Acute Exacerbations of Chronic Bronchitis

Document Sample
Acute Exacerbations of Chronic Bronchitis Powered By Docstoc
					Acute Exacerbations of Chronic
An International Comparison
Peter Ball, MD; and Barry Make, MD, FCCP

        The prevalence of chronic bronchitis is between 3% and 17% in most developed countries.
        However, higher rates in the range of 13 to 27% are encountered in less developed areas of the
        world. Acute exacerbations of chronic bronchitis (AECB) have usually been defined as the
        presence of increases in cough/sputum, sputum purulence, and dyspnea. However, recent
        investigations suggest that the severity of AECB may be divided into three stages based on the
        history of the patient: (1) previously healthy individuals; (2) patients with chronic cough and
        sputum and infrequent exacerbations; and (3) persons with frequent exacerbations or more
        severe chronic airflow limitation. Therapy for patients with less severe AECB include older and
        less expensive broad-spectrum antibiotics, while newer agents are indicated for patients with the
        most severe stage of AECB.                                         (CHEST 1998; 113:199S-204S)

A lthough socioeconomic status, geography, eth-
          strikingly disparate in
                                  and other fac-
                                                                dom in the early 1960s—17% of adult men9 or
                                                                more recently—16.4% in elderly persons.10
tors, the United States, Europe, Latin America,                    Specific factors may affect the prevalence in cer-
and the Asia-Pacific region have a remarkably                   tain countries in the southern hemisphere. Exposure
similar incidence of chronic bronchitis (CB). Re-               to firewood or other biomass smoke during cooking
cent studies in the United States suggest a preva-              may occur in up to 50% of the world’s households,11
lence of 5.1 to 5.4% in the middle-aged to elderly              and woodsmoke in particular is thought to be re-
population, with a lower prevalence in nonsmok-                 sponsible for almost 50% of all cases of obstructive
ers.1 Prevalence in Europe varies from 3.7% in                  airways disease. Wood and straw are used for cook-
Denmark,2 4.5% in Norway,3 6% and 6.4% in                       ing with inadequate or absent ventilation. In Nepal,
Barcelona and Valencia, Spain, respectively,4,5 and             Pandey12 commented on the very high 18% preva-
6.7% in Sweden.6 Data are more difficult to obtain              lence of CB in the region and its similar frequency in
in the Asia-Pacific area but appear to be in the                both men and women. Recently, a multivariate
same range. For example, Lai and coworkers7                     analysis of a case control study in Colombia13 found
found 6.8% of an elderly group of Chinese living in             woodsmoke to be more highly associated (odds ratio,
Hong Kong to be affected, and data obtained via                 3.43) with development of COPD in women than
pharmaceutical industry research indicate the in-               either tobacco use or passive smoking (odds ratios,
cidence, 6.9%, to be similar in Taiwan (Chan;                   2.22 and 2.05, respectively).
1997; personal communication). Figures from In-                    However, tobacco smoking is undoubtedly the
donesia are higher but do not differentiate acute               most common cause of CB. Estimates of smoking
bronchitis from CB or account for multiple epi-                 prevalence range from 23% of Filipinos (Reuter’s
sodes during the observation period. The age                    News Service, June 24, 1995) and 26.7% in Guate-
profile in an Indonesian sample of nearly 250,000               mala—male to female ratio, 2:1,5,14 to 62% of adult
patients noted the incidence in those aged 30 to 39             men in Korea.15
years to be 15.7%; for 40 to 54 years it was 19.3%;
and for     55 years it was approximately 6%.8
Nevertheless, the higher figures may not be so                  Pathogens Associated With Exacerbations
different from those reported in the United King-                       of Chronic Bronchitis

*From the Infectious Diseases Department (Dr. Ball), Victoria      It is estimated that 50 to 75% of infective exacerba-
 Hospital, Kirkcaldy, Fife, United Kingdom; and the Emphysema   tions are bacterial in origin. Studies from the northern
 and Pulmonary Rehabilitation Programs (Dr. Make), National     hemisphere consistently identify Haemophilus influen-
 Jewish Medical and Research Center, and the Division of
 Pulmonary Sciences and Critical Care Medicine, University of   zae as the major pathogen, with Moraxella catarrhalis
 Colorado School of Medicine, Denver.                           second in importance (Table 1).

                                                                          CHEST / 113 / 3 / MARCH, 1998 SUPPLEMENT   199S
       Table 1—Prevalence of Common Respiratory                      those with mild-to-moderate exacerbations manage-
                  Pathogens in AECB                                  able on an outpatient (home-care) basis; (2) patients
                               H             M             S         fulfilling the criteria for a severe acute exacerbation;
                          influenzae,   catarrhalis,   pneumoniae,   and (3) those requiring intensive care. However,
        Source, yr             %             %             %         these guidelines are empirically based and severity
Sputum isolates                                                      criteria are largely those of supervening parenchymal
  Mulder et al,28 1952      50.0                                     infection or respiratory failure. Further, the role of
  Lees and McNaught,29      54.0           Not            32.0       antibiotics in general received no more than scant
    1959                                identified*                  attention and choices between classes of antibiotic
  Gump et al,30 1976        57.0           Not            27.0
                                                                     received no attention.
  Davies et al,31 1986      58.5            15            16.5          Canadian recommendations on management of
  Chodosh,32 1991           22.5            14            10.0       CB22 provide more detail on AECB, the role of
  Lindsay et al,33 1992     50.0            19            17.0       antibiotic therapy, stratification of patients into risk
  Ball,34 1994              52.0            13            16.5       (severity) groups in addition to the classic Winnipeg
  Sportel et al,35 1995     35.0            42            10.0
PSB† isolates
                                                                     symptom criteria,18 and empirically anticipated at
  Fagon et al,36 1990       39.0‡            7            16.0       least some of the factors later shown to correlate with
  Monso et al,37 1995       58.0            12            12.0       poorer response to therapy. These include age 65
*M catarrhalis not considered pathogenic until 1970s.                years, significant comorbid illness, FEV1 50% of
  PSB protected specimen brush specimen taken at bronchoscopy.       predicted, and number of exacerbations per year.
  Including Haemophilus parainfluenzae.                              However, the resultant five-stage severity classifica-
                                                                     tion based on these numerous variables, although
                                                                     recommending a stepwise intensification of antibi-
                      Classification                                 otic potency for disease of increasing severity, ap-
                                                                     pears unlikely to find favor with busy practitioners
   Standard definitions of CB refer to persistent                    looking for a simple rule of thumb.
cough and sputum over 2 years, but not to airflow                       In parallel to these empiric recommendations, a
limitation, exacerbations, or the progressive nature                 number of studies have identified criteria that iden-
of the associated respiratory disability. Nevertheless,              tify at-risk patients and that thus allow a meaningful
the early but definitive observations of Oswald and                  severity classification. Unlike community-acquired
colleagues16 found 70% of patients had broncho-                      pneumonia, in which the outcome variable is usually
spasm, 88% had either spasmodic or constant                          death, these studies have used return to the practi-
breathlessness, and “spells of sickness for several                  tioner after therapy with no relief from or relapse of
weeks or a few months” with infection playing a part                 the original referring symptoms. The Winnipeg type
in all cases. The British Medical Research Council                   1 clinical criteria18 define a significant exacerbation
suggested the value of a classification for both clin-               but not its severity, which is predicted (in United
ical and epidemiologic purposes,9 the need for which                 Kingdom patients) by the numbers of exacerbations
was reiterated by Fisher and colleagues17 in 1969.                   in the previous years, 4 or more having a high odds
However, such an approach was largely ignored until                  ratio for failure or relapse and significant comorbid-
the landmark study by Anthonisen and coworkers18                     ity.23 Although length of history was not predictive in
that defined exacerbations as an increase in cough/                  the latter report, a Canadian group has recently
sputum, sputum purulence, and dyspnea and pro-                       demonstrated a 10-year history of CB was associated
vided proof of the short-term efficacy of antibiotic                 with a poor outcome in AECB (Ronald F. Grossman,
therapy. Although frequently cited as providing                      MD; 1997; personal communication). It is possible
guidelines for severity assessment, this study made                  on these bases to propose a three-level staging of
no such claims, although an empiric severity scoring                 exacerbations (Table 2). This type of classification
system largely based on the Winnipeg criteria was                    may allow a more rational choice of antibiotic on the
used by Allegra and colleagues19 in a further study                  basis of previous therapy, local prevalence of resis-
with similar conclusions. However, in the late 1990s,                tance, and other factors. The clarification in Table 2
there remain no generally accepted systems for                       has been accepted recently as the working basis for
classifying exacerbations of CB.                                     expert recommendations for antibiotic therapy for
   Specific guidelines for the evaluation of new anti-               AECB by large consensus groups in the Asia-Pacific
biotics in acute exacerbations of CB (AECB), pro-                    region and Latin America.
duced by the Infectious Diseases Society of America,                    Analysis of outcome of treatment of exacerbations
incorporated standard criteria for diagnosis but none                of CB and of the disease itself clearly requires more
for severity.20 The European Respiratory Society                     than assessments of clinical and bacteriologic re-
Guidelines21 described three levels of severity: (1)                 sponse. In addition to symptom complexes, out-

200S                                                                                     Disease Management of Pulmonary Infections
Table 2—Proposed Classification of Severity of AECB*             influenzae varies dramatically between countries.
Severity       Background Status       Exacerbation Definition
                                                                 Thus figures of 2.5% (Ecuador) and 10% (Colombia,
                                                                 Uruguay) are lower than the 25 to 30% in the general
Stage 1    Simple mucus               Acute tracheobronchitis    population of Argentina and Venezuela and increas-
             hypersecretion              in previously healthy
                                                                 ing to almost 50% in hospital isolates in Guatemala.
Stage 2    Simple CB                  Acute increase in             Clearly such resistance levels compromise therapy
           (2-3 yr history of cough   (a) dyspnea,               with the basic penicillins and some oral cephalospo-
              and sputum for 2-3      (b) sputum volume,         rins. Decisions by consensus groups incorporate
              mo/yr)                  (c) sputum purulence       these factors in their recommendations. Thus, such
Stage 3    Complicated CB             As in stage 2 plus
                                      (a) 4 AECB in previous
                                                                 groups in both Latin American and Asia-Pacific
                                         year,                   regions accept the use of amoxicillin for patients with
                                      (b) comorbidity,           few risk factors (stage 1 and stage 2) and a high
                                      (c) 10-yr history of CB    probability of spontaneous recovery, although cer-
*Modified from Ball and Wilson.38                                tain countries prefer to offer quinolones, macrolides,
                                                                 or tetracyclines as alternatives. For stage 3 patients,
                                                                 there is a significant consistency to the recommen-
comes should question the benefits to the patient’s              dations with alternative agents proposed being
lifestyle, activities, and sense of well-being. Such             quinolones (first choice in Latin America, first equal
“quality of life” assessment can be defined via ques-            in Southeast Asia), amoxicillin/clavulanic acid,
tionnaires that measure various aspects of the pa-               azithromycin, and the more active second- or third-
tients’ perspective of health24 and should be incor-             generation cephalosporins, eg, cefuroxime axetil and
porated into clinical trials of newer agents to obtain           cefixime.
valid comparisons of cost benefit.
                                                                    Management Guidelines in Developed
 Antibiotic Choices: Developing Countries
                                                                    Three sets of recommendations that address the
   The criteria used to make rational choices for
                                                                 management of AECB have recently been published
antibiotic use in AECB are not different from those
                                                                 in more highly developed communities.21,22,26 Two
in the affluent and more developed western coun-
                                                                 of these guidelines, one from the United States26 and
tries. They include local frequencies of pathogen
                                                                 the other from Europe,21 offer a comprehensive
resistance and the availability of registered agents,
                                                                 approach to patients with COPD. Since these guide-
which may be dramatically different in contiguous
                                                                 lines are very comprehensive and cover many differ-
countries. However, financial influences often inval-
                                                                 ent aspects of the management of COPD, they do
idate scientific imperatives, and patients may choose
                                                                 not provide any in-depth approach to the use of
to prioritize differently from physicians. In India,
                                                                 antibiotics during exacerbations and incorporate only
casual and daily wage earners were reported to
                                                                 a general outline of therapy. Only the Canadian
exhibit a paradoxical preference for expensive but
                                                                 document is focused solely on AECB.22 As outlined
more effective quinolone therapy. The higher costs
                                                                 in Table 3, these three guidelines each address four
of quinolone therapy were offset by reduction in loss
                                                                 important clinical issues in patient management: (1)
of earnings and prospect of continued employment
                                                                 sputum evaluation; (2) likely responsible bacteria; (3)
with a “sickness free” record.25
                                                                 when antibiotics should be prescribed; and (4) pri-
   Resistance rates to penicillin and amoxicillin
                                                                 mary and alternate antibiotic choices.
among the principal pathogens in the Asia-Pacific
region vary widely, ranging from 15% for H influen-
                                                                 Sputum Culture
zae in Malaysia and Korea to 30 to 40% in Singapore,
Indonesia, Thailand, Taiwan, Hong Kong, and the                     In outpatients with bronchitic exacerbations, the
Philippines. Pneumococcal resistance tends to be                 guidelines uniformly suggest that Gram’s stain and
lower in Singapore and Malaysia (10 to 20%) than                 culture of expectorated sputum are not cost-effective
elsewhere, where overall figures of 30 to 40% in                 and therefore not necessary; thus, antibiotic therapy
Hong Kong, Korea, and Taiwan are overshadowed                    should be empiric. However, sputum culture may be
by hospital figures of 70% in Manila, Philippines.               helpful in selected situations. The European guide-
   Prevalence of pneumococcal resistance is lower in             lines suggest culture of the sputum to guide therapy
Latin America in general than in southeast Asia but              when response to initial therapy is suboptimal. The
figures reach 15 to 25% in many areas. In Latin                  US recommendations support sputum culture to
America, the frequency of amoxicillin resistance in H            direct therapy in more severe exacerbations, in the

                                                                           CHEST / 113 / 3 / MARCH, 1998 SUPPLEMENT   201S
                                 Table 3—International Guidelines for Management of AECB

          Topic                      Europe21                              United States26                              Canada22

Initial sputum analysis   Not necessary                      Not necessary                                  Only with FEV1 50% pred, age
                                                                                                                65 yr, or 4 exacerbations/yr
Sputum analysis in        When response to initial therapy   In severe exacerbations, recent antibiotics,   With lack of improvement or
  other cases              is poor                             nursing home residence, or hospital           deterioration
Responsible bacteria      Streptococcus,                     Streptococcus,                                 Streptococcus,
                            Haemophilus, Moraxella;            Haemophilus,                                   Haemophilus,
                            increasing Staphylococcus and      Moraxella                                      Moraxella; increasing resistance
When to treat             Sputum purulence in an             Change in sputum color or consistency in        2 of: increased cough/sputum,
                            exacerbation                       an exacerbation                               sputum purulence, dyspnea
Initial antibiotics       Amoxicillin, tetracycline          Tetracycline, doxycycline, amoxicillin,        Aminopenicillin, tetracycline,
                            derivatives,                       erythromycin, trimethoprim-                   trimethoprim-sulfamethoxazole
                            amoxicillin/clavulanic acid        sulfamethoxazole, cefaclor
Secondary antibiotics     Newer cephalosporin, macrolide,    Broad-spectrum penicillin, cephalosporin       Cephalosporin (second or third
                            quinolone                                                                         generation),
                                                                                                              amoxicillin/clavulanic acid,
                                                                                                              newer macrolide, quinolone

face of recent prior antibiotic administration, when                      symptoms develop. The US guidelines focus more on
the patient resides in a nursing home, and when the                       exacerbations, which though not defined are implied
patient is sufficiently ill to require hospital admis-                    to indicate increasing dyspnea. A change in sputum
sion; the Canadian guidelines similarly recognize the                     color or consistency during an outpatient exacerba-
importance of Gram’s stain and culture in a more                          tion, and in all cases in which patients are hospital-
complicated subset of patients.                                           ized for an exacerbation, are suggested indications
                                                                          for antibiotics in the United States. The Canadian
Causative Bacteria                                                        guidelines, following the results of Anthonisen et
                                                                          al,18 recommend antibiotics with two or more of the
   All three publications recognize the importance of                     symptoms of increased cough/sputum, sputum puru-
bacteria such as Streptococcus pneumoniae, H influ-                       lence, and dyspnea.
enzae, and M catarrhalis. Although the problem of
antibiotic resistance is recognized by the European                       Antibiotic Choices
and Canadian guidelines, this issue is not addressed
by the Americans. Discussion on the increasing                               Since most exacerbations of CB occur in patients
incidence of Staphylococcus as a cause of exacerba-                       who have not had multiple recurrences and who may
tions and increasing antibiotic resistance of Hae-                        not even have airflow limitation, inexpensive, older
mophilus and of Streptococcus are prominent in the                        antibiotics are preferred. The US guidelines suggest
European publication, and the Canadian guidelines                         a broad set of initial antibiotic choices, including
indicate the potential for resistance in more compli-                     tetracycline, doxycycline, amoxicillin, erythromycin,
cated patients including those with more than four                        trimethoprim-sulfamethoxazole, or cefaclor. Both
exacerbations per year and thus prior antibiotic                          the Europeans and Canadians agree that less expen-
therapy. However, both the European and US doc-                           sive antibiotics are sufficient and recommend amoxi-
uments suggest that knowledge of local bacterial                          cillin and tetracycline derivatives as initial antibiotics.
resistance patterns and local experience in antibiotic                    The Europeans also recommend amoxicillin/clavu-
efficacy are helpful in guiding initial antibiotic                        lanic acid as an initial choice.
selection.                                                                   European guidelines suggest alternate antibiotics,
                                                                          including newer cephalosporins, macrolides, and
When To Use Antibiotics                                                   quinolones. The US recommendations for alternate
                                                                          antibiotics are broad-spectrum penicillins and ceph-
   Although detailed guidelines for when to use                           alosporins. The Canadians recommend similar alter-
antibiotics are not provided, the Europeans indicate                      nate antibiotics not only for patients who have
that administration of antibiotics is appropriate for                     suboptimal initial response but also suggest these
patients with purulent sputum, and further suggest                        agents are appropriate as a first choice in patients
that patients may be given a prescription for antibi-                     with more complicated disease.
otics for early initiation when appropriate signs or                         The repetitive demonstration of the equivalence of

202S                                                                                               Disease Management of Pulmonary Infections
new agents, eg, the macroazalides and quinolones, to                13 Dennis RJ, Maldonado D, Norman S, et al. Woodsmoke
ampicillin-amoxicillin despite their clear microbio-                   exposure and risk for obstructive airways disease among
                                                                       women. Chest 1996; 109:115-19
logical and kinetic advantages, is generally not re-
                                                                    14 Report of Comision Nacional Contra el Tabaquismo. 1994
flected in the initial antibiotic choices. This is per-             15 Medical Insurance Statistical Yearbook. 1994
haps best reflected by a recent draft of the British                16 Oswald NC, Harold JT, Martin WJ. Clinical pattern of
Thoracic Society COPD Guidelines that appear to                        chronic bronchitis. Lancet 1953; 2:693-43
provide no severity assessment for exacerbations and                17 Fisher M, Akhtar A, Calder M, et al. Pilot study of factors
that suggest the use of amoxicillin or tetracycline                    associated with exacerbations in chronic bronchitis. BMJ
should increased dyspnea, sputum volume, and pu-                       1969; 4:187-92
                                                                    18 Anthonisen NR, Manfreda J, Warren CPW, et al. Antibiotic
rulence be present. This fails to recognize the prev-
                                                                       therapy in acute exacerbations of chronic obstructive pulmo-
alence of bacterial resistance in the United King-                     nary disease. Ann Intern Med 1987; 106:196-204
dom, even though lower than elsewhere, the                          19 Allegra L, Grassi C, Grossi E, et al. Ruolo degli antibiotici nel
marginal pharmacokinetics of such agents in respi-                     trattamento delle riacutizza della bronchite cronica. Ital
ratory tissues, and the frequency of therapeutic                       J Chest Dis 1991; 45:138-48
failure of such therapy in even mild exacerbations                  20 Chow AW, Hall CB, Klein JO, et al. Evaluation of new
managed at home.27                                                     anti-infective drugs for the treatment of respiratory tract
                                                                       infections. Clin Infect Dis 1992; 15(suppl 1):S62-88
   In summary, the published guidelines from the
                                                                    21 Siafakis NM, Vermeire P, Pride NB, et al. ERS consensus
more developed countries fail to consistently and                      statement: optimal assessment and management of chronic
clearly identify when to treat patients with antibiotics               obstructive pulmonary disease (COPD). Eur Respir J 1995;
and when to use alternate antibiotics in patients with                 9:1398-1420
more complicated disease. The Canadian recom-                       22 Balter MS, Ryland RH, Low DE, et al. Recommendations on
mendations present a provocative, though clinically                    the management of chronic bronchitis: a practical guide for
untested, classification of severity of the underlying                 Canadian physicians. Can Med Assoc J 1994; 151(10, suppl):
disorder that may prove a useful guide to treatment.                23 Ball P, Harris JM, Lowson D, et al. Acute infective exacer-
                                                                       bations of chronic bronchitis. Q J Med 1995; 88:61-68
                                                                    24 Hyland ME. Quality of life assessment in respiratory disease.
                                                                       PharmacoEconomics 1992; 2:43-53
                        References                                  25 Arya SC. Paradoxical preference of even costly fluroquinolo-
 1 Enright PL, Krumla RA, Higgins MW, et al. Prevalence and            nes by casual workers and daily wage earners in developing
   correlates of respiratory symptoms on the elderly. Chest            countries. Fifth International Symposium of New Quinolo-
   1994; 106:827-34                                                    nes. Singapore, 1994
 2 Lange P, Groth S, Nyboe J, et al. Chronic obstructive lung       26 Celli BR, Snider GL, Heffner J, et al. ATS statement:
   disease in Copenhagen: cross-sectional epidemiological as-          standards for the diagnosis and care of patients with chronic
   pects. J Intern Med 1989; 226:25-32                                 obstructive pulmonary disease. Am J Respir Crit Care Med
 3 Bakke PS, Baste V, Hanoa R, et al. Prevalence of obstructive        1995; 152(5, pt 2):S77-120
   lung disease in a general population: relation to occupational   27 Ball P. Infective pathogenesis and outcomes in chronic
   title and exposure to some airborne agents. Thorax 1991;            bronchitis. Curr Opin Pulm Med 1996; 2:181-85
   46:863-70                                                        28 Mulder J, Goslings WRO, VanDerPlas MC, et al. Studies on
 4 Alonso J, Anto JM. Encuesta de salut de Barcelona 1986              the treatment with antibacterial drugs of acute and chronic
   (government publication). Barcelona, 1989                           mucopurulent bronchitis caused by Haemophilus influenzae.
 5 Brotons B, Perez JA, Sanchez-Toril F, et al. Prevalencia de la      Acta Med Scand 1952; 143:32-49
   enfermedad obstructiva cronica y el asma: estudio transversal.   29 Lees AW, McNaught W. Bacteriology of lower respiratory
   Arch Broncopneumol 1994; 103:481-84                                 tract secretions, sputum and upper respiratory tract secre-
 6 Lundback B, Nystrom L, Rosenhall L, et al. Obstructive lung         tions in “normals” and chronic bronchitics. Lancet 1959;
   disease in northern Sweden: respiratory symptoms assessed           2:1112-15
   by postal questionnaire. Eur Respir J 1991; 4:257-66             30 Gump DW, Phillips CA, Forsyth BR, et al. Role of infection
 7 Lai CKW, Ho SC, Lau J, et al. Respiratory symptoms in               in chronic bronchitis. Am Rev Respir Dis 1976; 113:465-74
   elderly Chinese living in Hong Kong. Eur Respir J 1995;          31 Davies BI, Maesen FPV, Teengs JP, et al. The quinolones in
   8:2055-61                                                           chronic bronchitis. Pharm Weekly (Sci) 1986; 8:53-59
 8 Indonesia Medical Drug Index, 1994                               32 Chodosh S. Treatment of acute exacerbations of chronic
 9 Royal College of General Practitioners. Chronic bronchitis in       bronchitis: state of the art. Am J Med 1991; 90(suppl
   Great Britain. BMJ 1961; 2:973                                      6A):87S-92S
10 Horsley JR, Sterling IJN, Waters WE, et al. Respiratory          33 Lindsay G, Scorer HJN, Carnegie CMD. Safety and efficacy
   symptoms among elderly people in the New Forest Area as             of temafloxacin versus ciprofloxacin in lower respiratory tract
   assessed by postal questionnaire. Age Ageing 1991; 20:325-31        infections: a randomized, double-blind trial. J Antimicrob
11 Smith KR, Aggarwal AL, Dave RM. Air pollution and rural             Chemo 1992; 30:89-100
   biomass fuel in developing countries: a pilot study in India     34 Ball P. Efficacy and safety of cefprozil versus other
   and implications for research and policy. Atmos Environ             beta-lactam antibiotics in the treatment of lower respiratory
   1983; 17:2343-62                                                    tract infections. Eur J Clin Microbiol Infect Dis 1994;
12 Pandey MR. Domestic smoke pollution and chronic bronchi-            13:851-56
   tis in a rural community of the hill region in Nepal. Thorax     35 Sportel JH, Koeter GH, van Altena R, et al. Relation between
   1984; 39:337-39                                                     beta-lactamase producing bacteria and patient characteristics

                                                                                 CHEST / 113 / 3 / MARCH, 1998 SUPPLEMENT         203S
   in chronic obstructive pulmonary disease. Thorax 1995; 50:            outpatients using the protected specimen brush. Am J Respir
   249-53                                                                Crit Care Med 1995; 152:16-20
36 Fagon J-Y, Chastre J, Trouillet J-L, et al. Characterization of    38 Ball P, Wilson R. The epidemiology and management of
   distal bronchial microflora during acute exacerbation of              chronic bronchitis, the Cinderella of chest diseases: epidemi-
   chronic bronchitis. Am Rev Respir Dis 1990; 142:1004-08               ology and management of chronic bronchitis: a selection of
37 Monso E, Ruiz J, Rosell A, et al. Bacterial infection in chronic      classic papers. Worthing, England: Cambridge Medical Pub-
   obstructive airways disease: a study of stable and exacerbated        lications, 1995

204S                                                                                         Disease Management of Pulmonary Infections