Multicenter Evaluation of Azithromycin and Cefaclor in Acute Lower Respiratory Tract Infections DIANA DARK, M.D., Kansas City, ~issoun

This was a randomized, third-party-blinded, multicenter study that compared once-daily azithromycin (500 mg on day 1, followed by 250 mg on days 2-5) to cefaclor (500 mg three times daily for 10 days) in the treatment of patients with acute bronchitis or pneumonia. A total of 546 patients were entered into the study and 272 patients were evaluable for effi- cacy analysis. Of these, 249 (176 azithromycin, 73 cefaclor) had bronchitis and 23 (15 azith- romycin, 8 cefaclor) had pneumonia. The com- bined clinical cure and improvement rate, as determined by the investigator, was 96% for azithromycin and 94% for cefaclor, with 88% bacteriologic eradication in both treatment groups. The elimination of Haemophilus influ- enzae was significantly better with azithromycin (94.5%) than with cefaclor (61.1%) (p CO.001; Fisher’s exact two-tail test). The two antibiotics were well tolerated during this study; the incidence of side effects re- ported was similar for azithromycin and cefaclor. Approximately two thirds of the side effects were mild. Only minor abnormalities in the screening laboratory tests were noted. This study shows that a 5-day course of once-daily azithromycin is as effective as a lo-day three times daily course of cefaclor in the treatment of patients with acute lower respiratory tract infections.

From the Veterans Administration Medical Center, Medical Services Office, Kan-

This iork was supported by a research grant from Pfizer Central Research,

Requests for reprints should be addressed to Diana Dark, M.D., Veterans Administration Medical Center, Medical Services Office, 4801 Linwood Boule- vard, Kansas City, Missouri 64128.

A zithromycin is an azalide, a new class of antibi- otics, now undergoing evaluation as a treat-

ment for a wide range of bacterial infections [l]. Azithromycin is similar in chemical structure to the widely used macrolide antibiotic erythromycin, but has a methyl-substituted nitrogen in the macrolide ring. This modification is responsible for marked alterations of in vitro antibacterial activity, espe- cially against Gram-negative organisms [2]. The antibacterial spectrum of azithromycin is similar to that of erythromycin against Gram-positive bacte- ria; however, there is increased activity against many Gram-negative organisms. Azithromycin also demonstrates significant in vitro activity against the pathogens causing atypical respiratory disease, such as Legionella pneumophila, Mycoplasma pneumoniae, and Chlamydia pneumoniae [2,3]. In addition, azithromycin has a pharmacokinetic pro- file distinct from erythromycin, with superior sta- bility in an acid environment such as the stomach. Studies in animals and humans have also demon- strated a uniformly long elimination half-life and have shown high concentrations in many tissues and body fluids, most notably lung tissue [41.

Macrolide antibiotics such as erythromycin are often prescribed for treatment of upper and lower respiratory tract infections (URTIs, LRTIs) both in children and adults. The clinical usefulness of erythromycin is often limited, however, by gastro- intestinal intolerance and its inconsistent effective- ness against many organisms common in LRTIs, notably Haemophilus influenxae. Other antibiotics commonly prescribed for acute LRTIs include cephalosporins, many of which require relatively short dosing intervals but demonstrate extended antibacterial spectra. However, the cephalosporins do not have activity against the aforementioned pathogens of atypical respiratory tract infections. Several studies have examined the effectiveness of short-course azithromycin in the treatment of acute bronchitis [5,6] and acute bacterial pneumonia [71.

The purpose of this multicenter trial was to eval- uate prospectively the efficacy and safety of a new orally administered antibiotic, azithromycin, com- pared with a commonly prescribed cephalosporin, cefaclor, in the treatment of acute LRTIs.

September 12, 1991 The American Journal of Medrcine Volume 91 (suppl 3A) 3A-31s

PATIENT POPULATION AND METHODS

This third-party-blinded, randomized trial, con- ducted at 26 centers, was designed to compare the safety and efficacy of azithromycin and cefaclor in the treatment of acute LRTIs. Both inpatients and outpatients aged 16 years or older with a clinical diagnosis of acute bronchitis or pneumonia were eli- gible to enter the study. Patients with acute exac- erbations of chronic bronchitis or chronic obstruc- tive pulmonary disease were included. Female pa- tients of childbearing age, unless lactating or preg- nant, were enrolled providing they used adequate contraception during and after the trial. Informed written consent was obtained from all patients. The study was conducted in compliance with institu- tional review board and informed consent regula- tions of each center.

Exclusion criteria for enrollment were: known hypersensitivity or intolerance to macrolide or cephalosporin antibiotics; the presence of active peptic ulcer disease or any condition affecting drug absorption (e.g., gastrectomy, pancreatic insuffi- ciency); treatment with another antibiotic during the 72 hours prior to enrollment; the presence of co-existing chronic bronchitis, bronchiectasis, or chronic obstructive pulmonary disease not accom- panied by acute infection; cystic fibrosis.

Patients were evaluated by a number of clinical, laboratory, and radiographic tests, looking particu- larly for evidence of fever, cough, purulent sputum, rales/rhonchi, leukocytosis, and infiltration on chest radiographs. This initial assessment was substanti- ated by a bacteriologic diagnosis based on culture demonstration of a likely etiologic organism in spu- tum obtained within 48 hours before starting treat- ment. Sputum samples with more than 25 polymor- phonuclear leukocytes and less than 10 squamous epithelial cells per low-power field of Gram-stained sputum were cultured. If a specimen of expecto- rated sputum could not be obtained, induction of sputum with nebulized saline, tracheal aspiration, or fiberoptic bronchoscopy was carried out.

Criteria for determining the presence of acute bacterial LRTI were: (a) purulent sputum with Gram-stain characteristics as just outlined; (b) clini- cal evidence of acute infection with at least one of the following: fever (~38°C); true rigors; leukocyto- sis (white cell count ~12,000); (c) local findings sug- gestive of LRTI @ales, rhonchi, wheezes, or evi- dence of consolidation).

Disease definitions include the above criteria plus: (a) for pneumonia: an acute pulmonary infil- trate, apparent on chest radiographs upon study entry or during the course of therapy, which could not be attributed to some other etiology (e.g., con-

gestive heart failure); (b) for acute purulent bron- chitis: the absence of an acute infiltrate on any chest radiograph taken immediately prior to or dur- ing the course of therapy; (c) for acute exacerbation of chronic obstructive pulmonary disease (COPD): a significant increase in the amount of sputum pro- duction (80% more than the usual daily sputum quantity) and an increase in purulence compared with that of the sputum usually produced by the patient, plus a clinically established diagnosis of chronic bronchitis, emphysema, asthma, or asth- matic bronchitis, and the presence or absence of an acute infiltrate on chest radiographs defined an acute bacterial exacerbation of COPD with pneu- monia or with purulent bronchitis, respectively.

Patient response to treatment was monitored by assessing clinical signs and symptoms (fever, dysp- nea, sputum production, cough, and chest sounds) at baseline and on study days 6 (-+l day), 11 (be- tween days 10 and 13), 18 (21 day), 30 (?l day), and at any other time, as clinically indicated. Clini- cal response was classified as satisfactory or unsat- isfactory: Satisfactory indicated a patient was cured if signs and symptoms of infection resolved during the study with no evidence of infection at day 11 (days 10-13). A patient was judged to be improved if signs and symptoms had subsided dur- ing the study but there was incomplete resolution by day 11 (days 10-13). Unsatisfactory indicated failure of therapy with no apparent clinical re- sponse at day 11.

Bacterial eradication was defined as elimination of the initial causative pathogen by day 11 (days 10-13). Culture and susceptibility studies were performed on sputum, samples at baseline and on study days 6 and 11 and repeated when clinically indicated. Patients with positive radiologic findings at baseline had follow-up chest radiographs at the end of therapy (day 11 visit), and on study comple- tion if infiltrates were not resolved at the day 11 visit.

Susceptibility of all causative organisms to the study drugs was determined using the Kirby-Bauer method (disk-diffusion or broth dilution methods). For the purposes of this study, the criteria for de- termining susceptibility to azithromycin (15 pg disks) were: zone-inhibition diameters of ~18 mm for susceptible category; 14-17 mm for intermedi- ate category; and 513 mm for resistant organisms. The corresponding minimum inhibitory concentra- tions were 52 mg/L, 4 mg/L, and 28 mg/L, respec- tively. If treatment randomization was delayed until the results of the bacteriology testing and an- tibiotic susceptibilities were available, the causa- tive organism had to be susceptible to both drugs. However, if patients were randomized to treatment

SYMPOSIUM ON TISSUE-DIRECTED ANTIBIOTIC THERAPY! DARK

3A-32s September 12, 1991 The American Journal of Medicine Volume 91 (suppl 3A)

before the bacteriologic results were reported, then the pathogen had to be susceptible to the assigned drug.

Patients were evaluated for side effects at each follow-up visit; screening laboratory tests, includ- ing urinalysis, hematology, and chemistry profiles, were performed at baseline and on all scheduled follow-up visits.

A total of 552 patients were randomized in a 2:l ratio by a third party to prevent the investigator knowing the patient’s study medication. Six pa- tients did not take any study medication and were excluded from analysis. In the remaining 546 pa- tients, 367 patients (189 males, 178 females) were treated with azithromycin (500 mg as two 250 mg capsules on day 1, followed by 250 mg on days 2-5 as a single daily dose) and 179 patients (87 males, 92 females) were treated with cefaclor (500 mg three times daily for 10 days). The majority of patients were diagnosed as having acute bronchitis: 93% of azithromycin patients and 92% of cefaclor patients. The remaining patients in each group had a diagno- sis of acute pneumonia. The mean age was similar in each group: 51.4 years (range, 16-91) in the azithromycin group and 51.0 years (17-87) in the cefaclor group.

RESULTS Total enrollment was 552 patients, 546 of whom

actually took the assigned antibiotic (azithromycin 367, cefaclor 179). Excluded from the efficacy anal- ysis were 274 patients, 176 in the azithromycin group and 98 in the cefaclor group. The presence of a resistant pathogen was the reason for exclusion in 26 azithromycin patients (7.1%) and in 14 cefaclor patients (7.8%). Other reasons for exclusion were: no baseline pathogen (118 azithromycin, 63 cefaclor); no end of therapy assessment (24 azithromycin, 13 cefaclor); no susceptibility results (3 azithromycin, 3 cefaclor); inadequate duration of therapy (1 azithromycin, 4 cefaclor); concomitant systemic antibiotic therapy (3 azithromycin, 1 cefaclor); inapplicable baseline diagnosis (1 azith- romycin, 0 cefaclor). The efficacy data were there- fore based on 272 evaluable patients, 191 azithromycin patients (52% of those entered) and 81 cefaclor patients (45%). Of these evaluable pa- tients, 249 (176 azithromycin, 73 cefaclor) had bron- chitis and 23 (15 azithromycin, 8 cefaclor) had pneu- monia as their primary diagnosis.

The clinical response for each evaluable patient, as judged by the investigator at the end of the study, is shown in Table I. The overall response (cured and improved) at the end of therapy was very good in each treatment group: 96.3% with azithromycin and 95.1% with cefaclor. The clinical

[ TABLE I Investigators’ Assessment of Clinical Response at the End of Therapy in the 272 Evaluable Patients

Treatment Group

Clinical Response

Cured Improved Failed Total evaluable

Azithromycin

No. %

36.1 1;: 60.2 3.7

19: 100.0

Cefaclor

No. %

30.9 :: 64.2

8: 4.9

100.0

TABLE II Organisms Isolated from Sputum on Entry to Study

Treatment Group

Aziiromycin Cefaclor

Organism No. % No. %

Moraxella (Lkanhame//a) catarMs 213:: ii 10.1 Haemophibs influenzae 18.2 Haemophiilus parahaemo@cus Haemophlus parainfluenzae 2: 1’: 4.0

17.2 K/ebsie//a pneumoniae 5.1 0 Staphylococcus aureus 13.8 1i.l Stre$xoccus pneumomae 11.8 :: 22.2

* 23.6 14.1 Total isolated lM1 i 100

L

SYMPOSIUM ON TISSUE-DIRECTED ANTIBIOTIC THERAPY / DARK

*Other organisms: Acinetobacter calcoaceticus anifratus, Acinetobacfer calcoaceficus Iwofi, Aeromonas hydrophila, Citrobacter freundii, Escherichia co/i, Enterobacter aero- genes, Enterobacter agg/omerans, Enterobacfer cloacae, Enterobacter sakazakii, Entero- batter spp., Enterococcus spp., Haemophilus spp., Klebsiella oxyfoca, Moraxella osloen- SE, Moraxella spp., Neisseria meningndis, Neisseria (acfamica, Streptococcus pyogenes, Streptococcus anginosus, .Sfreptococcus sanguis, Streptococcus agalacticae, Streptococ- cus group G, Streptococcus mitis.

cure rate was 36.1% in patients treated with azithromycin compared with 30.9% in patients treated with cefaclor. In addition, 60.2% and 64.2% of patients, respectively, were clinically improved.

Thirty different organisms were identified in the 272 evaluable patients, some of which were causa- tive pathogens. The most common organisms in the azithromycin group were H. influenzae (21.7%), Staphylococcus aureus (13.8%), Streptococcus pneumoniae (ll.S%), H. parainfluenxae (9.1%), and Branhamella catarrhulis (8.7%) (Table II). In the cefaclor group, S. pneumoniae accounted for 22.2% of the causative pathogens; H. influenxae (18.2%), H. parainfluenxae (17.2%), and S. aureus (14.1%) were the next most common isolates. The overall bacteriologic cure rate for the evaluable pa- tients in this study was similar for both antibiotics: 88.2% in the azithromycin group and 87.9% in the cefaclor group.

The bacteriologic cure, as judged by eradication in the sputum culture of the initial causative organ- ism in patients with pneumonia, was 100% in pa- tients treated with cefaclor and 94.1% in patients treated with azithromycin (Table III).

September 12, 1991 The American Journal of Medicine Volume 91 (suppl 3A) 3A-33s

TABLE III Bacteriologic Response by Primary Diagnosis

Primary Diagnosis

Treatment Group Bronchiis Pneumonia

Azithromycin Total no. organisms 237 17 Total eradicated 208 % organisms eradicated 81.8% ;i.l%

Cefaclor Total no. organisms Total eradicated ;; ;; % organisms eradicated 86.2% 100.0%

One patient with pneumonia caused by S. aureus had a persistent pathogen. Azithromycin and cefaclor were equally effective bacteriologically in the patients with bronchitis (87.3% and 86.2%, re- spectively).

The bacteriologic results revealed some notable differences in the response of individual pathogens to azithromycin and cefaclor (Table IV). The elimi- nation of H. influenzae was significantly better with the azalide (94.5%) than with the cephalo- sporin (61.1%) (p <O.OOl, Fisher’s exact two-tail test).

The safety reports were based on the 546 pa- tients who took at least one dose of study medica- tion. The compliance rate in this study was good overall; the mean duration of antibiotic treatment was 4.9 days for azithromycin and 8.8 days for cefaclor (duration in protocol was 5 and 10 days, respectively). A total of 112 patients (20.5%) re- ported adverse effects: 21.5% of azithromycin pa- tients (79 of 367) and 18.4% of cefaclor patients (33 of 179). Gastrointestinal disturbances (nausea, vomiting, diarrhea) accounted for 65.5% (azith- romycin) and 61.0% (cefaclor) of these adverse ef- fects. The great majority of these events were mild with only two azithromycin patients and five cefaclor patients withdrawn due to drug-related side effects.

A number of minor abnormalities in screening laboratory tests were found. The overall incidence of baseline abnormalities was 12.7% in the azithromycin group and 13.3% in the cefaclor group, although only 0.58% and 0.57% were attrib- uted to azithromycin or cefaclor, respectively. Drug-related changes in white blood cell counts and elevation of liver function tests were noted in a small number of patients in both treatment groups.

Patients in both groups reported a variety of in- tercurrent illnesses: 77 patients (21.0%) random- ized to the azithromycin group and 37 patients (20.7%) randomized to cefaclor. A number of these

illnesses were respiratory tract infections other than bronchitis or pneumonia.

COMMENTS

SYMPOSIUM ON TISSUE-DIRECTED ANTIBIOTIC THERAPY / DARK

Respiratory tract infections include one of the most common and most life-threatening illnesses. Acute bronchitis may present as nothing more than a transient cough with sputum production in an otherwise healthy individual. Acute bronchitis is a very common disease, accounting for approxi- mately 12 million physician visits per year in the United States [8]. Pneumonia, however, although often treated on an outpatient basis, is a common cause of hospitalization in adults, making up 10% of adult hospital admissions per year (91. Acute pneu- monia ranks sixth,among all causes of death in the United States and should be considered a potenti- ally fatal illness, especially in those patients with chronic underlying medical problems [91.

LRTIs, such as bronchitis and pneumonia, can be caused by a number of pathogens. Although many episodes of acute bronchitis are viral in origin, pa- tients with underlying chronic bronchitis or chronic obstructive pulmonary disease are especially prone to the development of acute exacerbations, which may be of bacterial etiology. The use of antibiotics as standard therapy in the treatment of acute exac- erbations of chronic obstructive pulmonary disease has been shown to shorten the course of illness in outpatients [lo-121 and is considered standard therapy by many physicians. The use of erythromy- tin and beta-lactam antibiotics is often limited in this patient population because of the number of bacterial pathogens that are resistant to these drugs, especially H. influenzae and B. catarrhulis. It is for this reason that azithromycin is particularly attractive. Moreover, its long elimination time and relatively low incidence of adverse effects may in- crease compliance, although a clear difference in compliance compared with cefaclor was not ob- served in this study.

Community acquired pneumonia is most com- monly caused by S. pneumoniae (50% or more of cases) with a lesser incidence of L. pneumophila, H. influenzae, M. pneumoniae, and S. aureus. Anaerobes and other Gram-negative enteric organ- isms account for some of these infections [9]. Clini- cians treating patients with community acquired pneumonia should strive for specific treatment based on Gram-stain and/or sputum culture results. An empiric drug regimen may be necessary when the Gram-stain is difficult to interpret or if sputum is not available for laboratory analysis. Azith- romycin demonstrates an optimal in vitro spectrum

3A-34s September 12, 1991 The American Journal of Medicine Volume 91 (suppl 3A)

SYMPOSIUM ON TISSUE-DIRECTED ANTlBlOTlC THERAPY / DARK

TABLE IV Bacteriologic Response of the Most Frequently Isolated Organisms

Treatment Group

Azithromycin Cefaclor

Organisms No. Eradicated/No. Assessed 91 Eradicated No. Eradicated/No. Assessed 16 Eradicated

Staphjiococcus aureus 32135 91 13114 Streptococcuspmmoniae 29130 ii* 21122 ;:

Haemophihn Muenzae 52155 11/18 Moraxeh (8ranharnela) catarrhalis 22122 100 9110 !J Haemo@ihn paminfbenzae 17123 74 16117 94

KlebsieUa pneumofriae 10113 112 Haemophihs parahaemo~cus 12116 fi 414 1: *p 10.001 (Fisher’s exact two-tail test) compared with cefaclor.

of activity against the most common community- acquired respiratory pathogens.

Ascertaining the true significance of bacterial isolates from sputum cultures remains a common problem. Some isolates may well have represented commensal organisms and not true pathogens and this is a limitation encountered in many trials of this nature.

In conclusion, azithromycin shows good in vitro and in vivo activity against many of the common bacterial pathogens of acute LRTIs. In this study, azithromycin had superior activity against H. influ- enzae when compared with cefaclor and was a safe and well-tolerated drug. A 5-day course of once- daily azithromycin was equally as effective as a lo- day three times daily course of cefaclor in the treat- ment of patients with acute LRTIs.

ACKNOWLEDGMENTS The 26 investigators who recruited patients into this study are gratefully acknowl- edged. The active participants were: Dr. A H. Arrington, 420 Lowell Drive, Suite 504. Huntsville, AL 35801; Dr. R. Bogin, National Jewish Center for Immunology, Respira- tory Medicine, 1400 East Jackson, Denver, CO 80206; Dr. B. D. Breland, 6765 Beaver Court, Midland, GA 31820; Dr. G. Carr, Sumrall Medical Clinic, Highway 42 and Center Avenue, Sumrall, MS 39482; Dr. D. I. Charles, University Health Service, OSU, 308 Wilce Hall, 1875 Milikin Road, Columbus, OH 43210; Dr. D. Dark, Veterans Administration Medical Center, Medical Services Office, 4801 Linwood Boulevard, Kansas City, MO 64128; Dr. A. C. DeGraff Jr, 85 Seymour Street, Suite 716, Hartford, CT 06106; Dr. J. C. Donaldson, Clmic for Chest Diseases, 10192 Coggins Drive, Sun City, AZ 85351; Dr. E. Fishman. Century Park Research Associates, 2080 Century Park East Suite 601, Los Angeles, CA 90067; Dr. W. J. Henry Ill, 554 Memorial Drive, Ext. Greer, SC 29651; Dr. R. Hooper, Southwest Chest Consultants, 525 N 18th Street, #302, Phoenix, AZ 85006; Dr. K. D. Jacobson, Oregon Research Group, 1180 Patterson Street, Suite 2A, Eugene, OR 97401; Dr. L. E. Kline, Division of Chest and Critical Care Medicine, Scripps Clinic, 10666 North Torrey Pines Road, La Jolla, CA 92037; Dr. J. Mack, The Portland Clinic, 800 SW 13th Street, Portland, OR 97205; Dr. W. Morowitz, The Delaware Valley Lung Center, 1949 Route 70 East, Cherry Hill, NJ 08003; Dr. A. Munim, 6405 North Federal Highway, Ft Lauderdale, FL 33308; Dr. J. W. Nielsen, 1015 W Washington Street, Sandusky, OH 44870; Dr. R. Price, 1300 North 12th Street, Suite 605, Phoenix, AZ 85006; Dr. C. Rose, Medical Associates

Clinic, Langworthy Drive, Dubuque, IA 52001; Dr. A. R. Rosenthal, Heartland Re- search Center, 100 E. Wayne Street, Suite 545, South Bend, IN 46601; Dr. T. J. Saddoris, Providence Professional Plaza, Suite 322, 510 NE 49th Street, Portland, OR 97213; Dr. G. Scott, University of Missouri Medical School, Pulmonary Section, Room MA 419, One Hospital Drive, Columbia, MO 65212; Dr. W. Stem, Respiratory Disease Associates, 920 East 2&h, Minneapolis, MN 55404; Dr. A R. Varraux, 6149 Chancellor Drive, Suite 800, Orlando, FL 32809; Dr. S. T. Westerman, 499 Broad Street, Shrewsbuty, NJ 07702; Dr. J. Zaremba, 9131 Parkway East, Birmingham, AL 35206.

REFERENCES 1. Bright GM, Nagel AA, Bordner J, et al. Synthesis, in vitro and in vivo activity of novel 9-deoxo-9a-aza-9a homoerythromycin A derivative; a new class of macrolide antibiotics, the azalides. J Antibiot 1988; 41: 1029-47. 2. Retsema J, Girard A, Schelkly W, et al. Spectrum and mode of action of azithromycin (CP-629931, a new 15.membered-ring macrolide with improved po- tency against Gramnegative organisms. Antimicrob Agents Chemother 1987; 31: 1939-47. 3. Chirgwin K, Roblin PM, Hammerschlag MR. In vitro susceptibilities of Ch/afnydia pneurnoniae (Chlamydia sp. strain TWAR). Antimicrob Agents Chemother 1989; 33: 1634-5. 4. Girard AE, Girard D, English AR, et a/. Pharmacokinetic and in vivo studies with azithromycin (CP-62,993), a new macrolide with an extended halFlie and excellent tissue distribution. Antimicrob Agents Chemother 1987; 31: 1948-54. 5. Bdhning W, Himanen P, Hoepelman IM, Stokke HP, Yernault JC. Three-dose (3. day) treatment of acute bronchitis with azithromycin. Br Med J (in press). 6. Balmes R, Clerc G, Dupont B, Labram C, Pariente R, Poirier R. A comparative study of azithromycin and amoxicillin in the treatment of lower respiratory tract infections. Eur J Clin Microbial Infect Dis 1991; 10: 437-39. 7. APRIM, Gaillat J, Sedallian A, Le Not P, Brion JP, Micoud M. Azithromycine versus josamycine: essai clinique prospectii au tours de pneumonies aigiies com- munautaire (abstract). Reunion lnterdisciplinaire de ChimiothCrapie Antiinfectieuse Paris, December 7, 1989. 8. Rodnick JE, Gude JK. The use of antibiotics in acute bronchitis and acute exacer- bations of chronic bronchitis. West J Med 1988; 149: 347-51. 9. Ellner JJ. Management of acute and chronic respiratory tract infections. Am J Med 1988; 85 (Suppl 3A): 2-5. 10. Anthonisen NR, Manfreda J, Warren CPW, Hershfield ES, Harding GKM, Nelson NA. Antibiotic therapy in exacerbation of chronic obstructive pulmonary disease. Ann Intern Med 1987; 1061 196-204. 11. Nicotra MB, Rivera M. Awe RJ. Antibiotic therapy of acute exacerbation of chronic bronchitis. Ann Intern Med 1982; 97: 18-20. 12. Bates JM. The role of infection during exacerbation of chronic bronchitis. Ann Intern Med 1982; 97: 130-l.

September 12, 1991 The American Journal of Medicine Volume 91 (suppl 3A) 3A-35s