This report presents the background and some preliminary findings during
1961 in a large-scale study of medical x-ray procedures in the offices
of private physicians in New York City.

M. B. Heller, M.S.; H. Bl-ztz, E.E.; B. Pasternack, Ph.D.; and M. Eisenbud, Sc.D.

THE benefits to mankind from med-                   lation than all other sources combined.
    ical x-ray diagnostic procedures in             (It does, without question, constitute
the past few decades have been sub-                 most of the total radiation received by
stantial and the practice of modern                 those individuals who are exposed.)
medicine is highly dependent on their                  It is quite clear that a large source
use. However, there are potential haz-              of the variation in dose estimates among
ards associated with these procedures               countries cited in the United Nations
due to the biological effects of ionizing           Report is a result of marked differences
radiation. In recent years, increased in-           in diagnostic x-ray technic. Most of the
terest in the magnitude of the popula-              references pertaining to dose estimates
tion's exposure to x-ray has developed              in the 1958 UN document were based
as part of the greater attention being              on research performed in large, modern
shown to all sources of exposure to ioniz-          hospitals or clinics in the respective
ing radiation, both artificial and natural.         countries. For many of the examinations,
   The principal sources of ionizing ra-            dosimetric data for only a limited num-
diation to which the population is ex-              ber of patients are presented. This num-
posed are the natural radioactive back-             ber of examinations may be sufficient
ground, occupational exposure, and the              for a valid representation in one hos-
clinical uses of x-rays and isotopes. The           pital or one clinic, but certainly con-
reports of the Federal Radiation Coun-              stitutes an inadequate representation of
cil' and the 1958 report of the United              an entire community whether city, state,
Nations Scientific Committee on the Ef-             or nation. Similarly, the recent esti-
fects of Atomic Radiation,2 the most                mates of gonadal dose included in the
authoritative compendia on the subject,             1962 UNSCEAR Report3 are again de-
present the estimates of these compo-               rived from data obtained from studies
nents shown in Table 1. This table in-              relating to large hospitals and medical
dicates two conditions which are sub-               centers mostly located in countries where
stantiated by reading most of the re-               medical care programs are either na-
maining literature on the subject: (1)              tionalized or strongly controlled. Esti-
There is more uncertainty regarding med-            mates derived from such data cannot be
ical radiation dose than any other com-             considered to constitute a reliable as-
ponent of population exposure, and (2)              sessment of the patient dose in nations
diagnostic exposure may account for                 where large proportions of medical diag-
more human radiation dose to the popu-              nostic x-ray examinations are performed

SEPTEMBER. 19964                                                                        1 551
                Table 1-Components of Population Radiation Exposure
                                                   Annual Dose (mrem)
                   Source                     Gonad Dosel     Marrow Dose2
               Cosmic rays                      32- 73             28
               Terrestrial radiation            25- 75             47
               Atmospheric radiation                 2              2
               K-40                                 19             11
               C-14                                  1.6            1.6
               Rn-Tn                                 2              2
               Ra                                   -               0.5
                        Total                   80-171             92

                                              Genetically      Per capita
                                              Significant        Mean
                                                 Dose         Marrow Dose
             Medical (exposure to patients)
               Diagnostic                      40-240       Ranges beyond 100
               Therapy                             12          No estimate
             Occupational                          <2              1-3

in private physicians' offices. Aside from     presenting dose estimates directly re-
this type of apparent biased estimation,       lated to the study by applying a set of
a large source of the variation among          laboratory or field data which was meas-
these nations can be ascribed, broadly, to     ured in an unrelated institution or area.
differences in diagnostic x-ray technic        Gonadal doses were then calculated in
among different institutions, specialties,     many cases without modification of the
and individuals.                               laboratory data with respect to physical
   Variation resulting from differences        and anthropometric factors except, per-
in diagnostic x-ray technic may be sub-        haps, for age and sex classification.
divided as follows: "(a) patient factors,      Clearly, there exists a need to study the
(b) procedural factors, (c) equipment         sources of variation which influence the
factors, and (d) professional factors."4      estimates of population radiation dose
Within these groups there are numerous        derived from medical x-ray procedures.
additional contributing factors such as       This report presents the background and
anthropometric relationships (including       some of the preliminary findings during
thickness of part), field location, field     1961 from our study of medical x-ray
size, peak kilovolts (KVP), filtration,       procedures in private physicians' offices
and individual variations in the per-         in New York City. The primary guide
formance of fluoroscopic examinations.        to the early objectives of the present
   Previous field studies relating to pa-     study was the ICRP/ICRU report en-
tient dose have been reported in this         titled "Exposure of Man to Ionizing
country.5'6 They generally present data       Radiation Arising from Medical Proce-
on examination frequency, but fail in         dures."7

1552                                                           VOL. 54, NO. 9. A.J.P.H.
                                                              MEDICAL X-RAY PROCEDURES

 Objectives                                        sociated with the physician's x-ray prac-
   The principal objectives and scope of          tice, viz.: type and manufacturer of unit,
the entire study may be summarized as             added filtration, dark adaptation time
follows:                                          for fluoroscopy, use of gonadal shields,
                                                  field size of radiographic cones, output
A. Collect, assemble, and furnish information     and half-value layer (HVL) at standard
   on the kind, operational state, and number
   of diagnostic and therapeutic x-ray units in   settings, and so forth.
   New York City.                                     The technical data for radiography
B. Estimate the frequency of various medical      and fluoroscopy, which are recorded on
   procedures which involve diagnostic and        the P-2 form, include: KVP, milli-
   therapeutic radiation exposure to residents    ampere-seconds (MaS), field size, name
   of New York City according to equipment
   used, technic, specialty of physician, age     of examination and anatomical empha-
   and sex of patients, and seasonal variation.   sis, projection, target to film distance,
C. Estimate per capita gonadal and bone mar-      film speed and size, and use of grids
   row radiation dose to the population of New    and screens. Information on the patient
   York City.
D. Determine the feasibility of using informa-    includes: age, sex, height, weight, race,
   tion obtained in objectives A, B, and C        borough of residence, and the thickness
   above in epidemiological studies. Identify     of part being examined.
   population exposure groups worthy of future
   study of possible delayed radiation effects.
                                                   Data Collection Method
Design and Methods                                   After careful consideration of all the
   The performance of the work entailed           methods proposed by the ICRP and UN
in the present study has been divided             Task Committees, the direct sampling of
into three phases. These, in sequence,            private physicians' offices was chosen.
are: (1) diagnostic private x-ray in-             The method selected permits measure-
stallations, (2) diagnostic institutional         ments of physical machine variables by
x-ray installations, and (3) therapeutic          trained personnel and allows the estab-
installations. The remainder of this re-          lishment of rapport between our field
port will concern itself primarily with           workers and the physicians' office staff.
the data collected during the early part          This rapport has contributed greatly to
of the survey of private medical offices.         the accurate reporting of data; particu-
It is to be understood that the private           larly when the personnel performing
patients exposed to units in these offices        routine radiography in private physi-
represent the population from which our           cians' offices are frequently not inti-
sample data has been assembled. Hos-              mately familiar with the technical vari-
pital outpatients, inpatients, and those          ables associated with x-ray technic.
exposed in clinics are considered in a               The forms used were designed in such
future phase of this study.                       a way as to permit the recording of all
                                                  pertinent data in the simplest possible
Required Data                                     way. The function of the P-1 form is the
                                                  collection of all information which is
   The data collected for each patient ap-        either constant or requires only one
pearing in this study are considered to           measurement. This is completed by the
be sufficient for the computation of dose         field worker during the initial interview.
by any of several methods. The tech-              The P-2 forms are designed for the re-
nical data are recorded by use of the             cording of examinations and variables
two forms designated as P-1 and P-2.              related to individual examinations and
The P-1 form is used in the initial inter-        patients. A further feature of the P-2
view to record general characteristics as-        form is that several examinations or ex-

SEPTEMBER, 1964                                                                        1 553
                                      Table 2-Main Study Sample
                                                                  Installations      Installation
                                  Total No. of     Total No. of      To Be            Sampling
              Specialty           Installations*   X-ray Units*     Sampled           Fraction
        GP                            2,401           3,696            302             0.1299
        Chest diseases                  220             351            114             0.518
        Dermatology                      17              23             11             0.649
        ENT                              47              59             28             0.595
        Gyn. and obst.                   56              80             28             0.50
       Internal med.                  1,098           1,879            222            0.202
       Neurology                          4               6              4             1.0
       Occupational med.                 99             208             50            0.505
       Ophthalmology                      4               5              4            1.0
       Orthopedics                      121             168             61            0.504
       Pediatrics                      205             261             104            0.507
       Radiology                       205             569             103            0.501
       Surgery                         211             336             106            0.501
       Urology                         134             170              67            0.50
       Gastroent.                         7              14              7            1.0
       Hematology                         2               2              2            1.0
       Allergy                            9               9              9            1.0
       Physical med.                      2               4              2            1.0
       Endocrinology                      2               2              2            1.0
       Mobile chest x-ray                 3               9              3            1.0
       Total                          4,847           7,851          1,229            0.2535
          *   As of March 15, 1961.

posures may be recorded on the same                      to all sample units during a stated pe-
page for one patient, but only one pa-                   riod of time. With regard to the se-
tient may be recorded per page.                          lected sample, patients who are under
   X-ray output, half-value layer, and                   the care of physicians other than the
field alignment are being obtained at in-                physician with whom the studied unit
stallations with an Electronic Instru-                   is associated are included, and patients
ments Limited ionization chamber and                    under the care of the physician being
electrometer of the type used in the                    studied who are referred elsewhere for
British study of medical exposure.8                     diagnostic x-ray procedures are ex-
                                                        cluded. Each installation is normally
The Sample Unit                                         sampled for a period of four calendar
   The sample unit, for the purpose of                     Names and addresses of patients are
this study, is defined as the x-ray equip-              not sought since the purpose of this
ment which belongs to or is registered                  study is the characterization of dose
in the name of a physician maintaining                  associated with specific diagnostic pro-
an office in New York City. All such                    cedures and an indication of exposure
installations comprise the universe from                groups according to type of examnination
which the sample is selected. Con-                      for possible future epidemiological study
ceptually, the exposed population of in-                rather than the identification of specific
dividuals consists of all patients exposed              heavily exposed individuals. Hence, it is

1554                                                                         VOL. 54, NO. 9. A.J.P.H.
                                                          MEDICAL X-RAY PROCEDURES

 Table 3-Sample Number of Patients and        ing criteria: (1) expected contribution
   Exposures, 1961                            to gonadal dose per examination per-
Radiography           No. of        No. of     formed, (2) expected patient workload,
   Procedure         Patients     Exposures    (3) expected variability in technic, (4)
                                              number of installations in the specialty,
Radiography only       7,167        19,276     and (5) size of installations within the
Radiography and                               specialty. Those specialties with less
  fluoroscopy          1,089        4,920
                                              than ten installations are being sampled
                       8,256       24,196     entirely. All others except general prac-
                                              tice, internal medicine, and dermatology
                                              are being sampled at 50 per cent. Gen-
Fluoroscopy           No. of        No. of    eral practice and internal medicine rep-
   Procedure         Patients      Sessions   resent the largest number of installations
Fluoroscopy only       1,281        1,645     and are therefore being sampled at only
Radiography and                               12 and 20 per cent respectively. Derma-
  fluoroscopy          1,089        1,096     tologists have an intermediate sampling
                                              fraction due to their expected low diag-
                      2,370         2,741     nostic x-ray workload. Deviations from
                                              the proposed sampling fractions result
                                              from the distribution of specialty instal-
emphasized that the identity of physi-        lation types within the boroughs.
cians and their patients is never released
from the study office and is known only       Organization of Field Work
by those personnel working directly on
the field and office staffs for the purpose      The boroughs were subdivided into
of communicating with the physician.          Area Work Zones (AWZ). Those AWZ's
Data are always presented in grouped          with a large number of installations
form within a given specialty.                were then further subdivided. The order
                                              of sampling the AWZ's was then selected
Sample Selection                              randomly. Each month the names of
                                              the physicians in the Area Work Zones
   The medical offices which are being        to be sampled are sent by our field super-
studied were chosen from the registry of      visor to the appropriate county med-
x-ray machines in New York City main-         ical societies. The executive secretaries,
tained by the Office of Radiation Con-        in turn, inform the physicians and urge
trol in the New York City Health De-          their cooperation by means of a letter.
partment. The total registration of diag-     Our field interviewers then call the
nostic medical installations arranged by      physicians for initial interview appoint-
specialty and number of units through         ments. The initial interview consists of
March, 1961, is shown in Table 2. This        providing detailed information to the
table also contains sampling fractions        physician and his staff on the objectives
and numbers of installations being            and motivations of this study, the col-
sampled. It is of interest to note that       lection of information for completion of
while 7.2 per cent of the x-ray units in      the P-1 form, and instructing a mem-
private offices are owned by radiologists,    ber of the physician's staff on the com-
71 per cent are in the hands of gen-          pletion of P-2 forms. Half-value layer,
eral practitioners and internists! A 25       output and beam alignment are meas-
per cent random sample, stratified by         ured subsequently by our physicist.
specialty and borough, was chosen after          The field interviewer makes a weekly
weighting each specialty by the follow-       follow-up visit at which time the P-2

SEPTEMBER, 1964                                                                    1 555
               CHEST EXAMINATIONS VA                                         22.7
                 EXPOSURES:                                                   Fm


                         TOTAL = 19,276
                        CHEST= 4,464
               % EXPOSURES:
                     < 30 YRS 20.4
                     < 40 YRS 37.4


Of) 2000-
                                                    7   24.8         22.0     7 22.6               21.3
                  1.9                     25.3                                                7/
                                                    /                                         /7
          0-       112.9_-    '730.0
                  0.9        10-19     20-29     30-39          40-49        50-59        >59
                                          AGE GROUP
Figure 1-Age Distribution. Chest Examinations           as a   Percentage of Total Exposures

forms are checked for completeness. The          computed. The general procedure con-
recorder is informed of proper proce-            sists of first computing the field size
dure, if necessary, and the forms are            of the x-ray beam at the patient from
brought back from the field for editing          the cone data and target to film distance.
and transcription to code sheets for key         The type of examination and anthropo-
punching. Preliminary pilot study field          metric information will then be used to
work was begun in February, 1961, and            determine the area and location of the
the main study field work commenced              beam projected on the patient. Given
in July, 1961.                                   this information, it will be determined
                                                 if the gonads are in the direct beam.
Data Processing                                  Since the output and HVL for each ma-
                                                 chine are measured at a standard KVP
   The study group has designed several          (with the amount of filtration normally
IBM cards to control and process the             used in place) the output and HVL at
information pertinent to the study and           any other KVP can be computed from
utilizes the following IBM equipment for
its routine work: keypunch, verifier,
sorter, and tabulator. More voluminous
and complex computations, such as long-          Table 4-Age Distribution of Persons
                                                   Exposed Radiographically
term statistical analyses and dose cal-
culations, are being programed for the                            UN Report          This Study
university's computer facilities.                  Age               %                   %

Dose Computation                                   0-14                 15              3.7
                                                  15-29                 20             16.8
  The gonadal dose to each patient ob-            30-49                 35             37.1
served in the study will be individually           >49                  30             42.4

1556                                                                 VOL. 54. NO. 9. A.J.P.H.
                                                               MEDICAL X-RAY PROCEDURES

  Table 5-Examination Distribution                  that year, 80 per cent of the 418 phy-
                                                    sicians contacted* cooperated with the
                                      Mean No.      study. The number of patients and ex-
                 Per cent               of Expo-
                  of Total             sures per   posures are presented in Table 3.
Radiography    Examinations          Examination      The age distribution in ten-year in-
                                                   tervals, for all exposures, is shown in
Chest               37                    1.28     Figure 1. Chest film exposures are in-
Lumbosacral                                        dicated as a percentage of the exposures
  spine              4.1                  2.47     in each age interval. The over-all per-
Upper GI             4.6                  7.50
Knee                 3.7                  4.01     centage of chest examinations is 37
                                                   per cent. The percentage of total exam-
Fluoroscopy                                        inations under ages 30 and 40 are of
Chest               65
                                                   interest, since it is estimated that 50
Upper GI            17                             per cent of all children are born to par-
Barium enema        11                             ents under 30 years of age and 90 per
                                                   cent are born to parents under 40 years
                                                   of age. The age distribution data in-
                                                   dicate an apparently lower percentage
standard data. The air dose at the pa-             of examinations in the 0-14 year age
tient's skin is computed as a function             group than that estimated by Laughlin4
of KVP and target to skin distance. The            in the UN/ICRP task report as shown
gonadal dose is then determined by di-             in Table 4. From this table it is seen
rect computation or by modification for            that the UN/ICRP estimates agree with
technical and anthropometric factors               our findings in the two middle ranges
and interpolation of a standard set of             and that the difference in the number
laboratory phantom data such as that               of exposures in the less than 15-year-age
developed at the Sloan-Kettering Insti-            group is made up in the greater than
tute.9 Bone marrow dose may be com-                49-year group.
puted in a similar manner.                            The most frequent examinations are
                                                   given in Table 5. The mean number of
Analysis                                           exposures for all radiographic examina-
                                                   tions was found to be 2.37. The most
  The data presented here are derived
                                                     * Excluding 95 who were unable to partici-
from the analysis of information col-              pate because of death or sale of equipment
lected during the year 1961. During                and/or practice.

                                   Table 6-Chest Radiography
                                      Field Size    Mean          Mean         Mean
              Specialty                  (cm)       MaS           KVP         TFD (in.)

      GP                                 96.15       17.3          70.5          70.2
      Chest                              87.46       16.0          70.3          69.7
      Internal medicine                  83.10       17.0          74.7          70.5
      Radiology                          74.78       21.1          71.3          69.6
      Occupational medicine*             62.58       28.0          78.0          65.9
      All spec.t                         77.1        17.6          71.8          71.1
         * Photoroentgen examinations included.
         t Excluding occupational medicine.

SEPTEMBER. 1964                                                                           I 557
 Table 7-Per cent of Total Chest Films                  Table 8 pertains to the field sizes
                                                     associated with radiographic chest exam-
                           Specialty                 inations. The relative size of field may,
 Projection     GP %      Chest % Radiology          perhaps, be better appreciated in remem-
                                                     bering that for the 14 by 17 inch film
    A.P.         17.3        5.3            5.3      the diagonal dimension is 22 inches or
    P.A.         68.8       86.5           67.8
    Lat.          3.7        7.0           26.0      56 centimeters. If the x-ray tube is
                                                     aligned with the film cassette, this is the
                                                     diameter of beam required to completely
                                                     cover the film. It is seen that among
 frequent type of examination involved               the installations within a given medical
 the chest. Table 6 summarizes the im-               specialty, there is relatively little varia-
portant technical variables in chest                 tion in field size. The magnitude of
 radiography for several specialties. The            field sizes, associated with various spe-
mean MaS for all specialties is 17.6. Our            cialties, are presented more clearly in
analyses show that this variable, which              Figure 2. The effects on dose of the
is proportional to dose, would be re-                large number of anterior-posterior pro-
duced by 15 per cent if appropriate in-              jections coupled with the field size used
tensifying screens were used for all                 by general practitioners can be clearly
examinations. The 28 MaS figure, the                 appreciated.
high KVP and low target to film dis-
tance for occupational medicine, reflect             Comments
the extensive use of photofluorographic
equipment in that specialty. The high                   It is hoped that the information to be
MaS value for radiology is due to the                gained from the present study will lead
relatively larger proportion of lateral              to a more realistic estimation of popula-
projections performed in that specialty              tion radiation exposure from medical
as indicated in Table 7. The higher                  x-rays than has heretofore been avail-
percentage of anterior-posterior projec-             able in the United States. Our data,
tions by general practitioners will tend             through the end of 1961, indicate that
to increase the gonadal dose contributed             many of the accepted estimatesI1012 (on
by this examination, since in females                which population dose estimates have
the iliac bones are not interposed be-               been based) may be low by as much
tween the target and the ovaries and in              as one or two orders of magnitude. These
males the muscles which would be inter-              estimates are misleading because they
posed in the posterior-anterior projec-              are based on the use of optimum tech-
tion are not interposed.                             nic and therefore indicate what the dose

               Table 8-Summary of Field Size Analysis for Chest Radiography

                                                    Mean        Standard     Coefficient
                               Number of          Field Size Error of Mean of Variation
              Specialty       Examinations           (cm)         (cm)     of Mean (%)
        GP                           274            96.15           8.00         8.33
        Chest                        649            87.46           5.88         6.72
        Internal med.                893            83.10           9.77        11.76
        Radiology                    512            74.78           2.59         3.46
        Occupational med.          1,849            62.58           5.13         8.12

I 558                                                                  VOL. 54, NO. 9. A.J.P.H.
                                                                         MEDICAL X-RAY PROCEDURES

INT. MED. 33.0"                 G.P.   38.0"               to the gonadal dose from chest examina-
                                  CFHEST       34.5"       tions-the most frequent radiographic
                                                           procedure and the one expected to be
                                                           most standardized.
                                       )                   REFERENCES
                                                           1. Background Material for the Development of Radia.
                                                              tion Protection Standards. Report No. 1. Washing-
                                                              ton, D. C.: Federal Radiation Council (May), 1960.
                                                           2. Report of the United Nations Scientific Committee
RADIOL. 29.5                    OCC I MED. 24.6"1             on the Effects of Atomic Radiation. 13th Session
                                                              Suppl. No. 17 (A/3838), New York, 1958.
                                                           3.              . 17th Session Suppl. No. 16 (A/5216),
                                                              New York, 1962.
                                                           4. "Recommendations for Plan of Study of Task I
                                                              Through V Within the United States," under
                                                              UNSCEAR, ICRP. In Hearings, 85th Session, United
                                                              States Congressional Joint Committee on Atomic
                                                              Energy, Special Subcommittee on Radiation, on the
                                                              Nature of Radiation Fallout and Its Effect on Man,
                                                              Part 1 (June), 1957, pp. 859-888.
                                                           5. Brown, R. F.; Heslep, J.; and Eads, W. Radiology
        HEIGHT 5' 8"   14" x 17" FILIM                        74:353 (Mar.), 1960.
                                                           6. Billings, M. S.; Norman, A.; and Greenfield, M. A.
Figure 2-Field Sizes for ChesiIt Examia-                      Ibid. 69:37-51 (July), 1957.
  tion                                                     7. "Exposure of Man to Ionizing Radiation Arising
                                                              from Medical Procedures." ICRP/ICRU Phys. in
                                                              Med. and Biol. 2:108, 1957.
                                                           8. Radiological Hazards to Patients. Second Report of
                                                              the Committee to the Ministry of Health, Depart-
should be rather than what tlhe dose is.                      ment of Health for Scotland, London, 1960.
                                                              Epp, E. R.; Weiss, H.; and Laughlin, J. S. Brit.
The true dose to the populati on is tem-                    m J. Radiol. 34:85100 (Feb.), 1961.
pered by technics which exiiibit great                     10. Webster, E. W., and Merrill, 0. E. New England
                                                               J. Med. 257:811, 1957.
variability. Our data, thus far, have                      11. Ardran, G. M., and Crooks, H. E. Brit. J. Radiol.
demonstrated the large variiability of                        30:295, 1957.
                                                           12. Wolfson, J. L., and Garrett, C. National Research
some technical factors relatinig directly              I
                                                               Council of Canada, No. 5377, 1959.

           The authors are associated with the Environmental Radiation Laboratory
         of the Institute of Environmental Medicine, New York University Medical
         Center, New York, N. Y.
           This study was performed under U. S. Public Health Service Contract No.
         SAph 73722.
           This report is a revision of a paper read at the annual meeting of the Health
         Physics Society, June 11-14, 1962, in Chicago, Ill.

SEPTEMBER. 1964                                                                                            155S9

To top