning with a         conducted a case study of a TB outbreak      tuberculin skin test conversion rates
cumulative percentage of conversion of         among emergency room personnel at a          among pulmonary physicians in
2.06% in the first year, 8.62% in the          Texas hospital. In this study, a 70 year     training. In this study, questionnaires
2nd year, 11.11% in the third year and         old male diagnosed with pulmonary TB         were sent annually, for 3 years, to
14.29% in the fourth year, resulting in        and undergoing treatment was diverted,       training programs located in the top 25
a linear conversion rate of 3.96% per          due to respiratory arrest, to Parkland       cities for TB in 1983. The purpose of the
year. As noted by the authors, this linear     Memorial Hospital while in route to          study was to compare the conversion
increase suggests the hospital                 another hospital. The man was admitted       rates of pulmonary disease fellows to
environment as the source of the               to the emergency room for                    the conversion rates of infectious
infection. In addition, the prevalence         approximately 4 hours until he was           disease fellows. It was presumed that
rate of conversions in the hospital            stabilized. Afterwards, the patient was      both groups have contact with patients
(17.24%) was much higher than would            placed in an intensive care unit, where      with TB but that pulmonary disease
have been expected in the community            he remained for 2 months until his           fellows are usually more involved with
for individuals of the same age.               death.                                       invasive procedures such as
   The authors suggested that these high          Six cases of active TB developed          bronchoscopies. Information requested
rates of conversion may have been a            among emergency room employees after         on the questionnaires included the type
result of the fact that the hospital in this   exposure to the TB patient, i.e., the        of fellowship (i.e., pulmonary or
study encounters 5 to 10 times more            index case. Five of these were among         infectious disease fellow), prior
active TB cases than most other urban          nurses who recalled contact with the         tuberculin skin test status, tuberculin
hospitals. In addition, the physicians in      index patient and a sixth case was an        status by the Mantoux technique at the
training also are expected to be the first     orderly who may have been infected           end of the 3 year fellowship program,
in line to perform physical evaluations        from one of the employee TB cases. In        history of BCG vaccination, age, sex and
and evaluate body fluids and secretions.       addition, a physician exposed while          ethnicity. In addition, the pulmonary
While the authors did not go into detail       administering treatment in the intensive     disease fellows were asked to give
about what, if any, TB infection control       care unit also developed active disease.     information on the number of
precautions were taken by these                   Skin test conversions were evaluated      bronchoscopies they performed and
physicians in training, they did note          for the 153 employees of the emergency       their use of masks during the procedure.
that the evaluation of body fluids and         room. Of 112 previously negative                Fourteen programs submitted data
secretions was often done in poorly            employees, 16 had positive skin tests,       that were usable. Only programs that
ventilated and ill-equipped laboratories.      including 5 nurses diagnosed with            had both pulmonary and infectious
   Increased rates of conversion were          active TB. Fifteen of the conversions        disease fellows in the same system were
observed among employees in a New              were a result of exposure to the index       used for the study. From this
Orleans hospital in a 1986 study by            case. Skin tests were also evaluated for     information, it was observed that 7 of 62
Ktsanes et al. (Ex. 7–6). Similar to           physicians in the intensive care unit. Of    (11%) of the pulmonary fellows at risk
Miami, New Orleans also has a high rate        21 resident physicians, two of whom          converted their tuberculin skin test from
of TB in the community. This study             had intubated the index patient, five        negative to positive during the two year
examined the skin test conversions             had newly positive reactions to the          training period. In contrast, only 1 of 42
among a cohort of 550 new employees            tuberculin skin tests. One of the            (2.4%) of the infectious disease fellows
who were followed for five years after         remaining three residents later              converted. The expected conversion rate
assignment to the adult inpatient              developed active disease.                    from previous surveys was 2.3%. In
services. Of these 550 employees who              The authors attributed the outbreak to    addition, the pulmonary disease fellows
were initially tuberculin negative, 17         several factors. First, the index case had   were grouped according to tuberculin
converted to positive status over the          a severe case of pulmonary TB in which       skin status. Skin test status was
five-year study period, resulting in an        he produced copious amounts of               evaluated for its relationship to the
overall five-year cumulative conversion        sputum. Second, sixty percent of the         number of bronchoscopies performed
probability of 5.2%.                           emergency room air was recirculated          and the pattern of mask usage. No
   Regression analyses were done to            without filtration adequate to remove        correlations were found with these
examine potential contributing factors.        TB bacilli, allowing for the recirculation   factors and tuberculin skin status at the
Factors examined in the regression             of contaminated air. Finally, employees      end of the fellowship. The authors
model included race, job, age at               in the emergency room were provided          suggested that the lack of correlation
employment, and department. Only race          surgical masks that were ineffective for     between mask usage during
(i.e., black vs. white employees) and job      protecting against transmission of           bronchoscopies and skin test conversion
(i.e., nursing vs. other jobs) were found      airborne TB droplet nuclei. This study       implies that masks worn by physicians
to be associated with skin test                illustrates that the lack of effective       may be inadequate. While little
conversion. To further examine the             measures for controlling TB                  information was presented to evaluate
potential job effect, conversions among        transmission can result in the infection     this suggestion, the study does suggest
blacks in nursing and blacks in other          and development of active disease in a       that high-hazard procedures such as
jobs were compared. Overall, the               relatively high number of employees          bronchoscopies that induce coughing,
cumulative probability of converting           even after exposure to only one case of      performed under uncontrolled
was higher among blacks in nursing,            active TB.                                   conditions, present a risk for TB
suggesting that the acquired infections           Similarly, the lack of effective          transmission.
resulted from employment at the                controls while performing high-hazard,          Pearson et al. (1992) conducted a
hospital rather than from the                  cough-inducing procedures on                 case-control study to investigate the
community at large. The authors thus           individuals with infectious TB has also      factors associated with the development
concluded that there is an increased risk      been shown to result in an increased         of MDR–TB among patients at a New
                Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules                             54179

York City hospital (Ex. 5–24). As a part      had negative pressure. Based on this         open, left their rooms, and/or removed
of this study, tuberculin skin test           evidence, the authors concluded that         their masks while outside their rooms.
conversion rates were compared among          the observed cases of MDR-TB were a          Patients with TB who were readmitted
health care workers assigned to wards         likely result of infections acquired in      to the HIV ward and who were receiving
where patients with TB were frequently        the hospital (i.e., primary TB) rather       anti-TB drugs were not admitted to
admitted (e.g., HIV unit, general medical     than as a result of the reactivation of      isolation. In some cases, these patients
ward, respiratory therapy) or rarely          infections acquired in the past. The         were later found to have infectious
admitted (operating room, orthopedic          authors attributed these nosocomial          MDR–TB.
ward, outpatient clinic, psychiatry           infections to the lack of adherence to          An environmental assessment of the
ward). In addition, infection control         recommended infection control                ventilation revealed that among 23
procedures and ventilation systems            procedures.                                  rooms tested with smoke tubes, 6 had
were evaluated.                                  While the primary focus of this study     positive pressure and many of the rooms
   Of 79 health care workers who were         was to investigate the transmission of       under negative pressure varied from
previously negative, 12 (15%) had             TB among patients, the increased             negative to positive depending on the
newly positive skin tests. Those health       likelihood of nosocomial infections          fan setting and whether the bathroom
care workers who were assigned to             among patients in the hospital would         door was open. Aerosolized
wards where patients with TB were             seem equally likely to apply to health       pentamidine administration rooms were
frequently admitted were more likely to       care workers working in the same             also found to have positive pressure
have skin test conversions (i.e., 11 of 32)   environment. A survey of tuberculin          relative to adjacent treatment areas. In
than health care workers assigned to          skin test conversions revealed an 18%        addition, the sputum induction rooms
wards where patients with TB were             conversion rate for health care workers      were found to recirculate air back to the
rarely admitted (i.e., 1 of 47).              who previously had negative skin tests       HIV clinic.
   Evaluations of the infection control       and were present during this outbreak of        Skin test conversions were evaluated
procedures and ventilation systems            MDR-TB. Although no statistics were          for all health care workers (i.e., nurses
revealed that patients who were               reported, the authors stated that the        and clerical staff) who tested negative
receiving isolation precautions for           pattern of skin test conversions             on the tuberculin skin test before the
suspected or confirmed TB were                suggested an ongoing risk over time          outbreak period, March 1988 through
allowed to go to common areas if they         rather than a recent increase during the     April 1990. Health care workers on the
wore a surgical mask. However, many of        outbreak period.                             HIV ward and in the HIV clinic
the patients did not keep their masks on         Based on an earlier 1990 report from      exhibited a significantly higher rate of
when out of their rooms. In addition,         the CDC (Ex. 5–22), Beck-Sague et al.        skin test conversion than health care
neither the isolation rooms nor rooms         1992 (Ex. 5–21) conducted a case-            workers on the thoracic surgery ward
used for cough-inducing procedures            control study to investigate an outbreak     (e.g., 13/39 vs. 0/15). Ten of the
were under negative pressure, thus            of MDR–TB among the staff and patients       conversions occurred among the 28
allowing contaminated air to exhaust to       in a HIV ward and clinic of a Miami          health care workers in the HIV ward.
the adjacent corridors.                       hospital. As part of the overall study the   Among these health care workers, the
   Edlin et al. (1992) (Ex. 5–9)              authors compared the skin test               authors reported a significant
investigated an outbreak of MDR–TB in         conversion rates of health care workers      correlation between the risk of infection
a New York hospital among patients            in the HIV ward and clinic to the skin       in health care workers and the number
with acquired immunodeficiency                test conversion rates of health care         of days that patients with infectious
syndrome (AIDS). This study compared          workers in the thoracic surgery ward         MDR–TB were hospitalized on the HIV
the exposure period of AIDS patients          where TB patients were rarely seen. In       ward. No correlation was observed
diagnosed with MDR–TB to the                  addition, the authors also evaluated the     between the risk of infection among
exposure period of AIDS patients with         relationship between the presence of         health care workers on the HIV ward
drug-susceptible TB. The date of              patients with infectious MDR–TB and          and the number of days that patients
diagnosis was defined as the date the         patients with infectious drug-               with infectious drug-susceptible TB
sputum sample was collected from              susceptible TB on the HIV ward and the       were hospitalized on the ward.
which tuberculosis bacteria were grown        risk of skin test conversion among the          Based on skin test conversions and
in culture. Patients were assumed to be       HIV ward health care workers. Infection      the evaluation of infection control
infectious two weeks before and two           control procedures in the HIV ward and       practices in the HIV ward and clinic, the
weeks after the date of diagnosis. The        clinic were also examined.                   authors concluded that the health care
period of exposure was the period in             All patients with suspected or            workers most likely were infected by
which the patient may have been               confirmed TB were placed in isolation.       patients on the HIV ward with MDR–
infected with TB. Because of the rapid        However, some patients whose                 TB. The factors most likely contributing
progression from infection to disease,        complaints were not primarily                to this increased risk of infection
the exposure period was defined as 6          pulmonary and whose chest X-rays were        included: (1) The prolonged
months preceding the date of diagnosis,       not highly suggestive of TB were not         infectiousness and greater number of
excluding the last two weeks.                 initially suspected of TB and were not       days that patients with infectious MDR–
   The patients with MDR–TB were              placed in isolation. Patients who were       TB were hospitalized, (2) the delayed
found to be more likely to have been          admitted to isolation rooms were             recognition of TB and failure to suspect
hospitalized during their exposure            allowed to leave TB isolation 7 days         infectious TB in patients receiving what
periods. Those who were hospitalized          after the initiation of chemotherapy         proved to be ineffective anti-TB
were more likely to have been on the          regardless of clinical or bacteriologic      treatment, (3) the inadequate duration
same ward and on the same day as a            response. Thus, in some instances,           of, and lapses in, isolation precautions
patient with infectious TB and were           patients with MDR–TB were allowed to         on the HIV ward, and (4) the lack of
more likely to have been near a room          leave isolation while they were still        negative pressure ventilation in
housing an infectious patient.                infectious, before drug resistance was       isolation and treatment rooms. While
Examination of the infectious patients’       recognized. In addition, patients in         the evidence in this study primarily
rooms revealed that only 1 of 16 rooms        isolation rooms sometimes left the doors     points to the transmission of MDR–TB
54180           Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules

from patients to health care workers,        frequent chest therapy and suctioning.        Lung X-rays were found to be irregular
many of the problems identified with         Three sputum samples were taken from          but were attributed to the lung cancer.
infection control procedures and             the patient for smears and cultures. All      Upon his death the autopsy revealed
ventilation would also increase the risk     AFB smears were negative. However,            extensive necrosis in the lung but
of acquiring drug-susceptible TB.            two cultures were positive for                tuberculosis was not suspected. Thus,
   In addition to MDR–TB outbreak            tuberculosis.                                 no cultures for mycobacteria were
investigations in Miami, in 1993 the            Despite the presence of positive           performed and no infection control
CDC reported an outbreak in New York         cultures the patient was not diagnosed        procedures were initiated. It was only
City in which health care workers            with active TB. The problem was not           upon histological examination of tissue
became infected after being exposed to       recognized until a physician on staff         samples one month later that the
patients with MDR–TB (Ex. 6–18). In          later developed symptoms of malaise           presence of TB was confirmed. Five
this investigation, for the period           and slight cough and requested a              months later one of the staff performing
December 1990 through March 1992, 32         tuberculin skin test and was found to be      the autopsy developed active TB. His
patients were identified with MDR–TB.        positive. Because the physician had           only history of exposure was to the
Twenty-eight of these patients had           been tuberculin negative 8 months             index case.
documented exposure to an                    earlier, a contact investigation was             As a result, a contact investigation
undiagnosed infectious MDR–TB                initiated. As a part of this investigation,   was initiated for hospital personnel who
patient while all of them were in the        all employees who previously had              had shared air with the patient during
HIV ward of the hospital.                    negative tuberculin tests and who also        his stay, including the autopsy staff. Of
   During November 1991, health care         worked in the intermediate and                susceptible hospital staff (i.e., those not
workers who were assigned to the HIV         intensive care units where the patient        previously found to react positive to the
inpatient unit and who were also             had been treated were given repeat skin       tuberculin skin test), infection
previously negative on the tuberculin        tests. Of 45 employees who previously         developed in 9 of 56 (16%) exposed
skin test, were given an additional skin     had negative tuberculin skin tests, 14        employees (4 autopsy staff, 4 nursing
test. Of 21 health care workers tested, 12   (31%) converted to positive status (6         staff and 1 radiology staff). Only 3 of
(57%) had converted to positive status       physicians, 3 nurses, 2 respiratory           333 unexposed personnel were found to
(7 nurses, 4 aides and 1 clerical worker).   therapists and 1 clerk). Ten of these         have converted to positive tuberculin
None of the health care workers had          conversions were among the 13                 status at the hospital during the same
used respiratory protection.                 previously tuberculin negative staff          period of investigation, thus indicating
   An investigation of infection control     members who were present at the time          a 17.8 fold increase in the infection rate
practices revealed that of 32 patients       bronchoscopies were conducted (10/            for the exposed group.
with MDR–TB, 16 were not initially           13=76.9%). Four of the conversions               Undiagnosed cases of TB at time of
suspected of TB and in these cases           were among 32 susceptible staff               autopsy were also indicated as the likely
isolation precautions either were not        members who were not present at the           cause for development of active TB
used or were instituted late during the      bronchoscopies (4/32=12.5%). The              among staff and students in an autopsy
patients’ hospitalization. In addition,      author thus concluded that being              room in a Swedish hospital (Ex. 5–19).
patients who were admitted to isolation      present during the bronchoscopy of the        In this study, three medical students
frequently left their rooms and when in      patient was a major risk factor in            and one autopsy technician, who were
their room the doors were frequently left    acquiring the TB infection. However,          present during the autopsy of a patient
open. Moreover, all rooms were found         the evidence did not show a significant       with previously undiagnosed
to be under positive pressure relative to    correlation between skin test conversion      pulmonary TB, developed active TB.
the hall. Thus, similar to the findings in   and the type of exposure, i.e., close         Both the medical students and the
Miami, the results of this study indicate    (administered direct contact) versus          autopsy technician had previously
that the inability to properly isolate       casual (in the room) contact. Thus,           received the BCG vaccine but none had
individuals with MDR–TB and also the         people who were present in the room           any other known contact with a
use of inadequate respiratory protection     during the bronchoscopy had an equal          tuberculosis subject. Thus, it was
may increase the risk of infection among     risk of infection as those administering      concluded that the tuberculosis
health care workers.                         direct patient care, presumably, as the       infections were most likely to have been
   Undiagnosed cases may also present a      author suggests, because droplet nuclei       transmitted during the autopsy. The
significant source for occupational          can disperse rapidly throughout the air       findings of this study further illustrate
transmission of TB. A case study by          of a room.                                    the risks that undiagnosed cases of
Cantanzaro (Ex. 5–14) described an              Similarly, Kantor et al. (Ex. 5–18)        active TB present to health care
outbreak of TB infection among hospital      reported an outbreak of TB infection          workers. The lack of recognition of an
staff at a San Diego hospital where the      among hospital staff exposed to a single      active case of TB often results in a
hospital staff were exposed to a single      undiagnosed case of TB. The index case        failure to initiate appropriate infection
patient with undiagnosed TB. In this         in this investigation was a 50 year old       control procedures and provide
case, a 64 year old man suffering from       man who was admitted for lung cancer          appropriate personal protective
generalized seizures was transferred         and was receiving chemotherapy,               equipment. In addition, this study
from a local jail to the emergency room      steroids and radiation treatment. After a     illustrates that, while TB is most often
and later admitted to a four bed             month of treatment, the patient               transmitted by individuals with
intermediate care unit. While in the         complained of a cough and chest pain          infectious pulmonary TB who generate
intermediate care unit he was treated        and was found to have emphysema               droplet nuclei when they cough or
with anticonvulsants but continued to        requiring additional drug treatment and       speak, the autopsy procedures on
have seizures accompanied with               a chest tube. However, even after the         deceased individuals with pulmonary
vomiting. He was therefore placed in         emphysema resolved, the patient               TB may also aerosolize bacteria in the
intensive care where he underwent a          complained of weakness, loss of               lungs and generate droplet nuclei.
variety of procedures including              appetite and fever. A sputum culture             Exposure during autopsy procedures
bronchoscopies and endotracheal              and smear were conducted for                  was also suspected as a possible route
intubation. During his stay, he received     mycobacteria and found to be negative.        of TB transmission in an upstate New
                Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules                           54181

York Medical Examiner’s Office (Ex. 7–      required drainage and irrigation. Due to     worn any personal protective
152). This Medical Examiner’s Office        the suspicion of TB, specimens for AFB       equipment.
conducted autopsies on deceased             smear and culture were obtained and             Thus, similar to other outbreak
inmates from upstate New York prisons.      the patient was placed in isolation.         investigations, the lack of appropriate
In 1991, the same year that an outbreak     While in isolation, drainage from the        ventilation and respiratory protection
of MDR–TB occurred among inmates            abscess continued and irrigation of the      stand out as the key factors in the
from an upstate New York prison, the        abscess cavity was initiated on an 8-        transmission of TB to employees who
Medical Examiner’s office conducted         hour schedule. After four days, acid fast    are exposed to individuals with
autopsies on 8 inmates with TB, six of      bacilli were observed in the AFB smears      infectious TB. Moreover, this particular
whom had infectious MDR–TB at death         and TB therapy was begun. The patient        case study demonstrates that certain
and who were also HIV positive and had      remained in isolation until his death        forms of extrapulmonary TB in
disseminated TB disease.                    except for three days that he spent in       conjunction with aerosolizing
   Skin tests were administered to          the Intensive Care Unit (ICU) due to         procedures, e.g., the irrigation of a
employees who had worked for at least       high fever.                                  tuberculous abscess, have the potential
one month during 1991 at the Medical           An investigation of skin test surveys     for presenting significant airborne
Examiner’s Office. Among 15 employees       among the hospital employees revealed        exposures to M. tuberculosis.
who had originally tested negative on a     55 skin test conversions among 442              Other aerosolizing procedures have
baseline skin test, 2 were found to have    previously nonreactive employees and 5       also shown evidence of presenting
converted. These two employees              conversions among 50 medical students.       airborne exposures to M. tuberculosis.
worked as morgue assistants and had         In addition, 5 of the employees who had      For example, tissue processing was
recent documented exposure to persons       conversions also had active TB,              associated with the skin conversion of
with extensive disseminated MDR–TB.                                                      two pathologists working at a
                                            including one who developed a
No potential exposure to TB outside the                                                  community hospital in California (Ex.
                                            tuberculous finger lesion at the site of a
Medical Examiner’s Office could be                                                       6–27). In this case study, after autopsy,
                                            needle-stick injury incurred during the
found.                                                                                   a 62 year old man who had died from
                                            incision and drainage of the patient’s
   The autopsy area of the office had a                                                  bronchogenic carcinoma was discovered
                                            abscess. All the skin test converters,
separate ventilation system. However,                                                    to have a caseating lung lesion. A stain
                                            except for two, recalled exposure to the
air was returned to a common air                                                         revealed a heavy concentration of acid-
                                            source case. Of the 442 susceptible
plenum, allowing the air to mix between                                                  fast bacilli, which were identified in
                                            employees, 108 worked at least one day
the autopsy area and other areas of the                                                  culture as M. tuberculosis. As a result,
                                            on one of the floors where the patient
office. In addition, the autopsy room                                                    a contact investigation was initiated.
                                            stayed (i.e., the surgical ward, the            This investigation found twenty
was found to be at positive pressure
relative to the adjacent hallway.           medical floor of the patient’s room and      employees who had contact with the
Employees performing or assisting at        the ICU). Four (80%) of 5 surgical suite     patient, including two pathologists and
autopsies on persons known to be            employees who had direct contact with        a laboratory assistant. All were given a
infected with HIV were required to wear     the patient through their assistance with    tuberculin skin test and found to be
plastic gowns, latex gloves and surgical    the incision and irrigation of the           negative. However, after follow-up skin
masks. Particulate respirators were not     patient’s abscess had skin test              testing three months later, the two
required until November of 1991, after      conversions. In addition, 28 (85%) of 33     pathologists had converted. Other than
the installation of germicidal UV lamps.    employees on the general medical floor       contact with the source case, the two
However, this was after the last MDR–       and 6 (30%) of 20 ICU employees had          had no other obvious sources of
TB autopsy. This study suggests that the    skin test conversions. All those             infection. One of the pathologists had
conversion of these two morgue              employees converting recalled exposure       been present at the autopsy. Both
assistants occurred as a result of          to the patient, some of whom had no          pathologists were present when the
exposure to aerosolized M. tuberculosis     direct contact with the patient.             frozen lung sections were prepared.
resulting from autopsy procedures,             Environmental studies revealed that       During this process, the lung tissue was
either as a result of participation in an   two of the areas in which the patient        sprayed with a compressed gas coolant,
autopsy in the autopsy area or from         stayed during his hospitalization did        which created a heavy aerosol. Masks
exposure to air contaminated with           not have negative pressure. The              were not routinely worn during this
aerosolized M. tuberculosis that was        isolation room was under positive            tissue processing. The investigators
exhausted into other areas of the           pressure relative to adjacent rooms and      suspected that this aerosol promoted the
Medical Examiner’s Office.                  the corridor. In addition, the patient’s     transmission of TB and was the likely
   In addition to autopsy procedures,       cubicle in the ICU had neutral pressure      cause of the observed infections.
other procedures, such as the irrigation    relative to the rest of the ICU.                While much of the health effects
of abscesses at sites of extrapulmonary     Employees in these two areas had skin        literature has focused on outbreaks of
TB, can result in the generation of         test conversions even in cases where         TB or MDR–TB, a more recent study
droplet nuclei. An outbreak                 there was no direct patient contact. The     investigated the status of infection
investigation in an Arkansas hospital       lack of negative pressure was thought to     control programs among ‘‘non-outbreak’’
(Ex. 5–17) reported the transmission of     have significantly contributed to the        hospitals (Ex. 7–147). Investigators from
TB among hospital employees exposed         dispersion of droplet nuclei generated       the Society of Health care Epidemiology
to a patient with a tuberculous abscess     from the irrigation of the tuberculous       of America (SHEA) and the CDC
of the hip and thigh. In this study, the    abscess. In the surgical ward, air was       surveyed members of SHEA to assess
source case was a 67 year old man who       directly exhausted to the outside.           compliance in the respondents’
was admitted to the hospital with a         However, all employees present in the        hospitals with the 1990 CDC Guidelines
fever of unknown origin and progressive     surgical ward when the patient was           for Preventing the Transmission of TB
hip pain. The patient did not present       being treated had direct contact with the    in Health Care Facilities for the years
any signs of pulmonary TB; however,         patient. There was no indication that        1989 to 1992. The survey included
the examination of soft tissue swelling     the surgical staff had taken any special     questions on tuberculin skin testing
in the hip area revealed an abscess that    infection control precautions or had         programs (e.g., frequency of testing,
54182           Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules

positivity at hire, and percent newly         (HEPA) filter respirators was by            efficacy of these particulate respirators
converted), AFB isolation capabilities        bronchoscopists and respiratory             in reducing conversion rates from the
(e.g., negative pressure, air changes per     therapists at 4 hospitals.                  reported survey data.
hour, HEPA filtration) and respiratory          As a second phase of this                    For hospitals with fewer than 6 TB
protection.                                   investigation, the survey responses were    patients or with fewer than 437 beds, no
   The survey showed that of the 210          analyzed to determine the efficacy of the   differences in conversion rates were
hospitals represented by the SHEA             TB infection control programs among         reported among health care workers
members’ survey results, 193 (98%)            the member hospitals participating in       from hospitals that had implemented
admitted TB patients from 1989 to 1992,       the survey (Ex. 7–148). In this analysis,   AFB isolation capabilities such as
40% of which had one or more patients         the reported conversion rates were          single-room occupancy, negative
with MDR–TB. In addition, the                 compared to reported infection control      pressure, or directly exhausted air and
proportion of hospitals caring for drug       measures (i.e., AFB isolation               those hospitals that had not. The
susceptible TB patients rose from 88%         capabilities and respiratory protection).   investigators suggested that this finding
to 92% and the proportion of hospitals        For purposes of comparison, hospitals       may support contentions that the
caring for MDR–TB patients rose from          were categorized as having either less      efficacy of TB infection control
5% to 30%. While the number of                than or ≥6 TB patients, less than or ≥437   measures vary depending on
hospitals caring for TB patients              beds, and admitting or not admitting        characteristics of the hospital or
increased, the majority of those              MDR–TB patients.                            community exposure. However, given
hospitals cared for a small number of           Conversion rates were higher among        the small sample size of the survey, as
patients. In 1992, approximately 89% of       health care workers from hospitals with     well as the reduced potential for
the hospitals reported 0 to 25 patients       ≥437 beds than among health care            exposure in hospitals with fewer than 6
per year, while approximately 5%              workers from smaller hospitals (0.9%        TB patients per year, it would be
reported greater than 100 patients per        vs. 0.6%, p≤0.05). This difference was      difficult to detect any differences in
year.                                         more pronounced among ‘‘higher-risk’’       conversion rates among health care
   Few hospitals reported routine             health care workers (i.e., health care      workers from hospitals with or without
tuberculin skin testing for each of the       workers including bronchoscopists and       certain levels of infection control.
years surveyed. For example, while 109        respiratory therapists). ‘‘Higher-risk’’    Where more opportunity does exist for
(52%) of the responding hospitals             health care workers from hospitals with     exposure (e.g., hospitals with ≥6 TB
reported tuberculin skin test results for     437 or more beds had a 1.9% conversion      patients), this analysis does show that
at least one of the years from 1989 to        rate compared to a conversion rate of       the implementation of TB infection
1992, only 63 (30%) reported results for      0.2% for ‘‘higher-risk’’ health care        control procedures can reduce the
each of these years. When examining           workers from smaller hospitals.             transmission of TB among health care
the conversion rates over time from           Similarly, health care workers from         workers.
1989 to 1992, the investigators limited       hospitals where 6 or more TB patients
their analysis to the 63 hospitals            were admitted per year had higher           Hospitals—Summary
reporting skin test data for each of these    conversion rates than health care              In summary, the evidence clearly
4 years. Among these hospitals the            workers from hospitals with fewer than      shows that in hospital settings,
median percentage of employees newly          6 TB patients per year (e.g., 1.2% vs.      employees are at risk of occupational
converting to positive skin test status       0.6%).                                      exposure to TB. Various studies and TB
remained constant over the 4 year               For hospitals with 6 or more TB           outbreak investigations have shown that
period at approximately 0.34% per year        patients, conversion rates also varied      employees exposed to individuals with
(i.e., 3/1000 per year). However, when        depending on the level of TB infection      infectious TB have converted to positive
including all hospitals in the analysis,      control practices that were in place in     tuberculin skin status and in some cases
from 1989 to 1992, the number of              the hospital. For example, among            have developed active disease. In these
hospitals reporting conversion rates          hospitals with 6 or more TB patients        reports, a primary factor in the
increased from 63 to 109 and the              and whose AFB isolation capabilities        transmission of TB has been a failure to
conversion rates increased from 0.26%         included at least single-room               promptly identify individuals with
(i.e., 2/1000) to 0.50% (i.e., 5/1000).       occupancy, negative pressure and            infectious TB so that appropriate
   With regard to AFB isolation               directly exhausted air, the conversion      infection control measures could be
capabilities, 62% of 181 responding           rates among health care workers were        initiated to prevent employee exposure.
hospitals reported that they had              lower than the conversion rates among       In addition, another major factor
isolation facilities consistent with the      health care workers at hospitals with 6     identified as contributing to
1990 CDC TB Guidelines (i.e., single-         or more TB patients but which did not       occupational exposures was the lack or
patient room, negative pressure, air          have similar isolation capabilities         ineffective implementation of
directly exhausted outside, and ≥6 air        (0.62% vs. 1.83%, p=0.03). For              appropriate exposure control methods
changes per hour). Sixty-eight percent of     respiratory protection, however, no         (e.g., lack of negative pressure in
the reporting hospitals had isolation         differences in conversion rates were        isolation rooms, lack of appropriate
facilities meeting the first three of these   observed among health care workers          respiratory protection for exposed
recommendations. For respiratory              wearing surgical masks (0.94%) and          employees, performance of high-hazard
protection, the majority of health care       health care workers using submicron         procedures under uncontrolled
workers in the hospitals used surgical        surgical masks, dust-mist respirators or    conditions). The lack of early
masks. However, there was an increase         dust-mist-fume respirators (0.98%).         identification and appropriate control
in the use of dust-mist or dust-mist-         Very few survey respondents reported        measures resulted in the exposure and
fume respirators. The use of dust-mist        use of HEPA filter respirators. For         subsequent infection of various hospital
respirators increased from 1 to 13%           example, only four hospitals reported       employees. These employees included
from 1989 to 1992 and the use of dust-        use of any HEPA respirators, and these      not only health care providers
mist-fume respirators increased from 0        were not the predominant type of            administering direct patient care to
to 10% for the same period. The only          respiratory protection used (Ex. 7–147).    individuals with infectious TB, but also
use of high efficiency particulate air        Thus, it is not possible to evaluate the    hospital staff providing support services
                Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules                              54183

to the infectious individuals, hospital       such as those listed above. Not all the      compared to an overall TB case rate of
staff working in adjacent areas of the        settings listed by the CDC as places         9.5 per 100,000 for the U.S. (Ex. 6–26).
hospital using shared air, autopsy staff      where TB transmission may be likely to       In 1995, TB cases reported among the
and laboratory staff working with             occur have been adequately studied and       foreign born accounted for 35.7% of the
infected culture and tissue samples.          thus can not be included in this             total reported cases, marking a 63.3%
                                              discussion. However, the discussion of       increase since 1986 (Ex. 6–34). Three,
Other Occupational Settings
                                              the following sectors clearly                many correctional facilities have a high
   While hospitals have been historically     demonstrates that the occupational           proportion of individuals who are
recognized as the primary type of work        transmission of TB is not limited to the     infected with HIV. The CDC reported
setting where TB presents an                  hospital setting. Occupational settings      that in addition to the growing increase
occupational hazard, there are other          where there is an increased likelihood       in AIDS among prisoners, the incidence
work settings where the transmission of       of exposure to aerosolized M.                of AIDS in prisons is markedly higher
TB presents a hazard to workers. There        tuberculosis present the same types of       than that for the U.S. general
are a variety of occupational settings in     occupational hazards as have been            population. In 1988, the incidence of
which workers can reasonably be               documented in the hospital setting.          AIDS cases in the U.S. population was
anticipated to encounter individuals
                                              Correctional Facilities                      13.7 per 100,000 compared to an
with active TB as a part of their job
                                                                                           estimated aggregate incidence for state/
duties. Several work settings have been          Many correctional facilities have a
                                                                                           federal correctional systems of 75 cases
identified by the CDC where exposure to       higher incidence of TB cases than occur
                                                                                           per 100,000 (Ex. 3–33). Individuals who
TB presents an occupational hazard:           in the general population. For example,
correctional facilities, long-term care       the CDC reported that the incidence of       are infected with HIV or who have AIDS
facilities for the elderly, homeless          TB among inmates of correctional             are at an increased risk of developing
shelters, drug treatment centers,             facilities was more than three times         active TB due to their decreased
emergency medical services, home-             higher than that for nonincarcerated         immune capacity. The likelihood of
health care, and hospices. Similar to the     adults aged 15–64, based on a survey of      pulmonary TB in individuals with HIV
hospital setting, these work settings         TB cases in 1984 and 1985 by 29 state        infection is reflected in the CDC’s
have a higher number of individuals           health departments (Ex. 3–33). In            Revised Classification System for HIV
with active TB than would be expected         particular, among inmates in the New         infection (Ex. 6–30). In this revised
for the general population. Many of the       York correctional system, the TB             classification system, the AIDS
clients of these work settings have many      incidence increased from an annual           surveillance case definition was
characteristics (e.g., high prevalence of     average of 15.4 per 100,000 during 1976      expanded to include pulmonary TB.
TB infection, high prevalence of HIV          to 1978 to 105.5 per 100,000 in 1986         Moreover, X-rays of individuals infected
infection, intravenous drug use) that         (Ex. 7–80) to 156.2/100,000 for 1990–        with HIV who have TB often exhibit
place them at an increased risk of            1991 (Ex. 7–137). Similarly, in 1987, the    radiographic irregularities that make the
developing active TB. These types of          incidence of TB among inmates in New         diagnosis of active TB difficult (Exs. 7–
work settings are also similar to             Jersey was 109.9 per 100,000                 76, 7–77, 7–78, and 7–79). HIV-infected
hospitals in that workers at these sites      (approximately 11 times higher than the      individuals may have concurrent
may also provide medical services and         general population in New Jersey) and        pulmonary infections that confound the
perform similar types of high-hazard          in California the incidence of TB among      radiographic diagnosis of pulmonary
procedures that are typically done in a       inmates was 80.3 per 100,000                 TB. In addition, it may be difficult to
hospital setting.                             (approximately 6 times higher than that      distinguish symptoms of TB from
   In addition to employees who provide       for the general population for California)   Pneumocystis carinii pneumonia or
medical services in these other types of      (Ex. 3–33). In 1989, the CDC reported        other opportunistic infections. This
work settings, there are other types of       that since 1985, eleven known outbreaks      difficulty in TB diagnosis can result in
workers (e.g., guards, admissions staff,      of TB have been recognized in prisons        true cases of active TB going
legal counsel for prisoners) who may          (Ex. 3–33).                                  undiagnosed in this population.
also be exposed to individuals with              The increased incidence of TB in          Undiagnosed TB has been shown to be
infectious TB. Similar to hospitals, these    correctional facilities has been             an important cause of death in some
work settings have an over-                   attributed to several factors (Ex. 7–25).    patients with HIV infection (Ex. 7–76).
representation of populations at high         One, correctional facilities have a higher   Fourth, environmental conditions in
risk for developing active TB, e.g.,          incidence of individuals who are at          correctional facilities can aid in the
individuals infected with HIV,                greater risk for developing active TB.       transmission of TB. For example, many
intravenous drug users, elderly               For example, the population in prisons       prisons are old, have inadequate
individuals, and individuals with poor        and jails may be dominated by persons        ventilation systems, and are
nutritional status and who are medically      from poor and minority groups, many of       overcrowded. In addition, inmates are
underserved. In addition to having a          whom may be intravenous drug users.          frequently transferred both within and
higher percentage of individuals with         These particular groups may also suffer      between facilities, thus increasing the
TB infection and a higher percentage of       from poor nutritional status and poor        potential for the spread of TB infection
individuals at an increased risk for          health care, factors that place them at      among inmates and staff. This increased
developing active TB, many of these           increased risk of developing active          potential for mobility among inmates
work settings also share environmental        disease. Two, special types of               also enhances the likelihood that
factors that facilitate the transmission of   correctional facilities, such as holding     inmates undergoing therapy for active
TB, such as overcrowding and                  facilities associated with the               disease will either discontinue their
inadequate ventilation, which increases       Immigration and Naturalization               treatment or inadequately follow their
the occupational hazard. The following        Services, may have inmates/detainees         prescribed regime of treatment. The
discussion describes some of the studies      from countries with a high incidence of      inadequacy of their treatment may give
available in the literature that have         TB. For foreign-born persons arriving in     rise not only to relapses to an infectious
examined the occupational transmission        the U.S., the case rate of TB in 1989 was    state of active disease, but also
of TB in other occupational settings          estimated to be 124 per 100,000,             potentially give rise to strains of MDR–
54184           Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules

TB. These strains of TB have a higher         exhausted to the hospital corridors and      A similar investigation in a California
incidence of fatal outcome and are            other patient rooms.                       state correctional institution identified
generally characterized by prolonged             A contact investigation in the prison   three active cases of TB (two inmates
periods of infectiousness during which        was conducted to identify other inmates    and one employee) during September
the risk of infection to others is            who might have been exposed during         and October 1991 (Ex. 6–5). As a result,
increased.                                    this outbreak of MDR–TB. Of those          an investigation was commenced to
   The high incidence of TB among the         inmates with previous negative             determine whether transmission of TB
inmate population presents an                 tuberculin skin tests and without active   was ongoing in the institution. Eighteen
occupational hazard to the staff in these     disease (306), ninety-two (30%) had        inmates with active TB were identified.
types of facilities. Recent outbreak          documented skin test conversions.          TB in 10 of these inmates was
investigations by the CDC have                There was no tuberculin skin test          recognized for the first time while they
documented the transmission of TB to          program for prison staff; therefore,       were in the institution during 1991,
exposed workers. In an investigation of       conversions among prison employees         resulting in an annual incidence of TB
a state correctional facility in New York     could not be evaluated.                    of 184 per 100,000, a rate greater than
for 1991 (Exs. 6–3 and 7–136), eleven            The primary factors identified as       10 times that for the state (17.4 per
persons with TB were identified (10           contributing to this outbreak were         100,000). Two of the 10 inmates had
inmates and one correctional facility         deficiencies in identifying TB among       negative tuberculin skin tests prior to
guard). Nine persons (8 inmates and the       transferred inmates, laboratory delays,    their entry into the institution. Three of
guard) had MDR–TB. All eight inmates          and lapses in isolating inmates with       the cases were determined to have been
were HIV positive. The guard was HIV          active TB within the facility. Inmates     infectious during 1991.
negative; however, he was also                with symptoms of active disease were         A review of skin test data revealed
immunocompromised as a result of              not sent for evaluation in some cases      that for the 2944 inmates for whom skin
treatment for laryngeal cancer. Seven of      until they became so ill they could not    test results were available, 324 tested
the inmates and the guard died from           care for themselves. Some of these         positive for the first time while in the
                                              inmates were placed in the infirmary       prison system. Of these, 106 were
MDR–TB. The eighth inmate was still
                                              with other inmates until their diagnosis   tuberculin negative before their entry
alive and receiving treatment for MDR–
                                              with TB. On other occasions, drug          into the prison system, 96 of which
TB 2 years after being diagnosed as
                                              susceptibility testing was not reported    occurred in the previous two years.
having the disease. DNA analysis
                                              until after an inmate’s death, which         The employee identified as having
identified the strains of tuberculosis
                                              means that appropriate patient             active TB had worked as a counselor on
bacteria from these individuals to be
                                              management was not initiated.              the prison’s HIV ward, where he
identical.                                       As a result of this outbreak, a         recalled exposure to one of the 3
   The investigation revealed that the        retrospective epidemiological              infectious inmates. This employee could
source case was an inmate who had             investigation was conducted to examine     recall no known exposures outside the
been transferred from another prison          the potential extent and spread of MDR–    prison. Similarly, two other prison
where he had been previously exposed          TB throughout the New York State           employees had documented skin test
to MDR–TB. He arrived at the prison           prison system during the years 1990–       conversions while working at the
with infectious TB but refused                1991 (Ex. 7–137). This investigation       prison. Neither recalled exposures
evaluation by the infirmary staff. This       revealed that 69 cases of TB were          outside the prison; one reported
inmate was placed in the general prison       diagnosed in 1990 and another 102 were     exposure to an inmate with possible TB.
population where he stayed for 6              diagnosed in 1991, resulting in a            No information was provided in this
months until he was admitted to the           combined incidence of 156.2 cases/         report as to whether any isolation
hospital where he later died. However,        100,000 inmate years for 1990 and 1991     precautions were implemented at this
before his hospitalization, he exposed        combined. Of the cases, 39 were            facility. However, the investigators
two inmates living in his cell block who      identified as being MDR–TB, 31 of          concluded that their findings suggested
later developed MDR–TB. These two             which were shown to be                     the likelihood that transmission of TB
inmates continued to work and live in         epidemiologically linked. Thirty-three     had occurred in the prison. Their
the prison until shortly before their final   of the individuals with MDR–TB never       conclusion was based on the fact that a
hospitalization. The other inmates who        received any treatment for MDR–TB, 3       substantial number of skin test
subsequently developed MDR–TB had             were diagnosed at death, and 23 died       conversions were documented among
several potential routes of exposure:         before drug susceptibility results were    the inmates and that at least two
social contact in the prison yard, contact    known. These inmates were also             inmates with active TB became infected
at work sites in the prison, and contact      discovered to be highly mobile. The 39     while at the prison.
at the prison infirmary where they            inmates lived in 23 different prisons        The transmission of TB was also
shared rooms with other inmates before        while they were potentially infectious.    reported in another California prison
diagnosis with TB.                            Twenty transfers were documented for       among prison infirmary physicians and
   The guard who developed MDR–TB             12 inmates with potentially infectious     nurses and correctional officers (Ex. 6–
had exposure to inmates while                 MDR–TB (9 shortly before diagnosis,        6). In this investigation, an inmate with
transporting them to and from the             one after diagnosis with TB but before     active MDR–TB spent 6 months during
hospital. The primary exposure for this       diagnosis with MDR–TB, and 2 after a       1990–1991 in the infirmary. The
guard apparently occurred when he was         diagnosis of MDR–TB).                      infirmary had no isolation rooms and
detailed outside the inmates’ room               Several factors were identified as      inmates’ cells were found to be under
during their hospitalization for MDR–         contributing to the spread of MDR–TB       positive pressure. Employees
TB. The inmates were hospitalized in an       throughout the New York prison system:     occasionally recalled wearing surgical
isolation room with negative pressure.        delays in identifying and isolating        masks when entering the rooms of TB
However, upon investigation it was            inmates, frequent transfers without        patients.
discovered that the ventilation system        appropriate medical evaluation, lapses       An analysis of available skin testing
for the room had not been working             in treatment, and delays in diagnosis      data revealed that of the 21 infirmary
correctly and had allowed air to be           and susceptibility testing.                health care providers, only 10 had been
                 Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules                             54185

tested twice during the period from            among the shelter residents and among        Cincinnati, Columbus, and Toledo (Ex.
1987 to 1990. Of these 10, two were            the shelter staff. A high prevalence of      7–51). In Cincinnati, 11 individuals
newly positive, one of whom had                TB infection and disease is common           with active TB were identified in a
recently converted in 1991 and had             among many homeless shelters. This is        shelter for homeless adults. The index
spent 5 months in the preceding year           not surprising, since the residents of       case was a man who had resided at the
providing health care to the source case       these facilities usually come from lower     shelter and later died from respiratory
in this investigation. Another health          socio-economic groups and often have         failure. He was not diagnosed with TB
care provider and a correctional officer       characteristics that place them at high      until his autopsy. Of these 11
who worked in the infirmary also were          risk. Screening of selected clinics and      individuals, of which the index case
identified as having newly converted           shelters for the homeless has shown that     was one, 7 were determined to be
while at the prison. There was no yearly       the prevalence of TB infection ranges        infectious. There was no indication as to
skin test screening, and thus their            from 18 to 51% and the prevalence of         whether any infection control measures
conversions could have occurred at any         clinically active disease ranges from 1.6    were in place in the shelter. DNA
time between 1987 and 1991. However,           to 6.8% (Ex. 6–15). The CDC estimates        analysis of 10 individual M. tuberculosis
13 other inmates were diagnosed with           this to be 150 to 300 times the              isolates showed identical patterns. The
pulmonary TB during that same period.          nationwide prevalence rate (Ex. 6–17).       similarity among these DNA patterns
An additional correctional officer who            In addition to having a high              suggested that transmission of the TB
did not work in the infirmary also was         prevalence of individuals with TB            occurred in the shelter.
found to have newly converted. His             infection in the shelters, many of the          While the primary focus of this
reported exposure occurred at a                shelter residents possess characteristics    investigation was on the active cases
community hospital where he was                that impair their immunity and thus          reported among the residents in this
assigned to an inmate with infectious          place them at a greater risk of              Cincinnati shelter, the risk of
TB. The officer was not provided with          developing active disease. For example,      transmission identified in this shelter
any respiratory protection. The lack of        homeless persons generally suffer from       also would apply to the shelter staff.
isolation precautions and the lack of          poor nutrition, poor overall health          Possible transmission of TB infection
appropriate respiratory protection             status and poor access to health care.       from the infectious individuals to the
suggest transmission of TB from                Many also suffer from alcoholism, drug       shelter staff might have been identified
infectious inmates in the infirmary to         abuse and psychological stress.              through tuberculin skin test
the prison staff, either as a result of        Moreover, a significant portion of           conversions. However, no tuberculin
exposure to the source case or other           homeless shelter residents are infected      skin test information for the staff was
inmates with pulmonary TB who were             with the HIV. In 1988, the Partnership       reported in this investigation.
also treated in the prison infirmary.          of the Homeless Inc. conducted a survey         Tuberculin skin testing results were
Because of the lack of contact tracing or      of 45 of the nation’s largest cities and     reported in the investigation of a
routine annual screening of inmates or         estimated that there were between 5,000      Columbus, Ohio shelter. In this
staff, the full extent of transmission         and 8,000 homeless persons with AIDS         investigation, a resident of a Columbus
from the source case or other TB cases         in New York City and approximately           homeless shelter was identified with
could not be determined.                       20,000 nationwide (Ex. 7–55). Due to         infectious pulmonary TB at the local
   Thus, similar to the evidence for the       these factors, homeless shelter residents    hospital in March of 1990. The patient
hospital setting, the evidence on              are at increased risk of developing          also had resided in a shelter in Toledo.
correctional facilities shows that the         active disease. Thus, there is the           As a result, a city-wide TB screening
failure to promptly identify individuals       increased likelihood that these              was initiated from April to May 1990
with infectious TB and provide                 individuals will be infectious as a result   among the residents and staff of the
appropriate infection control measures         of active disease and thereby present a      city’s men’s shelters. Tuberculin skin
can result in the exposure and                 source of exposure for other homeless        tests were conducted on 363 shelter
subsequent infection of employees with         persons and for shelter employees.           residents and 123 shelter employees.
TB. These employees include the                   In addition to having factors which       Among 81 skin-tested residents of the
correctional facility infirmary staff,         increase their risk of developing active     shelter in which the index case had
guards on duty at the facility, and            TB disease, homeless persons also are a      resided, 32 (40%) were positive
guards assigned to escort inmates during       very transient population. Because they      compared to 47 (22%) of 210 skin-tested
transport to other facilities (e.g., outside   are transient, homeless persons are more     residents of other shelters in Columbus
health care facilities and other               likely to discontinue or to erratically      who had positive skin test reactions.
correctional facilities).                      adhere to the prescribed TB therapy.         Similarly, among 27 employees of the
                                               Inadequately adhering to TB therapy          shelter where the index case resided, 7
Homeless Shelters                              can result in relapses to an infectious      (26%) had positive skin test reactions
  Tuberculosis has also been recognized        state of the disease or the development      compared to 9 (11%) of 85 employees in
as a health hazard among homeless              of MDR–TB. Both outcomes result in           other men’s shelters. These skin test
persons. The growth of the homeless            periods of infectiousness, during which      results suggest an increased risk of
population in the United States since          they present a source of exposure to         transmission of TB among residents and
the 1980s and the subsequent increase          other residents and staff. In addition,      employees of the homeless shelter
in the number of shelters for the              environmental factors at homeless            where the index case resided. However,
homeless, furthers heightens the               shelters, such as crowded living             due to the lack of baseline skin test
concern about the potential for the            conditions and poor ventilation,             information among these residents and
increased incidence and transmission of        facilitate the transmission of TB.           employees it is not possible to
TB among the homeless, especially in              Outbreaks of TB among homeless            determine when their conversion to
crowded living conditions such as              shelter residents have been reported.        positive status occurred and whether
homeless shelters.                             For example, during 1990, 17                 this index case was their source of
  A number of factors are present in           individuals with active pulmonary TB         exposure. These results, however, do
homeless shelters which increase the           were identified among residents of           indicate a high prevalence of TB
potential for the transmission of TB           homeless shelters in three Ohio cities:      infection among homeless residents
54186           Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules

(e.g., 40% and 22%). Many of these            TB diagnosed around the time of the            population for the transmission of TB.
individuals are likely to have an             outbreak were of the same phage type,          TB disease in persons over the age of 65
increased risk of developing active TB        suggesting that there was a predominant        constitutes a large proportion of TB in
and, as a result, they may present a          chain of infection, i.e., a single source      the United States. Many of these
source of exposure to residents and           of infection. However, there also were         individuals were infected in the past,
staff.                                        other phage types, suggesting several          before the introduction of anti-TB drugs
   The transmission of TB has also been       sources of infection. Therefore, the           and TB control programs when the
observed among residents and staff of         investigators suggested that there was         prevalence of TB disease was much
several Boston homeless shelters (Exs.        probably a mixture of primary and              greater among the general population,
7–75 and 6–25). From February 1984            reactivated cases.                             and have harbored latent infection over
through March 1985, 26 cases of TB               In addition to the similarity of phage      their lifetimes. However, with
were confirmed among homeless                 types among TB cases, tuberculin skin          advancing age, these individuals’
residents of three large shelters in          testing results suggested the ongoing          immune function starts to decline,
Boston. Nineteen of the 26 cases              transmission of TB in the shelter. For         placing them at increased risk of
occurred in 1984, thus giving an              example, 10 shelter clients who were           developing active TB disease. In
incidence of approximately 317 per            previously tuberculin negative in May          addition, they may have underlying
100,000, 6 times the homeless case rate       1985 were re-tested in January 1987 and        disease or overall poor health status.
of 50 per 100,000 reported for 1983 and       3 (30%) had converted. In addition, 43         Moreover, residents are often clustered
nearly 16 times the 1984 case rate of 19      clients who were negative in January           together and group activities are often
per 100,000 for the rest of Boston (Ex.       1987 were re-tested in June 1987 or            encouraged. TB case rates are higher for
6–25).                                        February 1988 and 10 (23%) had                 this age group than for any other. For
   Of the 26 cases of TB reported, 15 had     converted. Factors identified as               example, the CDC reports that in 1987,
MDR–TB. Phage typing of isolates from         contributing to the outbreak were the          the 6,150 cases of TB disease reported
13 of the individuals with drug-resistant     increased number of men with                   for persons ≥65 years of age accounted
TB showed identical phage types, thus         undiagnosed infectious pulmonary TB,           for 27% of the U.S. TB morbidity
suggesting a common source of                 the close proximity of beds in the             although this group only represented
exposure. As a result of this outbreak, a     shelter, and a closed ventilation system       12% of the U.S. population (Ex. 6–14).
screening program was implemented in          that provided extensive recirculation of          Because of the higher prevalence of
November 1984 over a four-night               unfiltered air.                                TB cases among this age group,
period. Of 362 people who received skin          As a result of the outbreak, a control      employees of facilities that provide
tests, 187 returned for reading, 42 (22%)     plan was implemented. This plan                long-term care for the elderly are at
were found to be positive and 3 were          included repetitive mass screening,            increased risk for the transmission of
recent converters. Screening also was         repetitive skin testing, directly observed     TB. More elderly persons live in nursing
reported for the shelter staff at the three   therapy, preventive therapy and                homes than in any other type of
homeless facilities. At the largest of the    modification of the ventilation system to      residential institution. The CDC’s
three shelters, 17 of 85 (20%) staff          incorporate UV light disinfection in the       National Center for Health Statistics
members had skin test conversions. In         ventilation duct work. After the control       reports that elderly persons represent
the other two shelters, 3 of 15 (20%) and     plan was in place, five additional             88% of the nation’s approximately 1.7
3 of 18 (16%) staff members had skin          individuals with active TB were                million nursing home residents. As
test conversions.                             observed over a 2-year follow-up period.       noted by the CDC, the concentration of
   Whereas MDR–TB was primarily                  While the primary focus in this study       such high-risk individuals in long-term
involved in the outbreak in Boston, an        was on clients of the shelter rather than      care facilities creates a high-risk
outbreak of drug-susceptible TB was           the shelter staff, the risk factors present    situation for the transmission of TB (Ex.
reported in a homeless shelter in             in the shelter before implementation of        6–14).
Seattle, Washington (Ex. 7–73). From          the control plan would have also                  In addition to having a higher
December 1986 to January 1987, seven          increased the likelihood for                   prevalence of active TB, the recognition
cases of TB from homeless residents           transmission of TB to shelter employees        of TB in elderly individuals may be
were reported to the Seattle Public           from infectious clients.                       difficult or delayed because of the
Health Department. The report of 7               Thus, similar to correctional facilities,   atypical radiographic appearance that
individuals with active TB in one             homeless shelters have a number of risk        TB may have in elderly persons (Exs. 7–
month prompted an investigation,              factors that facilitate and promote the        59, 7–81, 7–82, and 7–83). In this
including: (1) A mass screening to            transmission of TB (e.g., high incidence       situation, individuals with active TB
detect undiagnosed cases, (2) phage           of infected residents with an increased        may go undiagnosed, providing a source
typing of isolates from shelter clients to    likelihood of developing active disease,       of exposure to residents and staff.
detect epidemiologically linked cases,        crowded living conditions and poor                While the increased incidence of TB
and (3) a case-control study to               ventilation). Also, similar to                 cases among the elderly in long-term
investigate possible risk factors for the     correctional facilities, the evidence in       care facilities may be a result of the
acquisition of TB.                            homeless shelters shows that the failure       activation of latent TB infections, the
   A review of the case registries            to promptly identify homeless residents        transmission of TB infection to residents
revealed that 9 individuals with active       with infectious TB and the lack of             and staff from infectious cases in the
TB had been reported from the homeless        appropriate TB control measures (e.g.,         facilities has been observed and
shelter for the preceding year and four       lack of isolation precautions or prompt        reported in the scientific literature.
cases in the year previous to that. As a      transfer to facilities with adequate              For example, Stead et al. (1985)
result of the mass screening in late          isolation precautions) resulted in the         examined the reactivity to the
January 1987, an additional 6                 transmission of TB to shelter employees.       tuberculin skin test among nursing
individuals with active TB were                                                              home residents in Arkansas (Ex. 7–59).
detected. Phage typing of 15 isolates         Long-Term Care Facilities for the Elderly      This study involved a cross-sectional
from the shelter-associated cases               Long-term care facilities for the            survey in which tuberculin skin tests
revealed that 6 individuals with active       elderly also represent a high-risk             were given to all current nursing home
                Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules                              54187

residents. In addition, all newly-           negative, 38 (12%) converted to positive      his stay in the home, was not diagnosed
admitted nursing home residents were         status. Included among these were             with TB until after he was hospitalized
skin tested. For the three year period       employees in nursing (18), medical (3),       because of fever, loss of weight and
evaluated, 25,637 residents of the 223       dental (1), maintenance/engineering (3),      productive cough. The remaining 37%
nursing homes in Arkansas were tested.       supply (1), dietary (9), and clerical (2)     converted in the absence of a known
   Of 12,196 residents who were tested       services.                                     infectious case. Thus, the authors
within one month of entry, only 12              Occupational transmission of TB was        suggested that nosocomial infections are
percent were tuberculin positive,            also reported in a nursing home in            likely to result from persons
including those for whom a booster           Oklahoma (Ex. 6–28). In August 1978, a        unsuspected of having TB.
effect was detected. However, among          68 year old female residing in the east          Skin testing was also reviewed for
the 13,441 residents for whom the first      wing of the home was diagnosed with           employees of the nursing home.
test was delayed for more than a month,      pulmonary TB. She was subsequently            Questionnaires were completed by 108
20.8% were positive. In addition, the        hospitalized. However, by that time she       full-time employees. Eleven of 68
results of retesting 9,937 persons who       had already had frequent contact with         employees with follow-up skin tests
were tuberculin negative showed an           other residents in the east wing. As a        converted to positive skin status during
annual conversion rate of approximately      result, a contact investigation, in which     the study period. Ten of the 11 (91%)
5% in nursing homes in which an              all residents of the home were given          converters reported that they had been
infectious TB case had been recognized       skin tests, was initiated.                    in the nursing home in 1975, the same
in the last three years. In nursing homes       The investigation revealed that the        year in which many of the residents
with no recognized cases, the authors        reaction rate for residents in the east       were also found to have converted from
reported an annual conversion rate of        wing (34/48, 71%) was significantly           a single infectious case. In addition,
approximately 3.5%. The authors              higher than the reaction rates of             employees working at least 10 years in
concluded that their data supported the      residents living in the north and front
                                                                                           the home had a higher percentage of
contention that tuberculosis may be a        wings (30/87, 34%). No baseline skin
                                                                                           conversions (9 of 22, 40%) than
rather common nosocomial infection in        test information was presented for the
                                                                                           employees working less than 10 years (2
nursing homes and that new infections        residents to determine the level of
                                                                                           of 46, 4.4%). Based on the results of this
with tuberculosis is an important risk       conversion. However, it was noted that
                                                                                           study, the authors concluded that, in
for nursing home residents and staff.        half of the nursing home residents were
                                                                                           addition to occurrence of TB cases from
   Brennen et al. (Ex. 5–12) described an    former residents of a state institution for
                                                                                           the reactivation of latent infections
outbreak of TB that occurred in a            the developmentally disabled. A 1970
                                                                                           among the elderly, TB can also be
chronic care Veteran’s Administration        tuberculin skin test survey of that
                                                                                           transmitted from one resident to another
Medical Center in Pittsburgh. This           institution had shown a low rate of
investigation was initiated as a result of   positive reactions.                           resident or staff. Consequently, TB must
two skin test conversions identified            In addition to the nursing home            be considered as a potential nosocomial
through the employee testing program.        residents, nursing home employees             infection in nursing homes.
One converter was a nurse working on         were also skin tested. Of the 91                 Thus, long-term care facilities for the
ward 1B (a locked ward for                   employees tested, 61 (67%) were               elderly represent a high-risk situation
neuropsychiatric patients) and the other     negative and 30 (33%) were positive.          for the transmission of TB. These types
was a physician working in an adjacent       Similar to results observed among the         of facilities possess a number of
ward, 1U, who also had significant           residents, positive reaction rates were       characteristics that increase the
exposure to ward 1B. The source of           higher for employees who had ever             likelihood that active disease may be
infection in this investigation was          worked in the east wing (50%) than for        present among the facility residents and
traced to two patients who had resided       those who had never worked in the east        may go undetected. Similar to other
on ward 1B and who had either a              wing (23%). Retesting of the employees        high-risk settings, the evidence shows
delayed or undiagnosed case of TB. The       3 months later revealed 3 conversions.        that the primary factors in the
contact investigation revealed 8             These results suggested that there may        transmission of TB among residents and
additional conversions among patients,       have been occupational transmission of        staff have been the failure to promptly
4 in ward 1B and 4 in wards 2B and 4B        TB in this facility.                          identify residents with infectious TB
(units on the floor above 1B).                  Occupational transmission has also         and initiate and adequately implement
   Because the source cases were             been observed in a retrospective study        appropriate exposure control measures.
initially unidentified, no isolation         of residents and employees who lived or
                                                                                           Drug Treatment Centers
precautions were taken. Smoke tracer         worked in an Arkansas nursing home
studies revealed that air discharged         between 1972 and 1981 (Ex. 7-83). In             Another occupational setting that has
from the window air conditioning unit        this retrospective study, investigators       been identified as a high-risk
of one of the source patients discharged     reviewed the skin testing and medical         environment for the transmission of TB
directly into the courtyard. Air from this   chart data collected over a 10-year           is drug treatment centers. Similar to
courtyard was the air intake source for      period at an Arkansas nursing home.           other high-risk sites, drug treatment
window air conditioning units in the         Among the nursing home residents who          centers have a higher prevalence of TB
converters’ room on ward 2B and thus         were admitted between 1972 and 1982,          infection than the general population.
was one of the possible sources of           32 of 226 residents (17%) who were            For example, in 1989 the CDC funded
exposure.                                    initially tuberculin negative upon            25 state and city health departments to
   In addition to the contact                admittance became infected while in the       support tuberculin testing and
investigation on ward 1B and the             home, based on conversion to positive         administration of preventive therapy in
adjacent units, hospital-wide skin           after at least two previous negative tests.   conjunction with HIV counseling and
testing results were evaluated. Of 395       Twenty-four (63%) of these conversions        testing. In this project, 28,586 clients
employees tested, 110 (28%) were             were infected in 1975, following              from 114 drug treatment centers were
positive. The prevalence in the              exposure to one infectious resident.          given tuberculin skin tests. Of those,
surrounding community was estimated          This resident, who had negative skin          2,645 (9.7%) were positive (Ex. 6–8).
to be 8.8%. Of those employees initially     tests on three previous occasions during      When persons with previously
54188           Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules

documented positive tests were               converted to positive tuberculin status.      Therefore, individuals with TB would
included, 4167 (13.3%) were positive.        While the prevalence and incidence            go unidentified. In addition, the clients
   There is also evidence to suggest that    rates of TB infection were similar for the    were housed in a building with crowded
drug dependence is a risk factor for TB      two groups of patients, seropositive          dormitories for sleeping. The only
disease. For example, Reichman et al.        patients showed a higher incidence of         ventilation in this building was
(Ex. 7–85) evaluated the prevalence of       developing active disease. Active TB          provided by opening windows and
TB disease among different drug-             developed in 8 of the seropositive            doors. Thus, environmental conditions
dependent populations in New York: (1)       subjects with TB infection (4%),              were ideal for the transmission of TB.
An in-hospital population, (2) a             whereas none of the seronegative                 Consequently, the high-risk
population in a local drug treatment         patients with TB infection developed          characteristics of clients who frequent
center, and (3) a city-wide population in    active TB during the study period.            these centers (e.g., high prevalence of
methadone clinics. For the in-hospital          Among individuals who are infected         infection and factors increasing the
population of 1,283 patients discharged      with HIV or who have AIDS, TB disease         likelihood of developing active disease)
with drug dependence, 48 (3.74%) had         may be difficult to diagnosis because of      and environmental characteristics of the
active disease, for a prevalence rate of     the atypical radiographic appearance          center (e.g., crowding and poor
3,740 per 100,000. In comparison, the        that TB may present in these                  ventilation), lead to drug treatment
TB prevalence rate for the total inpatient   individuals. In these individuals, TB         centers being considered a high-risk
population was 584 per 100,000 and for       may go undiagnosed and present an             setting for the transmission of TB. The
New York City as a whole was 86.7 per        unsuspected source of exposure. Clients       available evidence shows that the
100,000. Screening of clients at a local     of drug treatment centers also may be         failure to promptly identify clients with
drug treatment center in Harlem              more likely to discontinue or                 infectious TB and to initiate and
revealed a TB prevalence of 3750 per         inadequately adhere to TB therapy             properly implement exposure control
100,000 in the drug-dependent                regimens in instances where they              methods (e.g., proper ventilation)
population. Similarly, in the New York       develop active disease. As in other           resulted in the infection of clients and
methadone program, the city-wide TB          instances, this increases the likelihood      staff at these facilities.
prevalence was 1,372 per 100,000. The        of relapse to active disease or possibly
authors also reported that although          the development of MDR–TB, both of
estimates of TB infection rates for both     which result in additional or even              The available evidence clearly
drug-dependent and non-drug                  prolonged periods of infectiousness           demonstrates that the transmission of
dependent people were similar, the           during which other clients or staff can       TB represents an occupational hazard in
prevalence of TB disease among the           be exposed.                                   work settings where employees can
drug-dependent was higher, thus                 There is evidence showing the              reasonably be anticipated to have
suggesting that drug dependency may be       transmission of TB in drug treatment          contact with individuals with infectious
a risk factor for disease.                   facilities among both the clients and the     TB or air that may reasonably be
   Clients of drug treatment centers not     staff. In a CDC case study (Ex. 5–6), a       anticipated to contain aerosolized M.
only have a high prevalence of TB            Michigan man who was living in a              tuberculosis as a part of their job duties.
infection, a majority of them are            residential substance abuse treatment         Epidemiological studies, case reports,
intravenous drug users. Of the estimated     facility and was undergoing therapy for       and outbreak investigations have shown
645,000 clients discharged each year         a previously diagnosed case of TB             that in various work settings where
from drug treatment centers,                 disease, was discovered by the local          there has been an increased likelihood
approximately 265,000 are intravenous        health department to have MDR–TB. As          of encountering individuals with active
drug users who either have or are at risk    a result, a contact investigation was         TB or where high-hazard procedures are
for HIV infection. In the Northeastern       initiated at the drug treatment facility in   performed, employees have become
U.S., HIV seroprevalence rates of up to      which he resided.                             infected with TB and in some cases
49% have been reported (Ex. 6–8).               Of the 160 clients and staff who were      developed active disease. While some
These individuals are at increased risk      identified as potential contacts, 146         infections were a result of more direct
of developing active TB disease.             were tested and given tuberculin skin         and more prolonged exposures, other
   To determine the risk of active TB        tests in November. No health screening        infections resulted from non-direct and
associated with HIV infection, Selwyn        program had been in place at the              brief or intermittent exposures. Because
et al. (Ex. 5–6) prospectively studied       facility. The following March repeat          of the variability in the infectiousness of
520 intravenous drug users enrolled in       skin tests were given. Of the 70 persons      individuals with active TB, one
a methadone maintenance program. In          who were initially tuberculin negative        exposure may be sufficient to initiate
this study, 217 HIV seropositive and 303     and were still present in the facility, 15    infection.
seronegative intravenous drug users,         (21%) had converted to positive status          Several factors, common to many of
who had complete medical records             (14 clients and 1 staff member). The          these work settings, were identified as
documenting their history of TB and          investigators noted that the number of        contributing to the transmission of TB:
skin test status, were followed from June    converters may have been                      (1) Failure or delayed recognition of
1985 to January 1988. On admission to        underestimated for two reasons. Many          individuals with active TB within the
the methadone program, and at yearly         of the clients were at risk for HIV           facility, and (2) failure to initiate or
intervals, all patients were given           infection and thus may have been              adequately implement appropriate
tuberculin skin tests.                       anergic and not responded to the              infection control measures (e.g.,
   Forty-nine (23%) of the seropositive      tuberculin skin tests. In addition, nearly    performance of high-hazard procedures
patients and 62 (20%) of the                 half of the clients who were initially        under uncontrolled conditions, lack of
seronegative patients had positive           negative were not available for repeat        negative pressure ventilation,
reactions to the skin test before entry      skin testing.                                 recirculation of unfiltered air, and lack
into the study. Among the patients who          Several factors may have contributed       of appropriate respiratory protection).
initially had negative skin tests, 15 of     to the observed conversions in this           Thus, in work settings where employees
131 (11%) seropositive patients and 62       facility. For example, no health              can reasonably be anticipated to have
of 303 (13%) seronegative patients           screening program was in place.               contact with individuals with infectious
                  Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules                                   54189

TB or air that may contain aerosolized          eliminated or lessened by a change in        infections (Ex. 7–269). For some
M. tuberculosis and where appropriate           practices’’ (448 U.S. 642).                  percentage of active cases, a more severe
infection control programs are not in              The Court in the Cotton Dust case         clinical course can develop which can
place, employees are at increased risk of       (American Textile Manufacturers              be attributed to various factors such as
becoming infected with TB.                      Institute v. Donovan, 452 U.S. 490           the presence of MDR–TB, an allergic
   Infection with TB is a material              (1981)), rejected the use of cost-benefit    response to treatment, or the synergistic
impairment of the worker’s health. Even         analysis in setting OSHA health              effects of other health conditions an
though not all infections progress to           standards. However, the Court                individual might have. Further, OSHA
active disease, infection marks a               reaffirmed its previous position in the      estimates that for 7.78% of active TB
significant change in an individual’s           ‘‘benzene’’ case that a risk assessment is   cases, TB is expected to be the cause of
health status. Once infected, the               not only appropriate, but also required      death. Section 6(b)(5) of the OSH Act
individual is infected for his or her           to identify significant health risk in       states that,
entire life and carries a lifetime risk of      workers and to determine if a proposed         The Secretary, in promulgating standards
developing active disease, a risk they          standard will achieve a reduction in that    dealing with toxic materials or harmful
would not have had they not been                risk. Although the Court did not require     physical agents under this subsection, shall
infected. In addition, many individuals         OSHA to perform a quantitative risk          set the standard which most adequately
with infection undergo preventive               assessment in every case, the Court          assures, to the extent feasible, on the basis of
therapy to stop the progression of                                                           the best available evidence, that no employee
                                                implied, and OSHA as a matter of policy
infection to active disease. Preventive                                                      will suffer material impairment of health or
                                                agrees, that assessments should be put       functional capacity even if such employee
therapy consists of very toxic drugs that       into quantitative terms to the extent        has regular exposure to the hazard dealt with
can cause serious adverse health effects        possible. The following paragraphs           by such standard for the period of his
and, in some cases, may be fatal.               present an overall description of            working life.
   Although treatable, active disease is        OSHA’s preliminary quantitative risk
also a serious adverse health effect.                                                           For this rulemaking, OSHA defines
                                                assessment for occupational exposure to      TB infection as a ‘‘material impairment
Some TB cases, even though cured, may           tuberculosis (TB).
result in long-term damage to the organ                                                      of health’’, for several reasons. First,
                                                   An earlier version of this risk           once infected with TB, an individual
that is infected. Individuals with active       assessment was reviewed by a group of
disease may need to be hospitalized                                                          has a 10% lifetime likelihood of
                                                four experts in the fields of TB             developing active disease and
while they are infectious and they must         epidemiology and mathematical
take toxic drugs to stop the progressive                                                     approximately 1% likelihood of
                                                modeling. The reviewers were George          developing more serious complications
destruction of the infected tissue. These       Comstock, MD, MPH, DPH, Alumni
drugs, as noted above, are toxic and may                                                     leading to death. Second, allergic
                                                Centennial Professor of Epidemiology,        reaction and hepatic toxicity due to
have serious side effects. Moreover,            The Johns Hopkins University; Neil
even with advancements in treating TB,                                                       chemoprophylaxis with isoniazid,
                                                Graham MBBS, MD, MPH, Associate              which is one of the drugs used in the
individuals still die from TB disease.          Professor of Epidemiology, The Johns
This problem is compounded by the                                                            recommended course of preventive
                                                Hopkins University; Bahjat Qaqish, MD,       treatment, pose a serious threat to a
emergence of multidrug-resistant strains        PhD, Assistant Professor of Biostatistics,
of TB. In these cases, due to the inability                                                  large number of workers. Third,
                                                University of North Carolina; and            defining infection with M. tuberculosis
to find adequate drug regimens which            Patricia M. Simone, MD, Chief, Program
can treat the disease, individuals remain                                                    as material impairment of health is
                                                Services Branch, Division of                 consistent with OSHA’s position in the
infectious longer, allowing the disease         Tuberculosis Elimination, CDC. The
to progress further and cause more                                                           Bloodborne Pathogens standard and is
                                                reader is referred to the peer review        supported by CDC and several
progressive destruction of the infected         report in the docket for additional
tissue. This increases the likelihood of                                                     stakeholders who participated in the
                                                details (Ex. 7–911). The revised version     pre-proposal meetings, as well as Dr.
long-term damage and death.                     of OSHA’s risk assessment, as published      Neil Graham, one of the peer reviewers
V. Preliminary Risk Assessment for              in this proposed rule, includes OSHA’s       of this risk assessment. In his comments
Occupational Exposure to Tuberculosis           response to the reviewers’ comments as       to OSHA, Dr. Graham stated,
                                                well as updated risk estimates based on
Introduction                                    recent purified protein derivative (PPD)       The focus of OSHA on risk of TB infection
                                                                                             rather than TB disease is appropriate. TB
  The United States Supreme Court, in           skin testing data made available to the
                                                                                             infection is a potentially adverse event,
the ‘‘benzene’’ decision (Industrial            Agency since the peer review was             particularly if exposure is from a MDR–TB
Union Department, AFL–CIO v.                    performed and is generally supported by      patient, or if the health-care or institutional
American Petroleum Institute, 448 U.S.          the reviewers or is consistent with          worker is HIV seropositive. In addition, a
607 (1980)), has stated the OSH Act             reviewers’ comments. (Note: PPD skin         skin test conversion will in most cases
requires that, prior to the issuance of a       test and tuberculin skin test (TST) are      mandate use of chemoprophylaxis for >6
new standard, a determination must be           synonymous terms.)                           months which is at least inconvenient and at
made, based on substantial evidence in             The CDC estimates that, once infected     worst may involve adverse drug reactions.
                                                with M. tuberculosis, an untreated           (Ex. 7–271)
the record considered as a whole, that
there is a significant health risk under        individual has a 10% lifetime                  The approach taken in this risk
existing conditions and that issuance of        probability of developing active TB and      assessment is similar to the approach
a new standard will significantly reduce        that approximately half of those cases       OSHA took in its risk assessment for the
or eliminate that risk. The Court stated        will develop within the first or second      Bloodborne Pathogens standard. As
that                                            year after infection occurs. Individuals     with bloodborne pathogens, the health
‘‘before he can promulgate any permanent
                                                with active TB represent a pool from         response (i.e., infection) associated with
health or safety standard, the Secretary is     which the disease may spread. Based on       exposure to the pathogenic agent does
required to make a threshold finding that a     data from the CDC, OSHA estimates that       not depend on a cumulative level of
place of employment is unsafe in the sense      every index case (i.e., a person with        exposure; instead, it is a function of
that significant risks are present and can be   infectious TB) results in at least 2 other   intensity and frequency of each
54190               Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules

exposure incident. However, unlike                   workers are appropriate to use as the           demonstrating the occupational
hepatitis B, where the likelihood of                 basis for describing the potential range        transmission of infectious TB. Reports
infection once an exposure incident                  of risks for workers in other work              of TB outbreaks and epidemiologic
occurs is known with some degree of                  settings where workers can be expected          surveillance studies have shown that
certainty, the likelihood of becoming                to come into close and frequent contact         health care and certain other workers
infected with TB after an exposure                   with individuals with infectious TB (or         are, as a result of their job duties, at
incident is not as well characterized.               with other sources of aerosolized M.            significantly higher risk of becoming
With TB, the likelihood of infection                 tuberculosis) as an integral part of their      infected than the average person.
depends on the potency of an exposure                job duties. As discussed in section IV
incident and the susceptibility of the               (Health Effects), epidemiological                  OSHA conducted a thorough search of
exposed individual (which is a function              studies, case reports, and outbreak             the published literature and reviewed
of the person’s natural resistance to TB             investigations have shown that workers          all studies addressing occupational
and his or her health status). Further,              in various work settings, including but         exposure to tuberculosis and TB
the potency of a given exposure incident             not limited to hospitals, have become           infection in hospitals and other work
is highly dependent on several factors,              infected with tuberculosis as a result of       settings. All published studies show
such as the concentration of droplet                 occupational exposure to aerosolized M.         positive results (i.e., workers exposed to
nuclei in the air, the duration of                   tuberculosis when appropriate infection         infectious individuals have a high
exposure, and the virulence of the                   control programs for tuberculosis were          likelihood of becoming infected with
pathogen (e.g., pulmonary and laryngeal              not in place.                                   TB). Because there are so many studies,
TB are considered more infectious than                  In this preliminary risk assessment,         OSHA selected a representative subset
other types).                                        OSHA presents risk estimates for TB             of the more recent studies conducted in
   The Agency has sufficient data to                 infections, cases of active disease, and        the U.S. to include in this section. These
quantify the risk associated with                    TB-related deaths (i.e., where TB is            studies were chosen because they show
occupational exposure to TB among                    considered the cause or a major                 occupational exposure in various work
health care workers in hospitals on a                contributing cause of death) for workers
                                                                                                     settings, under various working
state-by-state basis. In addition to                 with occupational exposure to
hospital employee data, OSHA has                                                                     conditions, and under various scientific
obtained data on selected health care                   A number of epidemiological studies          study designs.
employee groups from the TB Control                  demonstrate an increased risk of TB                OSHA’s summary of the studies is
Office of the Washington State Health                infection among health care workers in          presented in Table V–1(a) and Table V–
Department. These groups include                     hospitals and other work settings. A            1(b). These studies represent a wide
workers employed in long-term health                 brief review of a selection of these            range of occupational settings in
care, home health care, and home care.               studies is presented below, followed by         hospitals, ranging from TB and HIV
Small entities are encouraged to                     OSHA’s estimates of excess risk due to          wards in high prevalence areas, such as
comment and submit any data or studies               occupational exposure. Finally, OSHA            New York City and Miami, to hospitals
on TB infection rates relevant to their              presents a qualitative assessment of the        with no known TB patients located in
business.                                            risk of TB infection caused by                  low prevalence areas such as the state
   Because it is exposure to aerosolized             occupational exposure to tuberculosis in        of Washington. The studies include
M. tuberculosis that places workers at               correctional facilities, homeless shelters,     prospective studies of entire hospitals or
risk of infection, and not some factor               drug treatment centers, medical
unique to the health care profession, the                                                            groups of hospitals, retrospective
                                                     laboratories, and other high-risk work          surveys of well-controlled clinical
Agency concluded that the experience                 groups.
of these groups of health care workers                                                               environments, such as an HIV ward in
is representative of that of the other               Review of the Epidemiology of TB                a hospital, and case studies of single-
‘‘high-risk’’ workers covered by this                Infection in Exposed Workers                    source infection (i.e., outbreak
proposal. This means that the risk                     There are several studies in the              investigations).
estimates calculated for these groups of             published scientific literature

                                       TABLE V–1(A).—OUTBREAK INVESTIGATIONS OF TB INFECTION
       Authors/year                        Setting/source                Risk of TB in health care workers           Contributing factors

Catanzaro (1982) ...........    Hospital intensive care unit/San        14/45 (31%) PPD conversions, 10/13   Poor ventilation. No report on res-
                                  Diego/1 index case—7-day hospital       (77%) PPD conversions among          pirator use.
                                  stay.                                   health care workers present at
Kantor et al. (1988) ........   VA hospital in Chicago autopsy          9/56 (16%) PPD conversions among     No mechanical ventilation on medical
                                 room/1 index case undiagnosed            exposed workers vs. 3/333 (1%)       ward (autopsy room): no isolation.
                                 until histology exam of autopsy tis-     conversions among unexposed          Autopsy room had 11 air changes/
                                 sue.                                     (RR=17.8) 3 workers developed        hour and no air recirculation.
                                                                          active TB.
Beck-Sague (1992) ........      Jackson Memorial Hospital in Miami      13/39 (33%) PPD conversions on       Some rooms had positive pressure.
                                  MDR–TB in HIV/patients on HIV           HIV ward and clinic.                 Inadequate triage of patients with
                                  ward and clinic during 1989–91.                                              suspected TB. Delay in use of iso-
                                                                                                               lation. Early discharge from isola-
                       Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules                                                                  54191

                                                            Study pe-                                                     Risk of TB in health care
      Authors/year                   Setting/source                                       Population                                                         Comments
                                                               riod                                                                workers

Price et al. (1987) .....       19 Eastern North Caro-        1980–84          All Hospital workers .......               1.80% annual PPD con-
                                  lina hospitals.                                                                           version rate.
                                29 Central North Caro-      ................   ........................................   0.70% annual PPD con-
                                  lina hospitals.                                                                           version rate.
                                8 Western North Caro-       ................   ........................................   0.61% annual PPD con-
                                  lina hospitals.                                                                           version rate.
Aitken et al. (1987) ....       64 hospitals in Washing-      1982–84          All Hospital workers .......               0.1% PPD conversion         Strict adherence to CDC
                                  ton State.                                                                                rate/in 3 years.            guidelines.
Malasky et al. (1990)           14 urban hospitals in U.S               (1)    Physicians in training in                  11% PPD conversion/3
                                                                                 pulmonary medicine                         years among pul-
                                                                                 and infectious disease.                    monary fellows, 2.4%
                                                                                                                            PPD conversions/3
                                                                                                                            years among infectious
                                                                                                                            disease fellows.
Dooley et al. (1992) ..         Hospital in Puerto Rico       1989–90          Hospital workers (n=908)                   Prevalence study: 54/       Isolation rooms did not
                                 TB in HIV-infected pa-                                                                     109 (50%) nurses ex-        have negative pres-
                                 tients.                                                                                    posed to TB patients        sure. Recirculated air
                                                                                                                            had positive PPDs 35/       was not filtered.
                                                                                                                            188 (19%) clerical
                                                                                                                            workers with no expo-
                                                                                                                            sure to TB had posi-
                                                                                                                            tive PPDs (p<0.001).
NIOSH .......................   Jackson Memorial Hos-         1989–92          Hospital workers in se-                    60% annual PPD conver-      Incomplete isolation fa-
                                  pital, Miami.                                  lected wards (n=607).                      sion among 263 ex-          cilities. Improper appli-
                                                                                                                            posed workers, 0.6%         cation of isolation pro-
                                                                                                                            annual PPD conver-          cedures.
                                                                                                                            sion among 344 unex-
                                                                                                                            posed workers.
Cocchiarella et al.             Cook County Hospital,               1991       Graduating physicians                      18.8% 3-year PPD con-       Residents were offered
  (1996).                         Chicago.                                       with at least 1 year of                    version rate for house      limited respiratory pro-
                                                                                 clinical work at CCH                       staff in internal medi-     tection during expo-
                                                                                 (n=128).                                   cine vs. 2.2% PPD           sures. No protocol
                                                                                                                            conversion rate for         available for early
                                                                                                                            house staff in other        identification of sus-
                                                                                                                            specialties.                pect TB cases. PPD
                                                                                                                                                        testing program incom-
                                                                                                                                                        plete. Inadequate iso-
                                                                                                                                                        lation facilities.
  1 Mid   1980’s (3 years).

   Outbreak investigations describe                       the failures in control programs                                      studies often include large groups of
occupational exposure to tuberculosis                     contributing to these episodes.                                       employees with little or no exposure to
from single index patients or a well-                       Prospective and/or retrospective                                    TB. Results from such studies may
defined group of patients. Such                           surveillance studies are used to estimate                             reflect an overall estimate of risk in that
investigations are more likely to                         conversion rates from negative to                                     environment, but may underestimate
demonstrate an upper limit of                             positive in PPD skin testing programs.                                the occupational risk of those with
occupational risk in different settings,                  These conversion rates can be used to                                 frequent exposure.
usually under conditions of suboptimal                    estimate the excess incidence of TB                                      Other surveillance studies report PPD
environmental and infection controls.                     infection. Surveillance studies among                                 conversion rates of more narrowly-
Although outbreak investigations                          health care workers lend themselves to                                defined groups of workers, usually those
demonstrate the existence of                              a more systematic evaluation of the risk                              working in ‘‘high-risk’’ areas within a
occupational risk under certain                           of TB infection than outbreak                                         hospital such as the HIV or TB wards.
conditions and the importance of the                      investigations, for several reasons. First,                           Some of these studies have internal
early identification of suspect TB                        these studies better reflect the risk of TB                           control groups (i.e., they compare PPD
patients quite well, these studies do not                 experienced by workers under routine                                  conversion rates between a group of
provide information conducive to risk                     conditions of exposure. Second, these                                 workers with extensive exposure to TB
assessment estimations. Limitations of                    studies are usually based on a larger                                 and a group of workers with minimal or
outbreak investigations include the                       group of workers and therefore yield                                  no exposure to TB), thus making it
frequent absence of baseline PPD test                     more precise and accurate estimates of                                possible to more precisely quantify the
results, the difficulty of extrapolating                  the actual risk of infection. However,                                magnitude of excess risk due to
the results to non-outbreak conditions of                 the extent to which results from                                      occupational exposure. However, these
TB exposure, and, often, small sample                     surveillance studies can be generalized                               studies are also limited in their
sizes. Table V–1(a) lists some of the                     depends on a careful evaluation of the                                usefulness for risk assessment purposes.
published outbreak investigations and                     study population. Some studies report                                 They usually have small sample sizes,
shows the risks posed to health care                      skin test conversion rates for all workers                            making it more difficult to observe
workers by such outbreaks, as well as                     in the hospital(s) under study. Such                                  statistically significant differences. More
54192           Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules

importantly, internal control groups         Rb is the background rate of TB                    OSHA used a multiplicative function
may overestimate background risk, and             infection, and                             of each state’s background infection rate
thus underestimate excess occupational       ERRo is a multiplicative factor denoting        to estimate excess risk of TB infection
risk, unless painstaking efforts are made         the excess relative risk due to            because the probability of occupational
to eliminate from the control group               occupational exposure (ERRo).              infection can be viewed as a function of
those individuals with the potential for        Estimates of ERRo are derived from           the number of contacts and frequency of
occupational exposure, a difficult task      surveillance studies of workers with            contacts with infectious individuals.
in some hospital environments. Table         occupational exposure to TB. ERRo is            Thus, estimates of expected relative
V–1(b) contains a selected list of           defined as the relative difference              increase in risk above background due
published surveillance studies.              between the overall exposed worker risk         to occupational exposure are calculated
   In reviewing Table V–1(a) and Table       and the background (population) risk            for the three available studies and these
V–1(b), the reader should bear in mind       and is calculated as the difference             relative increases (i.e. ERRo) are
that these tables are not intended to        between overall worker and background           multiplied by background rates for each
present an exhaustive list of                risk divided by the background risk.            state to derive estimates of excess
epidemiologic studies with TB                   The annual excess risk due to                occupational risk by state. These state
conversion rates in occupational             occupational exposure is defined as a           estimates are then used to derive a
settings. Instead, these tables present      function of the background risk because         national estimate of occupational risk.
brief summaries of some of the               of data limitations. If data on overall            The CDC compiles and publishes
epidemiologic evidence of occupational       worker risk were available for each             national statistics on the incidence of
TB transmission found in the published                                                       active TB in the U.S. by state based on
                                             state, then the excess risk due to
literature; they are intended to convey                                                      reported cases. OSHA relied on these
                                             occupational exposure would simply be
the seriousness of the risk posed to                                                         data to estimate TB infection
                                             the difference between overall worker
health care workers and to illustrate                                                        background rates through the use of a
                                             risk and background risk. Instead, the
how failures in control programs                                                             mathematical model because
                                             annual excess risk due to occupational
contribute to this risk. Upon reviewing                                                      information on TB infection is not being
                                             exposure (i.e., p) is estimated using a
these studies, a consistent pattern                                                          collected nationwide by CDC. A more
                                             multiplicative model because data on
emerges: these work settings are                                                             detailed discussion on the methodology
                                             overall worker risk (i.e., Rw) were
associated with a high likelihood for                                                        and derivation of background risk
                                             available only for the states of                estimates by state is found in section 3,
occupational exposure to tuberculosis,       Washington, and North Carolina and for
and high rates of TB infection are being                                                     and discussion on the estimation of
                                             Jackson Memorial Hospital located in            occupational risk estimates by state is
observed among health care workers.          Miami, Florida. Therefore, the annual           found in section 4 of this risk
Quantitative Assessment of Risk              excess risk due to occupational                 assessment.
                                             exposure in state i (pi) is expressed as:          Because section 6(b)(5) of the OSH
   Data availability usually dictates the
                                                              (R                 ) ∗R
direction and analytical approach                                                            Act requires OSHA to assess lifetime
OSHA’s risk assessment can take. For                               wj   − R bj               risks, OSHA has converted the annual
this rulemaking, three health endpoints
                                                       pi =                             bi   excess risk due to occupational
                                                                    R bj                     exposure into an excess lifetime risk
will be used: (1) TB infection, which is
‘‘material impairment of health’’ for this   where:                                          based on a 45-year working lifetime.
proposed standard; (2) Active disease        Rwj is the overall worker risk estimated        The formula used to calculate lifetime
following infection; and, (3) Risk of             from surveillance studies (study j),       occupational risk estimates of the
death from active TB.                        Rbj is the study control group risk (i.e.,      probability of at least one occurrence of
   In order to account for regional               study background risk), and                TB infection due to occupational
variability in TB prevalence and             Rbi is the background rate for state i.         exposure in 45 years is expressed as { 1–
therefore to account for expected                                                            (1–p)45 }, where p is the annual excess
                                                When i=j (i.e., Washington State or          risk due to occupational exposure. Two
variability in the risk of TB infection in   North Carolina), the excess risk due to
different areas, the Agency chose to                                                         assumptions are critical in defining
                                             occupational exposure, is expressed as          lifetime risk: (1) the exposure period is
develop occupational risk estimates on       the straight difference between overall
a state-by-state basis. This approach was                                                    45 years, and (2) the annual excess risk
                                             worker risk and background risk.                remains constant. The implication of the
criticized by Dr. Neil Graham as being          OSHA calculated estimates of ERRo
too broad and ’’* * * insufficient in                                                        second assumption is that the worker’s
                                             based on three occupational studies: the        exposure profile and working
light of the tremendous variability          Washington State study, the North
* * * that can occur within a state.’’                                                       conditions, which may affect the level
                                             Carolina study, and the Jackson                 and intensity of exposure, and the
(Ex. 7–911). The Agency recognizes that      Memorial Hospital study (Exs. 7–263, 7–
risk estimates on a county-by-county                                                         virulence of the pathogen, remain
                                             7, 7–108). These estimates were                 unchanged throughout a working
basis would be preferable; however, the      expressed as percent change above each          lifetime. The merit of this assumption
unavailability of comprehensive county       study’s background. The derivation of           was questioned by Dr. Graham, because,
data has prevented the Agency from           these estimates is described in section 2.      as he states ‘‘* * * patient contact may
conducting such analysis.                       In order to estimate an overall range
   The annual excess risk of TB infection                                                    vary greatly throughout a career for
                                             of occupational risk of TB infection,           many HCWs [health care workers].’’ and
due to occupational exposure is defined
                                             taking into account regional differences        ‘‘ * * * physicians (and nurses) often
as a multiplicative function of the
                                             in TB prevalence in the U.S., OSHA: (1)         do not have extensive patient contact
background rate of infection and is
                                             Estimated background TB infection rates         until [their] mid-twenties, while other
expressed as:
                                             by state (Rbi), and (2) applied estimates       workers increasingly retire early.’’ Dr.
p = ERRo * Rb                                of ERRo, derived from the occupational          Graham recommends that OSHA’s risk
where:                                       studies, to the state background rates to       assessment be adjusted to account for
p is the annual excess risk due to           calculate estimates of excess risk due to       variable exposure levels and variable
     occupational exposure,                  occupational exposure by state.                 working lifetimes. Although accounting
                 Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules                               54193

for variable exposure levels could result      infected with TB due to occupational         study by Menzies et al. (Ex. 7–130). Dr.
in more precise risk estimates, the            exposure, and is estimated as a function     Graham admits that the ‘‘validity of the
unavailability of comprehensive                of the background risk. Lifetime             estimates in these reports [reviewed in
information on lifetime TB exposure            occupational risk is defined as the          the Menzies et al. study] must be open
scenarios by occupational group                excess probability of becoming infected      to serious question * * *’’ for the
prevented the Agency from developing           with TB due to exposure in the               following reasons, which were pointed
a more complex risk model.                     workplace, at least once, in the course      out by Dr. Graham: several of the studies
   OSHA has customarily assumed a 45           of a 45-year working lifetime and is         reviewed are very old and not relevant
year working lifetime in setting health        estimated as { 1–(1–p)45 } where p is the    to TB risk in the 1990s; four studies use
standards. The Agency believes that this       annual occupational risk of TB               tine tests and self-reports of skin test
assumption is reasonable and consistent        infection.                                   results, which are not useful for
with the Act. The Act requires the                                                          estimation of risk of TB infection; the
Secretary to set a standard for toxic          2. Data Sources for Estimating
                                                                                            studies were not consistent in the
substances that would assure ‘‘no              Occupational Risk
                                                                                            inclusion of high and low risk workers;
employee * * * suffer material                    The quantitative data needed to           two-step testing was not done; and the
impairment of health or functional             develop an overall national estimate of      participation rates were extremely low
capacity even if such employee has             risk for TB infection due to occupational    or unreported in many of the studies
regular exposures to the hazard for the        exposure are not available. The CDC          included in this review.
period of his working lifetime.’’ 29           does not publish occupational data              OSHA has chosen not to rely on the
U.S.C. § 655(b)(5) (emphasis added).           associated with TB infection incidence       Menzies et al. review study, because, in
The U.S. Court of Appeals for the              and active TB on a nationwide basis.         addition to Dr. Graham’s reservations
District of Columbia upheld the use of         There has been some effort to include        (which the Agency shares), OSHA is
a 45-year lifetime in the asbestos             occupational information on the TB           also concerned about the inclusion in
standard against an assertion by the           reporting forms, but only a limited          the Menzies et al. review article of
Asbestos Information Association that          number of states are currently using the     studies conducted outside the U.S.
the average duration of employment was         new forms that capture occupational          Factors known to affect the
five years. Building and Construction          information in a systematic way.             epidemiology of TB, such as
Trades Department, AFL–CIO v. Brock,              However, there are a number of            environmental conditions, socio-
838 F.2d 1258, 1264, 1265 (D.C. Cir.           sources that permit the risk in              economic status, and work practices, are
1988). The Court said that OSHA’s              occupational settings to be reasonably       expected to differ greatly from one
assumption ‘‘appears to conform to the         estimated and, with the aid of               country to another, and are not
intent of Congress’’ as the standard must      mathematical models, to develop              controlled for in the statistical analyses
protect even the rare employee who             estimates of excess relative occupational    of these studies. For all of these reasons,
would have 45 years of exposure. Id. at        risk (ERRo), which can then be               the Agency has chosen to rely solely on
1264. In addition, while working               multiplied by the state-specific             U.S. studies for its quantitative risk
lifetimes will vary, risk is significant for   background rates to yield estimates of       estimations.
some who work as little as one year and,       excess occupational risk. OSHA has              Estimates of excess risk due to
at any rate, individual and population         identified three data sources that are       occupational exposure are expressed as
risks are likely to remain the same so         suitable for assessing the excess risk of    the percent increase above background
long as employees who leave one job are        TB infection in health care workers with     based on relative risk estimates derived
replaced by others, and those who              occupational exposure. These include:        from occupational studies. Internal
change jobs remain within a covered            (1) A 1994 survey of tuberculin skin         control groups provided estimates of
sector. Nevertheless, the Agency solicits      testing in all health care facilities in     background risk for the Washington
information regarding the likelihood of        Washington State; (2) A state-wide           state and Jackson Memorial data sets. In
exposure to active TB in the workplace         survey of hospitals in North Carolina,       the absence of a suitable internal control
and duration of employment in various          conducted in 1984–1985, which                group, the estimated annual state-wide
occupational groups. Lifetime risk             addressed TB skin testing practices, TB      TB infection rate, as calculated in
estimates of TB infection by state are         infection prevalence, and TB infection       Section 3, was used as the background
described in section 4.                        incidence among hospital employees in        rate in the North Carolina study.
   Lifetime risk estimates of developing       that state; and (3) the employee                (a) Washington State Data: Initially,
active TB are calculated from lifetime         tuberculin skin test conversion database     OSHA relied on a three-year prospective
risk estimates of TB infection assuming        from Jackson Memorial Hospital in            study, conducted between 1982 and
that, once infected, there is a 10%            Miami, Florida. In addition to these         1984 in the state of Washington, to
likelihood of progressing to active TB.        hospital employee data, the Agency has       derive an estimate of excess risk for TB
These estimates are discussed in section       obtained data on selected other work         infection as a result of occupational
4. Further, the number of deaths caused        groups from the state of Washington.         exposure (Ex. 7–42). OSHA received
by TB is calculated from the lifetime          These groups include workers employed        several objections to the use of this
estimates of active TB using OSHA’s            in long-term health care, home health        study. The study used hospitals with no
estimate of TB case fatality rate, also        care, and home care.                         known TB cases as ‘‘controls’’ based on
discussed in section 4.                           On the issue of data availability for     the assumption that in those hospitals
                                               this risk assessment, Dr. Graham agrees      the risk of TB infection to employees
1. Definitions                                 with OSHA that there are no                  may be the same as for the general
   For the purpose of estimating               comprehensive data available with            population. Dr. Qaqish noted that this
incidence rates, TB infection rate is          respect to occupational risk of TB           assumption is highly questionable and
defined as the annual probability of an        infection in health care and other           that the use of such controls is not
individual converting from negative to         institutions in the U.S. Instead of          appropriate. Dr. Graham and Dr. Qaqish
positive in the tuberculin skin test.          relying on two state specific studies, Dr.   pointed out that the published results
Annual occupational risk is defined as         Graham recommends, though with               did not include conversions identified
the annual excess risk of becoming             serious reservations, the use of a review    through contact investigations, which
54194                       Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules

means that the conversion rate reported                                     survey of the state’s tuberculin skin                                facilities. Hospitals had the highest
in that study was likely to be an                                           testing program in 1994. This survey is                              survey response rate (85%) and home
underestimate of the true risk. In                                          conducted by the TB Control Office in                                health care had the lowest (77%). Every
addition, the commenters noted that the                                     the Washington State Health                                          employee at risk for TB infection (i.e.,
study was designed to estimate the                                          Department and it covers all hospitals in                            who was known to be tuberculin skin
effectiveness of the TB screening                                           the state, as well as long-term care,                                test negative at the start of the study
program and may have produced skin                                          home health care, and home care                                      period) in the participating hospitals
testing results biased toward the null;                                     facilities. OSHA was given access to the                             and long-term care facilities was given
the study is relatively old; and, the                                       database for the 1994 survey as well as                              a tuberculin skin test, including
study was conducted prior to the AIDS                                       data on conversions identified through                               administrators, housekeepers, business
epidemic and therefore the results may                                      contact investigations for the same year                             office staff, and all part-time employees.
not be relevant to the occupational risk                                    (Ex. 7–263). Table V–2 summarizes the                                Testing in home health care facilities
at present because the relationship                                         results of the 1994 survey. Of the 335                               was generally confined to those nursing
between HIV and TB is not reflected in                                      health care establishments in the state of                           staff who had direct client contact.
this study.                                                                 Washington, 273 responded to the                                     Employees in home care are those who
   In an effort to respond to reviewers’                                    survey, for an overall response rate of                              provide services to patients in home
comments, the Agency chose to update                                        81.5%. Of those, 76 were hospitals, 142                              health care and include food handlers,
the analysis by relying on a data set of                                    were long-term care, 47 were home                                    cleaning aides, personal care-givers, and
tuberculin skin testing results from a                                      health care, and 8 were home care                                    some social workers.

                                                          TABLE V–2—WASHINGTON STATE 1994 SURVEY RESULTS
                                                                                                                                    Number of a       Number of     Number of
                                                    Type of facility                                                                                                              rate of TB
                                                                                                                                   establishments     skin tests   conversions    conversion

Hospital ......................................................................................................................    76 (85%)               39,290            50     1.27/1,000
Long-term Care .........................................................................................................           142 (81%)              11,332           111     9.80/1,000
Home Health Care .....................................................................................................             47 (77%)                2,172            11     5.06/1,000
Home Care ................................................................................................................         8 (80%)                   537             1     1.86/1,000

      Total ....................................................................................................................   273 (81.5%)            53,331           173     3.24/1,000
   a Numbers        in parentheses are study response rates for each group.

   The overall rate of skin test                                            conversions observed might be late                                   respirators is expected when entering an
conversion for workers in the health                                        boosting because of BCG. However, an                                 isolation room; and (d) all isolation
care system in Washington State in 1994                                     almost two-fold increase in risk for long-                           rooms are under negative pressure, have
was 3.24 per 1,000 employees tested.                                        term care workers even as compared to                                UV lights, and exhaust to the outside. In
This is greater than a 4-fold increase                                      the significant excess risk among home                               addition, conversion data in hospitals
from the estimated state-wide                                               health care workers clearly indicates                                are more likely to represent true TB
background rate of 0.69 per 1,000 at                                        that the risk of TB infection for workers                            infections than in the other health care
risk, as calculated in section 3. The                                       in long-term care is high and not likely                             settings, because hospitals are more
annual rate of TB conversion ranged                                         to be fully explained by late boosting.                              likely to re-test converters in an effort to
from 1.27 per 1,000 tested for hospital                                     Beginning in 1995, two-step testing has                              eliminate false-positive cases.
employees to 9.80 per 1,000 tested for                                      been done on all new hires in                                           A more thorough analysis of the
long-term care employees.                                                   Washington State. Thus, tuberculin skin
   The annual rate of 9.8 per 1,000 for                                                                                                          hospital data is presented in table V–3.
                                                                            testing data for 1995 are not expected to                            Because the Washington State survey
long-term care employees probably                                           be influenced by possible late boosting;
reflects the high potential for exposure                                                                                                         was not designed to compare exposed
                                                                            OSHA will place the 1995 data in the                                 persons with matched controls who
to undiagnosed active TB in those                                           rulemaking record as they become
facilities. As a rule, long-term facilities                                                                                                      have had no exposure, several
                                                                            available.                                                           alternative definitions of an internal
in Washington State do not have AFB
isolation rooms. Therefore, residents                                          Hospital workers had the lowest                                   control (unexposed) group were used in
with no obvious TB symptoms but who                                         overall rate of conversion (overall rate of                          analyzing this data set. Three different
might be infectious spend most of their                                     1.27 per 1,000). This, in part, can be                               analyses, shown in table V–3, produced
time in open spaces exposing other                                          attributed to the existence of extensive                             estimates of annual occupational
residents and workers to infectious                                         TB control measures in that                                          infection rates ranging from 0.4 to 0.6
droplet nuclei. However, once a resident                                    environment in Washington State.                                     per 1,000 above control (i.e., ranging
has been identified as a suspect TB                                         Compliance with the CDC Guidelines                                   from a 47% to an 84% increase above
patient, that person is transferred to a                                    and OSHA’s TB Compliance Directive is                                control). In order to minimize the
hospital until medically determined to                                      quite high in Washington State because:                              likelihood of contaminating the control
be non-infectious.                                                          (a) There is a strong emphasis on early                              group with persons having significant
   Also, since employees who were 35                                        identification of suspect TB patients; (b)                           occupational exposure, OSHA defined
years of age or younger were not given                                      there is a strong emphasis on employee                               the control group as workers in
a two-step test at hiring, and a high                                       training and regular tuberculin skin                                 hospitals located in zero-TB counties
percentage of employees are foreign                                         testing (although on a less-frequent basis                           and with no known TB patients. This
born and therefore most likely to have                                      than recommended in the Guidelines:                                  analysis is summarized in table V–3 as
been vaccinated during childhood with                                       All employees are tested at hire and                                 Definition 1. If potential for
the BCG vaccine, some of the                                                annually thereafter); (c) the use of                                 occupational exposure is defined as
                          Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules                                                                                     54195

either working in a hospital in a county                                  significantly different from a 47%                                       estimate of TB infection rate resulted in
that has active TB or in a hospital that                                  increase and therefore these two                                         an estimate of the annual excess
has had TB patients, then the annual                                      ‘‘control’’ groups can be viewed as                                      occupational risk of approximately 84%
risk due to occupational exposure is                                      producing ‘‘statistically’’ equivalent                                   above background, shown in table V–3
47% above background. The excess                                          results. However, the Agency believes                                    as Definition 3. Estimates of the annual
annual risk due to occupational                                           that Definition 1 is more appropriate,                                   and lifetime occupational risk of TB
exposure appears to be approximately                                      though the risk estimates are higher if                                  infection for the average health care
60% above background, if workers in                                       the control group is defined based on                                    worker in hospitals by state,
hospitals with a transfer-out policy for                                  Definition 2, because there is a higher                                  extrapolated from this study and using
TB patients are considered to be the                                      likelihood of potential for exposure to a
                                                                                                                                                   Definition 1 as the control group, are
control group, shown as Definition 2 in                                   patient with undiagnosed TB under
                                                                                                                                                   presented and summarized in section 4.
table V–3. A 60% increase above                                           Definition 2 conditions. Comparisons of
background is not statistically                                           all hospital TST data to the state-wide

                                                                                                   Number of              Number of             Average con-    Overall con-       Relative risk
     Definition of exposed and control groups                              Sample size             skin tests            conversions             version rate   version rate
                                                                                                     given                observed                    1a             2b        Rate 1          Rate 2

                      Definition 1
Control: Hospitals in zero-TB counties and with
  no-known TB patients ......................................                             16                 1,142                        1            0.477        0.8756     ............   ..............
Exposed: Hospitals in counties reporting TB or
  having TB patients ...........................................                          60               38,148                       49             1.523        1.28447         3.19           1.47
                      Definition 2
Control: Hospitals that transfer out TB patients                                          35                3,645                        3             0.498        0.823      ............   ..............
Exposed: Hospitals with isolation rooms ............                                      41               35,645                       47             1.989        1.3185          3.99           1.602
                      Definition 3
Control: State-wide estimates of annual risk of
  infection ...........................................................    ....................   ....................   ....................         c0.69         c0.69      ............   ..............
Exposed: All PPD testing data ............................                                 76              39,290                        50            1.302        1.27            1.89           1.84

   aRate 1 is estimated as the arithmetic average of hospital specific conversion rates.
   bRate 2 is estimated as the ratio of the sum of all conversions reported divided by the total number of skin tests given within each group.
   cSource: Table V–3(b), state-wide rate of infection.

   Annual rates of excess risk due to                                     care, and a 1-fold increase in annual risk                               conversions divided by the number of
occupational exposure were estimated                                      for workers in home care (see Section 4).                                skin tests administered. (Since most
for long-term care, home health care,                                        Estimates of the range of annual and                                  employees were only given annual
and home care and are presented in                                        lifetime occupational risk for the                                       testing, the number of tests
Section 4. The same control group used                                    average health care worker in long-term                                  administered is a very close estimate of
in the hospital data analysis, Definition                                 care, home health care, and home care                                    the total number of people tested within
1 (i.e., 0.876/1,000 workers at risk) was                                 by state, extrapolated from the                                          a year and thus can be used as the
used to estimate the background risk                                      Washington State study, are presented                                    denominator in estimating infection
among workers in long-term care, health                                   in Section 4.                                                            incidence.) Only 56 out of 167 hospitals
care, and home care facilities and                                           (b) North Carolina Study: A state-                                    reported information on TB conversion
settings. Using 0.876 as the background                                   wide survey of all hospitals in North                                    rates (34% response rate). The authors
infection rate for workers in these                                       Carolina (NC) was conducted in 1984–                                     estimated a state-wide TB infection rate
settings (a) provided a level of                                          1985 (Ex. 7–7). The survey’s                                             of 11.9 per 1,000 per year for hospital
consistency among the Washington data                                     questionnaire was designed to address                                    employees in 1984 and a five-year mean
analyses, and (b) resulted in a lower                                     three main areas of concern affecting                                    annual infection rate of 11.4 per 1,000,
estimate of occupational risk for the                                     hospital employees: (1) Tuberculin skin                                  with a range of 0-89 per 1000 employees
non-hospital health care workplaces                                       testing practices; (2) TB infection                                      at risk for TB infection. An analysis of
than would have resulted had the state-                                   prevalence; and (3) TB infection                                         the data by region (i.e., eastern, central,
wide background risk estimate (i.e.,                                      incidence. The incidence of new                                          western) showed that the eastern region
0.67/1,000 see Section 3) been used.                                      infections among hospital personnel                                      had consistently higher rates (with an
When industry-specific risk data are                                      was assessed over a five-year period by                                  average infection rate of 18.0 per 1,000)
used, there is approximately a 10-fold                                    reviewing tuberculin skin test                                           followed by the central region (7.0 per
increase in annual risk for workers in                                    conversion data during calendar years                                    1,000) and the western region (6.1 per
long-term care, a 5-fold increase in                                      1980 through 1984 and was calculated                                     1000). Results of this study are shown
annual risk for workers in home health                                    as the number of TB skin test                                            in table V–4.
54196                       Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules

                                                                                       1980                     1981                  1992             1993              1984          5-year mean

Eastern ..................................................................                     19.3                     30.8              17.7               11.2             15.7              18.0
                                                                                               (7)                    (10)              (11)               (12)             (18)              (19)
Central ...................................................................                      3.0                     3.7               7.2                6.6             10.0               7.0
                                                                                               (6)                     (8)              (13)               (23)             (25)              (29)
Western .................................................................                        1.9                    13.5               5.3                4.1              7.2               6.1
                                                                                               (2)                     (4)               (4)                (4)              (8)               (8)
   a Conversion   rates are expressed as number of conversions per 1,000 workers tested.
   b In   parentheses is the number of hospitals included in the study.

   Use of this study’s overall results for                                  absence of an internal control group, the                            and issued a report (Ex. 7–108). In
risk estimates was criticized by the peer                                   Agency used the estimated state-wide                                 addition, NIOSH conducted a
reviewers because of design flaws in the                                    background rate of 1.20 per 1,000 as the                             retrospective cohort study of JMH to
study (e.g., high non-response rate,                                        background rate of infection for the                                 determine whether the risk of TB
inconsistent skin testing practices, and                                    western region in North Carolina (see                                infection was significantly greater for
limited two-step testing) and, most                                         Section 3) to estimate excess risk due to                            health care workers who work on wards
importantly, the presence of atypical                                       occupational exposure.1 Based on this                                having patients with infectious TB than
mycobacteria (contributing to false                                         study, annual occupational risk is                                   those who work on wards without TB
positive results) in the eastern part of                                    approximately four times greater than                                patients.
the state. Based on further input from                                      background [(5.98–1.2)/1.2 = 3.98].                                     For the data analysis of this study,
Dr. Comstock, the Agency chose to rely                                      Estimates of the annual and lifetime                                 ‘‘potential for occupational exposure’’
on the study results from the western                                       occupational risk of TB infection based                              was defined based on whether an
region only, because they are considered                                    on this study by state are presented in                              employee worked on a ward that had
to be more representative of the ‘‘true’’                                   Section 4.                                                           records of 15 or more positive cultures
risk of infection and are expected to be                                       (c) Jackson Memorial Hospital Study:                              for pulmonary or laryngeal TB during
less confounded by cross-reactions to                                       Jackson Memorial Hospital (JMH) is a                                 1988–1989. In other words, positive
atypical mycobacteria. Further, the                                         1500-bed general facility located in                                 culture was taken as a surrogate for
Agency chose to rely on the conversion                                      Miami, Florida, employing more than                                  exposure to infectious TB. The authors
rate estimated for 1984 because it was                                      8,000 employees. It is considered one of                             restricted the ‘‘exposed’’ group to
the most recent data reported in the                                        the busiest hospitals in the U.S. It is the                          employees on wards with exposures to
study. Therefore, the western region                                        primary public hospital for Dade County                              pulmonary or laryngeal TB because they
conversion rate of 7.2 per 1,000,                                           and the main teaching hospital for the                               intended to restrict the study to hospital
estimated based on responses to the                                         University of Miami School of                                        workers with exposure to patients with
survey from eight hospitals in 1984, was                                    Medicine. JMH treats most of the TB                                  the highest potential for being
used as an overall worker conversion                                        and HIV cases in Dade County and,                                    infectious. There were 37 wards at JMH
rate. Further, the 1984 rate was adjusted                                   consequently, there is a higher                                      that had submitted at least one positive
by the percent decrease of active TB                                        likelihood of occupational exposure to                               culture during 1988–1989. Seven wards
between 1984 and 1994 in North                                              TB in this facility than in the average                              met the criteria of 15 or more and were
Carolina so that the final worker                                           hospital in the U.S. From March 1988 to                              therefore included in the ‘‘exposed’’
conversion rate for 1994 based on the                                       September 1990, an outbreak of                                       group. These were the medical intensive
western region rates reported in this                                       multidrug-resistant TB (MDR–TB)                                      care unit, five medical wards, and the
study was estimated to be 5.98 (7.2 *                                       occurred among patients and an                                       emergency room. The ‘‘control’’ group
532/641 = 5.98) per 1,000 employees at                                      increased number of TST conversions                                  was defined as hospital workers
risk for TB infection.                                                      was observed among health care                                       assigned to wards with no TB patients
   The North Carolina study did not                                         workers on the HIV ward. This                                        (i.e., wards with no records of positive
have an internal control group to use as                                    prompted a re-evaluation of the                                      cultures during 1988–89). The ‘‘control’’
the basis for estimating excess risk due                                    hospital’s infection control practices                               wards were post-partum, labor and
to occupational exposure because the                                        and the installation of engineering                                  delivery, newborn intensive care unit,
conversion rates presented in this study                                    controls to minimize exposure to TB. As                              newborn intermediate care unit, and
were based on TST results for the entire                                    part of the evaluation of the outbreak,                              well newborn unit. The results of this
hospital employee population. In the                                        NIOSH did a Health Hazard Evaluation                                 analysis are presented in Table V–5.

                                                                                                                                      Exposed           Control          Relative
                                                             Year                                                                                                                       confidence
                                                                                                                                       group            group              risk           interval

1989 ..............................................................................................................................          62.2               6.2             10.1      2.3—44.2
                                                                                                                                       (13/209)           (2/324)

  1 Using the state-wide estimate of population risk                        excess risk due to occupational exposure, because                    the state-wide estimate, which is influenced by the
as the background estimate of risk for this study                           the true background estimate of risk for the western                 large number of infections found in the eastern
most likely results in an underestimate of the true                         region in North Carolina is expected to be less than                 region of that state.
                            Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules                                                                             54197

                                                                                                                                      Exposed          Control           Relative
                                                             Year                                                                                                                     confidence
                                                                                                                                       group           group               risk         interval

1990 ..............................................................................................................................         75.5              6.5             11.7     2.7—50.2
                                                                                                                                      (16/212)          (2/309)
1991 ..............................................................................................................................         31.7              3.5               9.0    1.1—73.8
                                                                                                                                       (6/189)          (1/282)
   a   Rates are expressed as number of conversions per 1,000 workers tested.
   b   Source: Ex. 7–108

   Table V–5 shows a substantially                                          estimate occupational risk of TB                                    in his comment ‘‘Although factors such
elevated risk for those workers with                                        infection in hospital workers with a                                as migration and distribution of the
potential exposure to patients with                                         relatively high likelihood of                                       population may influence this
infectious TB. The relative risk ranges                                     occupational exposure, for the following                            relationship it seems probable that this
from 9 to 11.7 between 1989 and 1991                                        reasons: (a) 1991 represents the most                               assumption is largely correct and
and is statistically significant for all of                                 recent year for which conversion data                               justifiable.’’ (Ex. 7–271). On the other
those years. This suggests that the                                         are available prior to the time when TB                             hand, Dr. Simone expresses concern
excess risk due to occupational                                             infection control measures were fully                               over this assumption and states ‘‘It is
exposure is approximately 8-fold above                                      implemented at JMH; and (b) The higher                              not necessarily true that a change in
background; this is an overall risk                                         conversion rates reported for 1990 and                              cases now reflects the risk of infection
estimate that reflects the occupational                                     1989 (75.5 and 62.2 per 1,000                                       now.’’ Dr. Qaqish demonstrates in his
risk of TB infection for JMH employees                                      respectively) may be atypical, i.e., they                           comment that the net effect of assuming
with patient contact, because this                                          may to some extent reflect the effect of                            a proportional relationship between the
analysis included everyone tested in the                                    the outbreak and not the long-term                                  number of active cases and the number
‘‘exposed’’ and ‘‘control’’ group,                                          occupational risk.                                                  of new infections is to introduce a
regardless of his or her specific job                                          Based on the results of this study,
                                                                                                                                                possible bias into the estimate of
duties or length of patient contact.                                        OSHA estimates that the annual excess
                                                                                                                                                background risk of TB infection,
   An analysis of various occupational                                      risk of TB infection due to occupational
                                                                                                                                                although such a bias could work in
groups within this cohort showed that                                       exposure is 7.95 times greater than
                                                                                                                                                either direction, i.e., toward increasing
nurses and ward clerks in the                                               background. Estimates of annual and
                                                                            lifetime occupational risk of TB                                    or decreasing the estimate of risk. Dr.
‘‘exposed’’ groups had the highest                                                                                                              Qaqish further states that in the absence
conversion rates: 182 and 156                                               infection for the average health care
                                                                            worker in hospitals by state,                                       of more ‘‘relevant data,’’ it is not
conversions per 1,000 workers tested,                                                                                                           possible to determine the actual net
                                                                            extrapolated from this study, are
respectively. Other studies have shown                                                                                                          effect in magnitude and direction of the
                                                                            presented and summarized in section 4.
that health care workers who provide                                                                                                            bias and ‘‘without obtaining additional
direct patient care are at greater risk for                                 3. Estimation of Background Risk of TB                              data, it would be impossible for the
infection than workers who do not                                           Infection                                                           Agency to improve on the accuracy of
provide direct patient care. The high                                         OSHA’s methodology for estimating                                 the risk estimates * * * ’’ OSHA has
risk seen in ward clerks was unexpected                                     population (background) TB infection                                considered all of the reviewer comments
since these workers are not involved in                                     rates relies on the assumption that TB                              and is aware of the inherent uncertainty
direct patient care. However, in the                                        infection occurring in an area can be                               and the potential for bias associated
emergency room, the risk for TST                                            expressed as a numerical function of                                with the use of this assumption;
conversion for the ward clerks was                                          active TB cases reported in the same                                however, in the absence of the
almost three times higher than for the                                      area. If the likelihood of observing any                            additional ‘‘relevant’’ data to which Dr.
nurses, 222 and 83 per 1,000,                                               infection in a population is minimal,                               Qaqish refers, the Agency believes this
respectively. Ward clerks in the                                            then the likelihood of observing active                             approach to be justifiable.
emergency room are responsible for                                          disease diminishes. Conversely, the
clerical processing of patients after                                                                                                             In defining the model used to estimate
                                                                            presence of active TB implies the
triage, handling specimens for the                                                                                                              the annual infection rates occurring in
                                                                            presence of infection, since active
laboratory, and gathering clothing and                                                                                                          a geographical area based on data on
                                                                            disease can only progress from
valuables from admitted patients.                                                                                                               active disease cases reported for the
                                                                            infection. Therefore, there is a
During these interactions, there may                                                                                                            same area, infections progressing to
                                                                            functional relationship linking TB
have been less strict adherence to                                                                                                              active disease are assigned to one of
                                                                            infections to active disease being
infection control measures, and this                                        observed in a particular area during a                              three distinct groups: those occurring
could explain the high conversion rate.                                     specified time period.                                              this year, last year, and in previous
   OSHA used the results from the 1991                                        Peer reviewer comments on this                                    years.
analysis of the data in the JMH study to                                    assumption varied. Neil Graham states                               BILLING CODE 4510–26–P
54198             Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules

BILLING CODE 4510–26–C                           Based on Dr. Graham’s                       Pi(j) be the estimated prevalence of adult
   TB cases reported to CDC each year        recommendation to rely on the                         TB infection in state i during year
are a combination of new and old             progression rates from the MRC study,                 j.
infections that have, for various reasons,   OSHA changed the assumption on the              Ri be the ratio of the number of adult TB
progressed to active disease. Until          progression parameters from 2.5% (first               cases reported in 1993 to the
recently, it was believed that most of the   year), 2.5% (second year), and 5%                     number of adult cases reported in
active cases were the product of old         (remaining lifetime) to 4%, 2% and 4%,                1994 in state i.
infections. However, with the use of         respectively. Therefore the total 10%
DNA fingerprinting techniques,                                                                  The number of TB cases reported in
                                             progression from infection to active            1994 can be expressed as a function of
researchers have reported that a larger      disease is partitioned into 3 groups:
percentage of active cases may be                                                            TB infections expected to have
                                             progression during the first year after         progressed to active disease, by the
attributed to new or recent infections.      infection (40% of all infections that
Small et al. reported, in an article on                                                      following formula:
                                             eventually progress, for a net probability      Ai(1994)=.04*Ii(1994)+.02*Ii(1993)+(.04/
tracing TB through DNA fingerprinting,       of 4%), progression during the second
that as many as 30% of the active cases                                                            73)*Ii(1992)*prob(alive in 1994)
                                             year (20% of all infections that                   +(.04/73)*Ii(1991)*prob(alive in 1994)
reviewed in the study may be the result      eventually progress, for a net probability
of recent infections (Ex. 7–196).                                                               +....
                                             of 2%), and progression during all                 +....
   In this risk assessment, the Agency
                                             subsequent years (the remaining 40% of             +(.04/73)*Ii(1919)*prob(alive in 1994)
assumes the lifetime risk that an
                                             progressing infections). This last                 This can be expressed as:
infection will progress to active TB to be
                                             probability (4%) is assumed to be
approximately 10%. This estimate is                                                          Ai(1994)=.04*Ii(1994)+.02*Ii(1993)+(.04/
                                             uniformly distributed across the
supported by CDC and in her comment,                                                               73)*∑ [Ii(j)*prob(alive in 1994)],
                                             remaining lifespan.
Dr. Simone states that: ‘‘The assumption         TB rates vary considerably by               where j ranges from 1919 to 1992. The
* * * is generally agreed upon.’’ Dr.        geographic area, socio-economic status,         quantity inside the summation symbol
Comstock and Dr. Qaqish both                 and other factors. In an attempt to             is the sum of all people who were
questioned the validity and accuracy of      account for some of those factors, to the       infected with TB between 1919 and
CDC’s estimate. Their comments suggest       extent possible, background TB                  1992 and are still alive in 1994. This
that the true lifetime rate of progression   infection rates have been estimated             summation can be approximated by the
from infection to active disease for         separately for each state. The derivation       prevalence of TB infection in 1992,
adults may be less than 10 percent.          of background infection rates involves          Pi(1992). Therefore, the number of active
However, as Dr. Graham points out, the       several steps for which the process and         TB cases reported in 1994 can be
10% assumption is a widely accepted          formulae are presented below.                   expressed as:
‘‘rule of thumb’’ and is also in relative        Step 1: Background rate of TB               Ai(1994)=.04*Ii(1994)+.02*Ii(1993)+(.04/
agreement with data from the                 infection for state i in year j is defined            73)*Pi(1992)      (2)
unvaccinated control group of the            as:                                             Further, if we assume that the number
British Medical Research Council (MRC)
                                             Bi(j)=Ii(j)/Xi(j)    (1)                        of new infections is directly
vaccination trial in adolescents (Ex. 7–
                                             where:                                          proportional to the number of active
   In the MRC study, 1,338 adolescents’      Bi(j) is the background TB infection rate       cases, then Ii(1993) can be expressed as
skin tests converted following TB                   for state i in year j                    follows:
exposure where the precise date of           Ii(j) is an estimate of the number of new       Ii(1993)=Ii(1994)*(Ai(1993)/Ai(1994))  (3)
conversion was known. Of these, 108                 infections that occurred in state i in   and (2) can be expressed as:
(8.1%) individuals developed active TB              year j                                   Ai(1994)=[(.02*(Ai(1993)/
during follow-up. Of these, 54%              Xi(j) is the population at risk for TB                Ai(1994))+.04]*Ii(1994)+(.04/73)*Pi(1992)
developed active TB within one year                 infection in state i in year j.          Ai(1994)=[(.02*Ri+.04]*Ii(1994)+(.04/
and 78% within 2 years. This results in          Step 2: Estimation of Ii(j), the number           73)*Pi(1992)      (4)
a risk of approximately 4% at one year,      of new TB infections:
                                                                                             then solving for Ii(1994) becomes: 2
6% at two years, and an overall risk of      Let:
8%. Given that the risk of TB                Ai(j) be the total number of adult TB              2 Using the prevalence of TB infection in 1992

reactivation increases with age, the                cases reported to CDC by state i in      (i.e., Pi(1992)) to approximate the quantity inside the
lifetime risk is expected to be higher              year j.                                  summation sign (i.e., everyone infected between
                                                                                             1919 and 1992 and alive in 1994) slightly
than the 8% attained in this study and,      A(j) be the total number of adult TB            overestimates the quantity inside the summation
as Dr. Graham points out, a 10% overall             cases reported to CDC by all states      (i.e., Pi(1992) is slightly larger than the quantity it
lifetime risk seems reasonable.                     in year j.                               approximates.) It includes a small number of people
                            Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules                                                                                 54199

Ii(1994)=[Ai(1994)¥.04/73*Pi(1992)]/                                            Step 4: Estimation of population                                  infection prevalence has been designed
      (.02*Ri+.04)      (5)                                                  currently infected as of 1993 by state,                              to capture the transmission dynamics of
   Step 3: Estimation of Xi(1994):                                           Pi(1993):                                                            TB by modeling transfers between a
                                                                                The prevalence of TB infection in                                 series of age-stratified compartments
   Xi(1994), the population at risk for TB
                                                                             each state is estimated as a function of                             using a system of differential equations.
infection in state i in 1994, is estimated
                                                                             TB infection prevalence in the U.S. in                               The model adjusts for various
as follows:
                                                                             1993 and the percent TB case rate for                                epidemiological factors known to
Xi(1994)=Ni¥Pi(1993)      (6)                                                each state.                                                          influence the course of active TB, such
Where:                                                                       Pi(1993)=P(1993)*(Ai(1993)/A(1993)) (7)                              as onset of infection (i.e., old vs. new
Ni is the adult population for state i as                                    Where:                                                               infections) and the impact of
      reported by U.S. Census in 1994.                                                                                                            immigration rates and the HIV
                                                                             P(1993) is the prevalence of TB infections
                                                                                                                                                  epidemic. However, it does not
Pi(1993) is the estimated number of                                               in the U.S. in 1993 (Ex. 7–66) and
                                                                             A(1993) is the total number of adult TB                              differentiate among gender or race
      infected adults in state i in 1993
                                                                                  cases reported in 1993.                                         categories. The model has been
      (i.e., prevalence of TB infection in
                                                                                                                                                  successfully validated using actual
      state i among adults).                                                    Estimates of TB infection prevalence                              epidemiological data on active TB from
To estimate the number of adults                                             in the U.S. were developed for OSHA by                               1965 to 1994. The estimates of TB
currently at risk for TB infection in each                                   Dr. Christopher Murray of the Harvard                                prevalence rates presented here are
state, the number of already infected                                        Center for Population and Development                                specific for adults (i.e., older than 18
adults (i.e., prevalence of TB infection Pi                                  Studies and are presented in Table V–                                years of age), which make them more
in 1993) is subtracted from the adult                                        6 (Ex. 7–267). The mathematical model                                appropriate for estimating risk of
population in 1994.                                                          used by Dr. Murray to estimate TB                                    transmission in an occupational setting.

                                          TABLE V–6.—NATIONAL PREVALENCE OF TB INFECTION IN ADULTS (18+)                                                            ab

                                                                  Year                                                                          Expected           Minimum              Maximum

1992 ........................................................................................................................................         6.87%              6.53%                7.22%
                                                                                                                                                (12,978,461)       (12,336,150)         (13,639,663)
1993 ........................................................................................................................................         6.64%              6.31%                6.97%
                                                                                                                                                (12,667,062)       (12,037,524)         (13,296,599)
1994 ........................................................................................................................................         6.47%              6.14%                6.79%
                                                                                                                                                (12,449,445)       (11,814,465)         (13,065,182)
   a Numbers        in parentheses are population prevalence figures.
   b Estimated       for OSHA by Christopher Murray MD, PhD, Harvard University, Center for Population and Development Studies (Ex. 7–267).

  To estimate the number of previously                                       equation (6)). The number of new                                     Table V–7(b), infection rates are based
infected adults in each state (Pi), the                                      infections expected to have occurred in                              on the minimum value of TB infection
estimated national TB prevalence figure                                      1994 was estimated using equation (5).                               prevalence in the U.S. (i.e., 12,037,524).
was multiplied by the active cases for                                         The background rate of TB infection                                In Table V–7(c), infection rates are
each state and divided by the total                                          for 1994 was then estimated by dividing                              based on the maximum value of TB
number of active cases reported [see                                         the number of new infections (Ii) by the                             infection prevalence in the U.S. (i.e.,
equation (7)] (i.e., the national                                            number of susceptible adults in each                                 13,296,599). An overall range of
prevalence estimate was apportioned                                          state (Xi) (see equation (1)).                                       background annual TB infection rates
among the states based on each state’s                                         Results on estimated TB background                                 was constructed by combining all three
percent contribution to active TB                                            annual infection rates for each state are
                                                                                                                                                  sets of infection rates and was estimated
reported for 1993). To estimate the                                          presented in Table V–7(a)—Table V–
                                                                                                                                                  to be between 0.194 and 3.542 per 1,000
number of adults at risk of TB infection,                                    7(c). In Table V–7(a) TB infection rates
                                                                                                                                                  individuals at risk of TB infection, with
(Xi), the number of already infected                                         are based on an average value of TB
adults was subtracted from the adult                                         infection prevalence, as estimated by Dr.                            a weighted average of 1.46 per 1,000
population estimate for each state (see                                      Murray, in the U.S. (i.e., 12,667,062). In                           using state population size as weights.

                                         TABLE V–7(a).—ESTIMATES OF ANNUAL BACKGROUND TB INFECTION RATES a
                                                                                                 [Referent Year 1994]

                                                                                                                              Population                                               Annual popu-
                                                                      TB cases re-                 Population                                    Population at     Estimate of
                            State                                                                                            currently in-                                             lation rate of
                                                                     ported in 1994                  size a                    fected b              risk         new infections        TB infection

                                                                               Ai                         Ni                      Pi(1993)            Xi                 Ii                 Bi

Alabama (01) ............................................                             413                     3,139                  250,083        2,888,917                  4,779             1.65
Alaska (02) ................................................                           78                       414                   27,787          386,213                  1,182             3.06
Arizona (04) ..............................................                           233                     2,936                  118,231        2,817,769                  2,858             1.01
Arkansas (05) ............................................                            235                     1,813                  107,334        1,705,666                  2,906             1.70
California (06) ...........................................                         4,291                    22,754                2,437,044       20,280,956                 47,852             2.36

who were infected with TB and were alive as of                               technically, are not included in the summation.                      be, resulting in an underestimate of the number of
1992 and who were therefore included in the                                  This implies that, in equation (5), a slightly larger                new infections in 1994 and an underestimate of the
prevalence figure, but who died before 1994, and,                            number is being subtracted from Ai(1994) than should                 occupational risk.
54200                      Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules

                           TABLE V–7(a).—ESTIMATES OF ANNUAL BACKGROUND TB INFECTION RATES a—Continued
                                                                                     [Referent Year 1994]

                                                                                                        Population                                            Annual popu-
                                                                   TB cases re-        Population                         Population at    Estimate of
                           State                                                                       currently in-                                          lation rate of
                                                                  ported in 1994         size a                               risk        new infections
                                                                                                         fected b                                              TB infection

                                                                        Ai                 Ni             Pi(1993)             Xi               Ii                 Bi

Colorado (08) ............................................                      90            2,686              52,850      2,633,150                1,045             0.40
Connecticut (09) ........................................                      144            2,487              81,182      2,405,818                1,665             0.69
Delaware (10) ...........................................                       51              531              26,152        504,848                  671             1.33
D.C. (11) ...................................................                  116              451              80,092        370,908                1,162             3.13
Florida (12) ................................................                1,675           10,691             846,687      9,844,314               20,545             2.09
Georgia (13) ..............................................                    676            5,162             396,646      4,765,354                7,082             1.49
Hawaii (15) ................................................                   234              875             132,942        742,058               25,890             3.49
Illinois (17) .................................................              1,021            8,669             622,211      8,046,789               10,994             1.37
Indiana (18) ...............................................                   201            4,279             129,673      4,149,327                2,083             0.50
Iowa (19) ...................................................                   62            2,180              31,056      2,068,943                  859             0.42
Kansas (20) ...............................................                     77            1,864              37,049      1,826,951                1,065             0.58
Kentucky (21) ............................................                     316            2,857             203,227      2,653,773                3,273             1.23
Louisiana (22) ...........................................                     412            3,080             185,792      2,894,208                5,582             1.93
Maine (23) .................................................                    31              934              14,712        919,289                  419             0.46
Maryland (24) ............................................                     344            3,743             211,399      3,531,601                3,582             1.01
Massachusetts (25) ...................................                         299            4,617             183,067      4,433,933                2,889             0.65
Michigan (26) ............................................                     438            6,971             246,269      6,724,731                5,036             0.75
Minnesota (27) ..........................................                      127            3,326              68,105      3,257,895                1,413             0.43
Mississippi (28) .........................................                     262            1,913             141,659      1,771,341                3,120             1.76
Missouri (29) .............................................                    241            3,899             128,583      3,770,417                2,922             0.78
Montana (30) .............................................                      22              618              11,987        606,013                  290             0.48
Nebraska (31) ...........................................                       22            1,181              12,531      1,168,469                  233             0.20
Nevada (32) ..............................................                     111            1,181              50,670      1,130,330                1,514             1.34
New Hampshire (33) .................................                            17              845              13,076        831,924                  182             0.22
New Jersey (34) ........................................                       764            5,973             456,579      5,516,421                8,150             1.48
New Mexico (35) .......................................                         78            1,156              35,415      1,120,585                  944             0.84
New York (36) ...........................................                    3,414           13,658           2,044,797     11,613,203               34,728             2.99
North Carolina (37) ...................................                        532            5,314             298,574      5,015,426                6,000             1.20
North Dakota (38) .....................................                         10              466               3,813        426,186                  132             0.29
Ohio (39) ...................................................                  318            8,248             161,274      8,086,726                3,763             0.47
Oklahoma (40) ..........................................                       231            2,378             101,886      2,276,114                3,064             1.35
Oregon (41) ...............................................                    146            2,303              78,457      2,224,543                1,793             0.81
Pennsylvania (42) .....................................                        583            9,154             379,211      8,774,789                5,886             0.67
Rhode Island (44) .....................................                         47              757              31,601        725,399                  495             0.68
South Carolina (45) ...................................                        362            2,712             205,406      2,506,594                4,273             1.70
South Dakota (46) .....................................                         26              513               8,173        504,827                  342             0.68
Tennessee (47) .........................................                       494            3,878             283,863      3,594,137                5,759             1.60
Texas (48) .................................................                 2,276           13,077           1,199,200     11,877,800               27,306             2.30
Utah (49) ...................................................                   47            1,236              23,973      1,212,027                  427             0.35
Vermont (50) .............................................                      10              434               2,724        431,276                  160             0.37
Virginia (51) ...............................................                  330            4,949             226,110      4,722,890                3,220             0.68
Washington (53) ........................................                       241            3,935             142,729      3,792,251                2,554             0.67
West Virginia (54) .....................................                        80            1,393              40,318      1,352,682                  919             0.68
Wisconsin (55) ..........................................                      104            3,735              50,126      3,684,874                1,307             0.35
Wyoming (56) ............................................                       12              339               3,814        335,186                  188             0.56
   a Expressed     in thousands.
   b Based     on 6.64% rate of TB infection prevalence in the U.S. (expected)

                                            TABLE V–7(b).—Estimates of Annual Background TB Infection Rates
                                                                                     [Referent Year 1994 a]

                                                                                                        Population                                            Annual popu-
                                                                   TB cases re-        Population                         Population at    Estimate of
                           State                                                                       currently in-                                          lation rate of
                                                                  ported in 1994         size a          fected b             risk        new infections       TB infection

                                                                        Ai                 Ni            Pi(1993)              Xi               Ii                 Bi

Alabama (01) ............................................                      413            3,139             237,654      2,901,346                4,871             1.68
Alaska (02) ................................................                    78              414              26,406        387,594                1,196             3.09
Arizona (04) ..............................................                    233            2,936             112,355      2,823,645                2,913             1.03
Arkansas (05) ............................................                     235            1,813             102,000      1,711,000                2,967             1.73
California (06) ...........................................                  4,291           22,754           2,350,136     20,403,864               48,956             2.40
Colorado (08) ............................................                      90            2,686              50,223      2,635,777                1,066             0.40
Connecticut (09) ........................................                      144            2,487              77,147      2,409,853                1,700             0.71
Delaware (10) ...........................................                       51              531              24,853        506,147                  681             1.34
D.C. (11) ...................................................                  116              451              76,111        374,889                1,192             3.18
                           Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules                                                          54201

                                 TABLE V–7(b).—Estimates of Annual Background TB Infection Rates—Continued
                                                                                      [Referent Year 1994 a]

                                                                                                         Population                                            Annual popu-
                                                                    TB cases re-        Population                         Population at    Estimate of
                           State                                                                        currently in-                                          lation rate of
                                                                   ported in 1994         size a                               risk        new infections
                                                                                                          fected b                                              TB infection

                                                                         Ai                 Ni            Pi(1993)              Xi               Ii                 Bi

Florida (12) ................................................                 1,675           10,691             804,607      9,886,393               20,944             2.12
Georgia (13) ..............................................                     676            5,162             376,933      4,785,067                7,275             1.52
Hawaii (15) ................................................                    234              875             126,335        748,665                2,652             3.54
Illinois (17) .................................................               1,021            8,669             591,288      8,077,712               11,260             1.39
Indiana (18) ...............................................                    201            4,279             123,228      4,155,772                2,136             0.51
Iowa (19) ...................................................                    62            2,180              29,513      2,070,487                  869             0.42
Kansas (20) ...............................................                      77            1,864              35,208      1,828,792                1,079             0.59
Kentucky (21) ............................................                      316            2,857             193,126      2,663,874                3,357             1.26
Louisiana (22) ...........................................                      412            3,080             176,558      2,903,442                5,667             1.95
Maine (23) .................................................                     31              934              13,980        920,020                  425             0.46
Maryland (24) ............................................                      344            3,743             200,893      3,542,107                3,677             1.04
Massachusetts (25) ...................................                          299            4,617             173,969      4,443,031                2,983             0.67
Michigan (26) , ..........................................                      438            6,971             234,030      6,736,970                5,144             0.76
Minnesota (27) ..........................................                       127            3,326              64,721      3,261,279                1,448             0.44
Mississippi (28) .........................................                      262            1,913             134,619      1,778,381                3,183             1.79
Missouri (29) .............................................                     241            3,899             122,193      3,776,807                2,978             0.79
Montana (30) .............................................                       22              618              11,391        606,609                  294             0.48
Nebraska (31) ...........................................                        22            1,181              11,909      1,169,091                  240             0.21
Nevada (32) ..............................................                      111            1,181              48,152      1,132,848                1,536             1.36
New Hampshire (33) .................................                             17              845              12,426        832,574                  185             0.22
New Jersey (34) ........................................                        764            5,973             433,887      5,539,113                8,357             1.51
New Mexico (35) .......................................                          78            1,156              33,655      1,112,345                  965             0.86
New York (36) ...........................................                     3,414           13,658           1,943,173     11,714,827               35,735             3.05
North Carolina (37) ...................................                         532            5,314             283,735      5,030,265                6,138             1.22
North Dakota (38) .....................................                          10              466               3,624        462,376                  134             0.29
Ohio (39) ...................................................                   318            8,248             153,259      8,094,741                3,845             0.48
Oklahoma (40) ..........................................                        231            2,378              96,822      2,281,178                3,116             1.37
Oregon (41) ...............................................                     146            2,303              74,558      2,228,442                1,825             0.82
Pennsylvania (42) .....................................                         583            9,154             360,365      8,793,635                6,047             0.69
Rhode Island (44) .....................................                          47              757              30,030        726,970                  506             0.70
South Carolina (45) ...................................                         362            2,712             195,197      2,516,803                4,356             1.73
South Dakota (46) .....................................                          26              513               7,766        505,234                  350             0.69
Tennessee (47) .........................................                        494            3,878             269,756      3,608,244                5,875             1.63
Texas (48) .................................................                  2,276           13,077           1,139,601     11,937,399               27,853             2.33
Utah (49) ...................................................                    47            1,236              22,782      1,213,218                  446             0.37
Vermont (50) .............................................                       10              434               2,589        431,411                  162             0.37
Virginia (51) ...............................................                   330            4,949             214,873      4,734,127                3,311             0.70
Washington (53) ........................................                        241            3,935             135,654      3,799,346                2,621             0.69
West Virginia (54) .....................................                         80            1,393              38,315      1,354,685                  941             0.69
Wisconsin (55) ..........................................                       104            3,735              47,634      3,687,366                1,332             0.36
Wyoming (56) ............................................                        12              339               3,624        335,376                  190             0.57
   a Expressed     in thousands.
   b Based     on a 6.31% rate of TB infection in the U.S.

                                         TABLE V–7(c).—ESTIMATES OF ANNUAL BACKGROUND TB INFECTION RATES
                                                                                      [Referent Year 1994 a]

                                                                                                         Population                                            Annual popu-
                                                                    TB cases re-        Population                         Population at    Estimate of
                           State                                                                        currently in-                                          lation rate of
                                                                   ported in 1994         size                                 risk        new infections
                                                                                                          fected b                                             TB infection,

                                                                         Ai                 Ni            Pi (1993)             Xi               Ii                 Bi

Alabama (01) ............................................                       413            3,139             262,512      2,876,488                4,685             1.63
Alaska (02) ................................................                     78              414              29,168        384,832                1,167             3.03
Arizona (04) ..............................................                     233            2,936             124,107      2,811,893                2,801             1.00
Arkansas (05) ............................................                      235            1,813             112,669      1,700,332                2,843             1.67
California (06) ...........................................                   4,291           22,754           2,595,951     20,158,049               46,720             2.32
Colorado (08) ............................................                       90            2,686              55,476      2,630,524                1,024             0.39
Connecticut (09) ........................................                       144            2,487              85,216      2,401,784                1,629             0.68
Delaware (10) ...........................................                        51              531              27,452        503,508                  661             1.31
D.C. ...........................................................                116              451              84,072        366,928                1,131             3.08
Florida (12) ................................................                 1,675           10,691             888,766      9,802,234               20,137             2.05
Georgia (13) ..............................................                     676            5,162             416,359      4,745,641                6,884             1.45
Hawaii (15) ................................................                    234              875             139,539        735,451                2,526             3.43
Illinois (17) .................................................               1,021            8,669             653,134      8,015,866               10,721             1.34
54202                     Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules

                            TABLE V–7(c).—ESTIMATES OF ANNUAL BACKGROUND TB INFECTION RATES—Continued
                                                                                   [Referent Year 1994 a]

                                                                                                      Population                                             Annual popu-
                                                                 TB cases re-        Population                         Population at    Estimate of
                          State                                                                      currently in-                                           lation rate of
                                                                ported in 1994         size                                 risk        new infections
                                                                                                       fected b                                              TB infection,

                                                                      Ai                 Ni            Pi (1993)             Xi               Ii                  Bi

Indiana (18) ...............................................                 201            4,279             136,117      4,142,883                2,029              0.49
Iowa (19) ...................................................                 62            2,180              32,600      2,067,401                   849             0.41
Kansas (20) ...............................................                   77            1,864              38,891      1,825,109                1,052              0.58
Kentucky (21) ............................................                   316            2,857             213,327      2,643,673                3,187              1.21
Louisiana (22) ...........................................                   412            3,080             195,025      2,884,975                5,496              1.91
Maine (23) .................................................                  31              934              15,442        918,558                   413             0.45
Maryland (24) ............................................                   344            3,743             221,905      3,521,095                3,484              0.99
Massachusetts (25) ...................................                       299            4,617             192,166      4,424,834                2,793              0.63
Michigan (26) ............................................                   438            6,971             258,508      6,712,492                4,925              0.73
Minnesota (27) ..........................................                    127            3,326              71,490      3,254,510                1′,377             0.42
Mississippi (28) .........................................                   262            1,913             148,700      1,764,300                3,057              1.73
Missouri (29) .............................................                  241            3,899             134,973      3,764,027                2,865              0.76
Montana (30) .............................................                    22              618              12,582        605,418                   286             0.48
Nebraska (31) ...........................................                     22            1,181              13,154      1,167,846                   227             0.20
Nevada (32) ..............................................                   111            1,181              53,189      1,127,811                1,491              1.32
New Hampshire (33) .................................                          17              845              13,726        831,274                   178             0.21
New Jersey (34) ........................................                     764            5,973             479,270      5,493,730                7,938              1.44
New Mexico (35) .......................................                       78            1,156              37,175      1,118,825                   922             0.82
New York (36) ...........................................                  3,414           13,658           2,146,421     11,511,421               33,696              2.92
North Carolina (37) ...................................                      532            5,314             313,413      5,000,587                5,859              1.17
North Dakota (38) .....................................                       10              466               4,003        461,997                   129             0.28
Ohio (39) ...................................................                318            8,248             169,289      8,078,711                3,678              0.46
Oklahoma (40) ..........................................                     231            2,378             106,949      2,271,051                3,011              1.33
Oregon (41) ...............................................                  146            2,303              82,357      2,220,643                1,760              0.80
Pennsylvania (42) .....................................                      583            9,154             398,057      8,755,943                5,722              0.66
Rhode Island (44) .....................................                       47              757              33,171        723,829                   483             0.67
South Carolina (45) ...................................                      362            2,712             215,614      2,496,386                4,188              1.68
South Dakota (46) .....................................                       26              513               8,579        504,421                   334             0.67
Tennessee (47) .........................................                     494            3,878             297,971      3,580,029                5,641              1.58
Texas (48) .................................................               2,276           13,077           1,258,799     11,818,201               26,746              2.26
Utah (49) ...................................................                 47            1,236              25,165      1,210,835                   408             0.34
Vermont (50) .............................................                    10              434               2,860        431,140                   158             0.37
Virginia (51) ...............................................                330            4,949             237,347      4,711,653                3,126              0.66
Washington (53) ........................................                     241            3,935             149,843      3,785,157                2,485              0.66
West Virginia (54) .....................................                      80            1,393              42,322      1,350,679                   896             0.66
Wisconsin (55) ..........................................                    104            3,735              52,617      3,682,383                1,283              0.35
Wyoming (56) ............................................                     12              339               4,003        334,997                   185             0.55
   a Expressed     in thousands.
   b Based     on 6.97% rate of TB infection prevalence in the U.S. (maximum estimate).

   Step 5 Model validation:                                         years, formula (8) predicts that infection           with moderate TB prevalence and
   An alternative, but less sophisticated,                          rates can range from 1.45 to 1.61 per                inadequate TB infection control
way to estimate annual risk of infection,                           1,000. These results are in close                    programs. Finally, the Jackson Memorial
if prevalence is known in a specific age                            agreement with OSHA’s weighted                       Hospital data are representative of
group, is to use the following formula:                             average estimate of the national TB                  county hospitals serving high-risk
Annual Rate of Infection = -ln(1-P)/d                               infection rate, which is 1.46 per 1,000.             patients whose employees have a high
     (8)                                                                                                                 frequency of exposure to infectious TB.
                                                                    4. Occupational Risk Estimations                     These data sources provide information
P is the percent prevalence of infection                              OSHA used the three different data                 on the magnitude of the expected excess
     and                                                            sources to obtain estimates of risk of TB            risk in three different environments, and
d is the average age of the population                              infection for health care employees: the             are used to provide a range of possible
     (Ex. 7–265).                                                   Washington State data, the North                     values of excess risk.
   In order to validate the model used by                           Carolina study, and the NIOSH Health                    Based on the Washington State data,
OSHA to estimate background infection                               Hazard Evaluation (HHE) from Jackson                 the annual risk is expected to be 1.5
rates, estimates of TB infection                                    Memorial Hospital (Exs. 7–263, 7–7, 7–               times the background rate for hospital
prevalence for 1994 were used to                                    108). The Washington State data                      employees, approximately 11 times the
calculate predicted infection rates using                           represent workplaces located in low TB               background rate for long-term care
equation (8). Based on Murray’s model,                              prevalence areas, where TB infection                 employees, 6 times the background rate
TB infection prevalence is expected to                              control measures and engineering                     for home health care workers, and
range from 6.31% to 6.97% in 1994                                   controls are required by state health                double the background rate for home
among adults (18+). Using these figures                             regulations. The North Carolina data                 care employees. Based on the North
and assuming the average age to be 45                               represent workplaces located in areas                Carolina data, the annual risk is
               Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules                       54203

expected to be approximately 5 times     calculated by applying the excess          minimum and maximum estimates of
the background rate. Based on the        relative risk ratios, derived from the     excess risk among all states for each
Jackson Memorial Hospital data, the      three occupational studies, to the         data source. These results are presented
annual risk is expected to be            overall background rate of infection for   in table V–9 and table V–10 for workers
approximately 9 times the background.    each state and are presented in table V–   in hospitals and for workers in other
  Estimates of expected excess risk of   8(a)—table V–8(c). A range of excess       work settings, respectively.
TB infection for workers with            risk of TB infection due to occupational
                                                                                    BILLING CODE 4510–26–P
occupational exposure by state are       exposure is constructed by using the
54204   Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules
                  Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules   54205

54206                      Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules

                                         TABLE V–9.—OCCUPATIONAL RISK ESTIMATES FOR HOSPITAL EMPLOYEES a
                                                                                                                                           Excess risk       Range of excess occupational
                                                                                          Overall risk/                                     based on                    risk d
                                     Source                                                                        risk based on
                                                                                           (exposed)                                          study
                                                                                                                        study               (percent)           Annual            Lifetime

Washington State 1994 data ............................................                         1.24/1000                0.88/1000                    47         0.09–1.66          4.1–72.2
North Carolina Western Counties .....................................                         b 5.98/1000               d 1.20/1000                  398         0.77–14.1         34.2–472
Jackson Memorial (1991) .................................................                       31.7/1000                  3.5/1000                  795         1.54–28.2         67.1–723
   a Background  TB infection rate ranges from 0.194 to 3.542 per 1,000 at risk for TB infection.
   b Ajustedfor 1994, i.e., 5.98=7.2*(532/641)
   c Therange reflects regional differences in TB prevalence as well as inherent uncertainty in the estimate of TB infection prevalence in the U.S.,
as estimated by Dr. Christopher Murray, and used in the internal calculations of annual background TB infection rate.
  d State-wide estimate of population risk for North Carolina, shown in Table V–3(a).

                                     TABLE V–10.—OCCUPATIONAL RISK ESTIMATES FOR OTHER WORK SETTINGS a,b
                                                                                                                                           Excess risk       Range of excess occupational
                                                                                          Overall risk/                                     based on                    risk d
                                      Type                                                                           risk State-
                                                                                           (exposed)                                          study
                                                                                                                        wide c              (percent)           Annual            Lifetime

Long-term Care .................................................................                 9.8/1000             0.8756/1000                   1019         1.98–36.1           85–807
Home Health Care ............................................................                   5.06/1000             0.8756/1000                    478         0.93–16.9         40.9–526
Home Care .......................................................................               1.86/1000             0.8756/1000                    112         0.22–3.97          9.7–164
   a Background TB infection rate ranges from 0.194 to 3.542 per 1,000 employees at risk of infection.
   b Basedon the Washington State data.
  c Background rate for this analysis is assumed to be the same as in the case-control analysis of the Washington State hospital data (i.e.
0.8756 per 1,000 employees).
  d The range reflects regional differences in TB prevalence as well as inherent uncertainty in the estimate of TB infection prevalence in the U.S.,
as estimated by Dr. Christopher Murray, and used in the internal calculations of annual background TB infection rate.

  Lifetime estimates of the excess risk of                               risk estimates of developing active TB                                death caused by TB is calculated from
TB infection were estimated based on                                     are calculated from lifetime risk                                     the lifetime estimates of active TB using
the annual excess risk by using the                                      estimates of TB infection assuming that,                              OSHA’s estimate of the TB case fatality
formula {1–(1–p) 45}, where p is the                                     once infected, there is a 10% likelihood                              rate (also presented in table V–11 and
annual excess risk. Lifetime excess                                      of progressing to active TB; these                                    table V–12). The methodology used to
estimates of TB infection are presented                                  estimates are presented in table V–11                                 estimate a TB case fatality rate is
in table V–9 and table V–10. Lifetime                                    and table V–12. Further, the risk of                                  presented below.

                                                                                                                                                                                Death caused
                                                               Source                                                                      TB infection d    Active disease e      by TB

Washington State (1994) .........................................................................................................               4.1–72.2            0.4–7.2            0.03–0.6
North Carolina Western Region ...............................................................................................                  34.2–472             3.4–47.2            0.3–3.7
Jackson Memorial Hospital (Miami) .........................................................................................                    67.1–723             6.7–72.3            0.5–5.6
   a Risk estimates reflect excess risk due to occupational exposure and are expressed per 1,000 employees at risk.
  b Estimates of death caused by TB due to occupational exposure are derived based on an estimated TB case death rate of 77.85 per 1,000
TB cases and are estimated by multiplying the lifetime active disease rate by .07785.
  c The ranges of risk presented in this TABLE reflect expected variance in the annual background TB infection rate by state. They are estimated
based on the assumption that the annual background TB infection rate ranges from 0.194 to 1.542 per 1,000 employees at risk.
  d Lifetime infection rate is estimated by (1–(1–p) 45), where p is the annual excess TB infection rate due to occupational exposure.
  e Lifetime active disease rate is estimated to be 10% of lifetime infection rate.

              TABLE V–12—LIFETIME OCCUPATIONAL RISK ESTIMATES FOR EMPLOYEES IN OTHER WORK SETTINGS                                                                               abc

                                                                                                                                                                                Death caused
                                                          Work setting                                                                    TB infection d    Active disease e       by TB

Long-term Care ......................................................................................................................           85–807            8.5–80.7             0.7–6.2
Home Health Care .................................................................................................................             40.9–536           4.1–53.6             0.3–4.2
Home Care .............................................................................................................................         9.7–164           1.0–16.4             0.1–1.3
   a Risk estimates reflect excess risk due to occupational exposure and are expressed per 1,000 employees at risk of TB infection.
   b Estimates  of death caused by TB due to occupational exposure are derived based on an estimated TB case death rate of 77.85 per 1,000
cases and are estimated by multiplying the lifetime active disease rate by .07785.
  c The ranges of risk presented in this TABLE reflect expected variance in the annual background TB infection rate by state. They are estimated
based on the assumption that the annual background TB infection rate ranges from 0.194 to 3.542 per 1,000 employees at risk.
  d Lifetime infection rate is estimated by (1–(1–p)45), where p is the annual excess TB infection rate due to occupational exposure.
  e Lifetime active disease rate is estimated to be 10% of lifetime infection rate.
                            Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules                                                                                    54207

  As outlined in the Health Effects                                          reported deaths caused by TB from table                                    cases in 1990 (Ex. 7–268). In this study,
section, several possible outcomes are                                       8–5 of the Vital Statistics for the U.S.                                   all confirmed TB cases reported in the
possible following an infection.                                             and cases of TB reported in CDC’s TB                                       county in 1990 were tracked and the
Approximately 90% of all infections                                          Surveillance system for 1989 through                                       number of deaths where TB was the
never progress to active disease. An                                         1991 (Exs. 7–270, 7–264). As shown in                                      direct or contributing cause was
estimated 10% of infections is expected                                      table V–13, the TB case death rate                                         ascertained. ‘‘Contributing cause’’ was
to progress to active disease; most of                                       ranged from 69.94 to 89.18 per 1,000                                       defined as a case of TB of such severity
these cases are successfully treated.                                        with a 3-year average of 77.85 per 1,000                                   that it would have caused the death of
However, a percentage of active TB                                           TB cases. The Agency used the 3-year                                       the patient had the primary illness not
cases develop further complications.                                         average (77.85 per 1,000) for its estimate                                 caused death earlier. Of the 1,724 cases
Approximately 7.8% of active TB cases                                        of deaths caused by TB. This estimate is                                   included in the study, TB was
may take a more severe clinical course                                       in close agreement with published                                          considered the cause of death or the
and lead to death. The TB case fatality                                      results from a retrospective cohort study                                  contributing cause of death in 135 cases
rate was estimated using information on                                      conducted in Los Angeles County on TB                                      (78.31 per 1,000).

                                                          TABLE V–13.—TB CASE DEATH RATES FOR ADULTS (18+)
                                                                                                                                                        Number of       Number of TB      TB case death
                                                                     Year                                                                                deaths a         cases b             rate c

1991 ..............................................................................................................................................            1,700            24,307            69.94
1990 ..............................................................................................................................................            1,796            23,795            75.48
1989 ..............................................................................................................................................            1,956            21,934            89.18
3-year Average .............................................................................................................................                   1,817            23,345            77.85
   a Source:    Vital Statistics for the U.S., Table 8–5, (age 20+).
   b Source:    CDC, TB surveillance system, (age 18+).
   c Rate    expressed per 1,000 TB cases. Any deaths caused by TB in persons 18 or 19 years of age are not included in the numerator.

   National estimates of annual and                                          disease and death caused by TB due to                                      weighted averages of the state estimates
lifetime risk for TB infection, active                                       occupational exposure are computed as                                      and are presented in table V–14.

                           TABLE V–14.—AVERAGE OCCUPATIONAL RISK ESTIMATES                                                               a, b   PER    1,000 WORKERS AT RISK
                                                                                                                           Annual TB in-              Lifetime TB in-   Lifetime active   Death caused
                                                 Work setting                                                                 fection                      fection             TB           by TB c

    WA ...........................................................................................................                         0.68                   30               3.0              0.2
    NC ...........................................................................................................                         5.7                   219              22.0              1.7
    JM ............................................................................................................                       11.8                   386              38.6              3.0
Long-term Care ..............................................................................................                             14.6                   448              44.8              3.5
Home Health Care ..........................................................................................                                6.9                   225              25.5              2.0
Home Care .....................................................................................................                            1.6                    69               6.9              0.5
   a Weightedby each state’s population in 1994.
        estimates reflect excess risk due to occupational exposure and are expressed per 1,000 employees at risk.
   b Risk
   c Number of deaths caused by TB due to occupational exposure are derived based on an estimated TB case death rate of 77.85 per 1,000
cases and are computed by multiplying the lifetime active disease rate by .07785.

  (a) Risk Estimates for Hospital                                            expected overall risk due to                                               estimates of the number of active TB
Employees: Logistic regression analysis                                      occupational exposure is estimated to be                                   case per 100 exposed workers are
of the Washington state hospital data                                        4 times the background rate. This results                                  expected to range between 7 and 72,
indicated an increase in annual risk                                         in an expected range of lifetime risk                                      resulting in as many as 6 deaths per
(47% above background) for employees                                         between 34 and 472 infections per 1,000                                    1,000 exposed employees at risk for TB
with potential exposure to TB. For this                                      employees at risk for TB infection.                                        infection.
particular analysis the control group                                        Lifetime estimates of active TB cases                                         In summary, table V–9 and table V–
was defined as those hospitals with no-                                      resulting from these infections are                                        14 show that the annual occupational
known TB patients that are located in                                        expected to range between 3 and 47,                                        risk of infection is expected to range:
counties that did not report any active                                      resulting in as many as 4 deaths per                                          (a) From .09 to 1.66 with a weighted
TB cases in 1994. However, an                                                1,000 exposed employees at risk of TB                                      average of 0.68 per 1,000 for workplaces
increased risk of 47% above background                                       infection. As done previously, the North                                   located in relatively low TB prevalence
in the annual infection rate is expected                                     Carolina study results were adjusted to                                    areas, and where TB infection measures
to produce a range of 4 to 72 TB                                             reflect 1994 TB disease trends.                                            and engineering controls are required;
infections per 1000 exposed workers in                                          Based on the data from Jackson                                             (b) From 0.77 to 14.1 with a weighted
a working lifetime, which could result                                       Memorial Hospital, the overall risk due                                    average of 5.7 per 1,000 for workplaces
in as many as 7 cases of active TB and                                       to occupational exposure is estimated to                                   located in areas with moderate TB
approximately 1 death per 1,000                                              be 8 times the background rate. This                                       prevalence and inadequate TB control
exposed workers.                                                             results in an expected range of lifetime                                   programs; and
  Based on the survey of hospitals in                                        risk between 67 and 723 infections per                                        (c) From 1.54 to 28 with a weighted
North Carolina’s western region, the                                         1,000 employees at risk. Lifetime                                          average of 11.8 per 1,000 for workplaces
54208           Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules

located in high TB prevalence areas,         were still being performed in certain        of lifetime risk of between 85 and 800
serving high risk patients, with high        areas of the hospital without adequate       infections per 1,000 employees at risk
frequency of exposure to infectious TB.      engineering controls, such as the            for TB infection. Lifetime estimates of
   Similarly, the lifetime occupational      Special Immunology clinic where HIV–         the number of active TB cases resulting
risk is expected to range:                   TB patients received pentamidine             from these infections range from 9 to 81
   (a) From 4 to 72 with a weighted          treatments. Like the hospitals in the        and are projected to cause as many as
average of 30 per 1,000 for workplaces       North Carolina study, Jackson Memorial       6 deaths per 1,000 exposed employees
located in relatively low TB prevalence      represents a working environment that        at risk of TB infection. Similarly, the
areas, and where TB infection measures       serves a patient population known to         overall annual excess risk of TB
and engineering controls are required;       have high TB prevalence. In addition,        infection for workers in home health
   (b) From 34 to 472 with a weighted        Jackson Memorial only tested                 care is estimated to be approximately
average of 219 per 1,000 for workplaces      employees with patient contact in areas      500% above background. This results in
located in areas with moderate TB            where active TB had been detected.           an expected range of lifetime risk of
prevalence and inadequate TB control            (b) Risk Estimates for Workers in         between 41 and 536 infections per 1,000
programs; and                                Other Work Settings: In long-term care       employees at risk for TB infection.
   (c) From 67 to 723 with a weighted        facilities for the elderly there is also a   Lifetime estimates of the number of
average of 386 per 1,000 for workplaces      significantly increased likelihood that      active TB cases range from 4 to 54 per
located in high TB prevalence areas,         employees will encounter individuals         1,000, and are projected to cause as
serving high risk patients, with high        with infectious TB. Persons over the age     many as 4 deaths per 1,000 exposed
frequency of exposure to infectious TB.      of 65 constitute a large proportion of the   employees at risk of TB infection.
   Risk estimates derived from either        TB cases in the United States. In 1987,      Similarly, the overall annual excess risk
study (Washington State or North             CDC reported that persons aged 65 and        of TB infection for workers in home care
Carolina) represent an overall rate of       over accounted for 27% (6150) of the         is estimated to be approximately 100%
occupational risk, because both studies      reported cases of active TB in the U.S.,     above background. This results in an
include PPD skin testing results from        although they account for only 12% of        expected range of lifetime risk of
the entire hospital employee                 the U.S. population. Many of these           between 10 and 164 infections per 1,000
population, whereas the Jackson              individuals were infected in the past        employees at risk for TB infection.
Memorial study addresses the                 and advancing age and decreasing             Lifetime estimates of the number of
occupational risk to workers where           immunocompetence have caused them            active TB cases range from 1 to 16, and
exposure to infectious TB is highly          to develop active disease. In 1990 the       are expected to result in approximately
probable.                                    CDC estimated that approximately 10          1 death per 1,000 exposed employees at
   Although the exact compliance rate is     million people were infected with TB.        risk of TB infection.
not known, hospitals in Washington           As the U.S. population steadily ages,           Clearly, employees in all three groups
State have been required to implement        many of these latent infections may          (long-term care for the elderly, home
the CDC TB guidelines with respect to        progress to active disease. Because          health care, and home care) have higher
engineering controls (requiring isolation    elderly persons represent a large            risks than hospital employees in
rooms with negative pressure) and            proportion of the nation’s nursing home      Washington. This could be attributed, in
infection control measures (advocating       residents and because the elderly            part, to the lack of engineering controls
early patient identification, employee       represent a large proportion of the active   in these work settings. That respirators
training, respiratory protection, and PPD    cases of TB, there is an increased           may be used only intermittently may
testing).                                    likelihood that employees at long-term       also play a role. Although workers in
   Neither the facilities in North           care facilities for the elderly will         these three groups are encouraged by
Carolina nor Jackson Memorial had            encounter individuals with infectious        local health authorities to use
engineering controls fully implemented       TB.                                          respiratory protection while tending to
at the time these data were collected.          Similarly, there are other                a suspect TB patient, the actual rate of
Early identification of suspect TB           occupational settings that serve high-       respirator usage is difficult to ascertain.
patients has always been recommended         risk client populations and thus have an     A third factor that may contribute to
in North Carolina. However, engineering      increased likelihood of encountering         higher risk in these work settings is
controls in isolation rooms were either      individuals with infectious TB. For          delayed identification of suspect TB
not present or did not function properly     example, hospices, emergency medical         patients due to confounding symptoms
because of modifications in the physical     services, and home-health care services      presented by the individuals. For
structure of the building (i.e., isolation   provide services to client populations       example, many long-term care residents
rooms had been subdivided using              similar to those in hospitals and thus       exhibit symptoms of persistent coughing
partitions, air ducts had been re-           are likely to experience similar risks.      from decades of smoking. Consequently,
directed because of remodeling, etc.).          OSHA used information from the            an individual in long-term care with a
Tuberculin skin testing was very             1994 Washington state PPD skin testing       persistent cough may be infectious for
inconsistent and sporadic. In addition,      survey to estimate occupational risk for     several days before he or she is
employee training and use of respiratory     workers in long-term care, home health       identified as having suspected
protection were not emphasized.              care, and home care. Annual estimates        infectious TB.
   By 1991, Jackson Memorial had most        of excess risk for TB infection are
of the engineering controls in place in      presented in TABLE V–10 and lifetime         Qualitative Assessment of Risk for Other
the HIV ward (where the first outbreak       estimates for TB infection, active TB,       Occupational Settings
took place) and in selected areas with       and death caused by occupational TB            The quantitative estimates of the risk
high TB exposure, but not in the entire      are presented in TABLE V–12.                 of TB infection discussed above are
hospital. However, the staff training           Based on the Washington State data,       based primarily upon data from
program was still being developed and        the overall annual excess risk for TB        hospitals and selected other health care
respiratory protection was not always        infection is estimated to be 10-fold over    settings. Data from hospitals and certain
adequate. Although exposures had been        background for workers in long-term          health care settings were selected
greatly reduced, ‘‘high risk’’ procedures    care. This results in an expected range      because OSHA believes that these data
                Federal Register / Vol. 62, No. 201 / Friday, October 17, 1997 / Proposed Rules                                54209

represent the best information available     settings that have overcrowded                working in these facilities will
to the Agency for purposes of                conditions or poor ventilation will           encounter individuals with infectious
quantifying the occupational risks of TB     facilitate the transmission of TB. Thus,      TB. In addition, environmental factors
infection and disease. However, as           given that a case of infectious TB does       such as overcrowding and poor
discussed above, it is their exposure to     occur, the conditions at the work setting     ventilation facilitate the transmission of
aerosolized M. tuberculosis that places      itself may promote the transmission of        TB. Thus, given that a case of infectious
these workers at risk of infection and       disease to employees who share                TB does occur, the conditions in the
not factors unique to these particular       airspace with the individual(s) with          facility itself promote the transmission
kinds of health care activities. Thus,       infectious TB.                                of the disease to other inmates and
OSHA believes that the risk estimates           The second step in extrapolating the       employees in the facility who share
derived from hospitals and selected          quantitative risks is to identify the types   airspace.
other work settings can be used to           of work settings which have some or all          As discussed in the Health Effects
describe the potential range of risks for    of the risk factors outlined above. Once      section, a number of outbreak
other health care and other occupational     these work settings have been                 investigations (Exs. 6–5, 6–6) have
settings in which workers can                identified, OSHA believes that it is          shown that where there has been
reasonably anticipate frequent and           reasonable to assume that the                 exposure to aerosolized M. tuberculosis
substantial exposure to aerosolized M.       quantitative risk estimates calculated for    in correctional facilities, the failure to
tuberculosis.                                hospitals and other selected health care      promptly identify individuals with
                                             settings can be used to describe the risks    infectious TB and provide appropriate
   In order to extrapolate the
                                             in the identified work settings.              infection control measures has resulted
quantitative risk estimates calculated for
                                                                                           in employees being infected with TB.
hospital employees and other selected        Correctional Facilities
                                                                                           These studies demonstrate that, as in
health care settings, OSHA, as a first         Employees in correctional facilities or     hospitals or health care settings, where
step, identified risk factors that place     other facilities that house inmates or        there is exposure to aerosolized TB
employees at risk of exposure. Some          detainees have an increased likelihood        bacilli and where effective control
amount of exposure to TB could occur         of frequent exposure to individuals with      measures are not implemented, exposed
in any workplace in the United States.       infectious TB. Many correctional              employees are at risk of infection. Thus,
TB is an infectious disease that occurs      facilities have a higher incidence of TB      estimates based on the risk observed
in the community and thus, individuals       cases in comparison to the incidence in       among employees in hospitals and in
may bring the disease into their own         the general population. In 1985, the          selected other work settings that involve
workplace or to other businesses or          CDC estimated that the incidence of TB        an increased likelihood of exposure can
work settings that they may visit.           among inmates of correctional facilities      be appropriately applied to employees
However, there are particular kinds of       was more than three times higher than         in correctional facilities.
work settings where risk factors are         that for nonincarcerated adults aged 15–         Recently, scientists at NIOSH have
present that substantially increase the      64 (Ex. 3–33). In particular, in states       completed a prospective study of the
likelihood that employees will be            such as New Jersey, New York, and             incidence of TB infection among New
frequently exposed to aerosolized M.         California, the increased incidence of        York State correctional facilities
tuberculosis. First among these factors is   annual TB cases in correctional facilities    employees (Ex. 7–288). This study is the
the increased likelihood of exposure to      ranged from 6 to 11 times greater than        first prospective study of TB infection
individuals with active, infectious TB.      that of the general population for their      among employees in correctional
Individuals who are infected with TB         respective states (Exs. 7–80 and 3–33).       facilities in an entire state. Other studies
have a higher risk of developing active      A major factor in the increased               have reported on contact investigations,
TB if they are (1) immunocompromised         incidence of TB cases in correctional         which seek to identify recent close
(e.g., elderly, undergoing chemotherapy,     facilities is the fact that the population    contacts with an index case and
HIV positive), (2) intravenous drug          of correctional facilities is over-           determine who might subsequently have
users, or (3) medically underserved and      represented by individuals who are at         been infected. Studies based on contact
of generally poor health status (Exs. 6–     greater risk of developing active disease,    investigations have the advantage of a
93 and 7–50). Thus, in work settings in      e.g., persons from poor and minority          good definition of potential for exposure
which the client population is               groups who may suffer from poor               and they serve to identify infected
composed of a high proportion of             nutritional status and poor health care,      persons for public health purposes. On
individuals who are infected with TB,        intravenous drug users, and persons           the other hand, prospective studies of
are immunocompromised, are                   infected with HIV. Similarly, certain         an entire working group have the
intravenous drug users or are of poor        types of correctional facilities, such as     advantage of covering the entire
general health status, there is a greatly    holding facilities associated with the        population potentially at risk, of
increased likelihood that employees          Immigration and Naturalization Service,       considering all inmate cases
will routinely encounter individuals         may have inmates/detainees from               simultaneously as potential sources of
with infectious TB and be exposed to         countries with a high incidence of TB.        infection, and, most importantly, of
aerosolized M. tuberculosis. A second        For foreign-born persons arriving in the      permitting the calculation of incidence
factor that places employees at high risk    U.S., the case rate of TB in 1989 was         rates and risk attributable to
of exposure to aerosolized M.                estimated to be 124 per 100,000,              occupational exposure.
tuberculosis is the performance of high-     compared to an overall TB case rate of           Following an outbreak of active TB
hazard procedures, i.e., procedures          9.5 per 100,000 for the U.S. (Ex. 6–26).      among inmates that resulted in
performed on individuals with                Moreover, in the period from 1986 to          transmission to employees in 1991, the
suspected or confirmed infectious TB         1989, 22% of all reported cases of TB         state of New York instituted a
where there is a high likelihood of the      disease occurred in the foreign-born          mandatory annual tuberculin skin
generation of droplet nuclei. A third        population. Given the increased               testing program to detect TB infection
factor that places employees at risk of      prevalence of individuals at risk for         among employees. The authors used
exposure is the environmental                developing active TB, there is an             data from the first two years of testing
conditions at the work setting. Work         increased likelihood that employees           to estimate the incidence of TB infection

Shared By: