Equal effectiveness of older traditional antibiotics and newer broad-spectrum antibiotics in treating patients with acute exacerbations of chronic bronchitis.

Background: Choice of antibiotic therapy for acute exacerbations of chronic bronchitis (AECB) is important because of cost and concerns about development of resistance.

Methods: A retrospective cohort study was conducted in outpatients with chronic obstructive pulmonary disease and documented AECB treated with antibiotics.

Results: We compared outcomes and costs of AECB treated with first-line antibiotics having partial coverage against organisms associated with AECB (eg, amoxicillin), first-line antibiotics with full coverage against organisms associated with AECB (eg, sulfamethoxazole-trimethoprim), and newer broad-spectrum antibiotics (eg, azithromycin). There were no significant differences among the three antibiotic groups in failure rate, hospitalization rate, or time until subsequent AECB. Pharmacy costs were significantly less with first-line antibiotics ($3.18 [+ or -] 0.64, $3.00 [+ or -] 0.48, and $36.70 [+ or -] 16.29, respectively; P < 0.0001), but there was no significant difference among all three groups with regard to total costs.

Conclusion: We found no difference in outcome between older traditional antibiotics with adequate coverage against organisms associated with AECB and newer broad-spectrum antibiotics.

**********

Although antibiotics are routinely given for the treatment of acute exacerbations of chronic bronchitis (AECB), this practice has been questioned because the overuse of antibiotics promotes resistance and because not all AECB are infectious. Despite this, clinical trails have found that patients with AECB benefit from antibiotic therapy. (1,2) Although some studies have assumed all antibiotics are equally effective in the treatment of AECB, (1) the choice has recently become an important consideration because of the development of resistance to the older traditional antibiotics commonly used to treat AECB.

In addition to the antibiotic concerns, the economic implications of treating AECB are also significant. The costs of treating AECB include not only the cost of antibiotic therapy but also the high cost of hospitalization when outpatient treatment fails. In a recent study using data from 1994, there were more than 10 million outpatient visits and 280,000 hospital admissions for AECB at an estimated cost of $1.5 billion in hospital stays alone. (3) Therefore, if a more expensive antibiotic could prevent hospitalization, it would be considered highly cost-effective.

Several recent reviews on the choice of antibiotics for the treatment of AECB have suggested that the newer, expensive broad-spectrum antibiotics (eg. fluoroquinolones and newer macrolides) might be preferable to the older traditional antibiotics (eg, amoxicillin and sulfamethoxazole-trimethoprim [SMX-TMP]). (4-6) This recommendation is based on the belief that these newer antibiotics are associated with better outcomes and thus an overall lower cost of care; however, the evidence relies primarily on an industry-sponsored, retrospective study by Destache et al (7) that was published in a journal supplement.

The study by Destache et al (7) evaluated the efficacy and related costs for the treatment of AECB. Antibiotics were grouped into first-line, second-line, or third-line agents on the basis of use at their facility, with first-line agents being older traditional antibiotics and third-line agents being newer broad-spectrumantibiotics. The authors concluded that although third-line agents were more expensive, they decreased the failure rate of antibiotics, decreased the need for hospitalization, prolonged the time between exacerbations, and resulted in lower total costs as compared with first-line agents. On the basis of these results, it was concluded that third-line agents might be more cost-effective. Although this study has been quoted to support the use of newer broad-spectrum antibiotics in the treatment of AECB, it has several limitations that compromise the conclusions. Therefore, we sought to verify the findings of Destache et al (7) in a similarly designed trial at our institution, where we evaluated the clinical and economic implications of treating AECB with older traditional antibiotics compared with newer broad-spectrum antibiotics.

Materials and Methods

Potential patients were identified from our pharmacy database, and the medical records and pulmonary function laboratory results of these patients were reviewed for potential inclusion in the study. Patients who had a clinical diagnosis of chronic obstructive pulmonary disease (COPD) and who received outpatient antibiotic therapy for an AECB between January 1999 and December 1999 were studied. Data collected included pulmonary function tests (PFTs), supplemental oxygen dependence, current and past smoking status, age, race, sex, long-term oral steroid use ([greater than or equal to]3 mo), and coexisting disease (diabetes mellitus, hypertension, coronary artery disease, cancer, congestive heart failure, liver disease, or renal insufficiency, defined as a serum creatinine value >2 mg/dl). In addition, we classified the severity of each patient's COPD, based on their PFTs, as Stage 1 (mild) if forced expiratory volume in one second (FE[V.sub.1]) was [greater than or equal to]50% predicted, Stage 2 (moderate) if FE[V.sub.1] was 35 to 49% predicted, or Stage 3 (severe) if FE[V.sub.1] was <35% predicted. (8)

Patients were excluded if they 1) had no documented COPD; 2) received an antibiotic that was not used for an AECB; 3) received antibiotic therapy for an AECB while hospitalized; 4) had a standing prescription for self-administered antibiotics; 5) had no PFTs within the past 3 years; 6) had an AECB within 3 months before their first visit in 1999; 7) had radiographic evidence of pneumonia; or 8) had documented acquired immunodeficiency syndrome, unstable angina, New York Heart Association Class IV heart failure, clinical or laboratory data suggestive of atypical pathogens, tuberculosis, Pneumocystis carinii pneumonia, post-obstructive pneumonia, active lung cancer, lymphoma, leukemia, ongoing dialysis, or lack of documentation of antibiotic use in the medical records.

We analyzed patient information for 12 months from the time of the initial AECB in 1999. Primary outcomes for the study were treatment failure rate, defined as the number of episodes that required further antibiotics within 2 weeks of the AECB, need for hospitalization within 2 weeks of initial outpatient treatment, time between episodes up to 12 months after the initial exacerbation, and costs associated with the treatment of AECB from the perspective of our institution, a university-affiliated Veterans Affairs medical center. We calculated inpatient costs using our institution's average cost for an AECB hospitalization, and we determined outpatient costs by adding our institution's costs of all laboratory work, radiology, physician or emergency care visits, and pharmacy costs for each episode. Pharmacy costs included the drug acquisition cost and the pharmacy dispensing fee.

We classified antibiotics according to the categorization by Destache et al, (7) which was derived from the consensus of resident pulmonologists' preference in antibiotic choice for the treatment of AECB. Whereas Destache et al (7) classified antibiotics as first-line, second-line, or third-line agents, we simplified the classification of antibiotics into two groups on the basis of antibiotic use and availability at our institution during the study period. Antibiotics were categorized as either first-line agents (amoxicillin, cephalexin, doxycycline, erythromycin, and SMX-TMP) or second-line agents (amoxicillin-clavulanate, azithromycin, cefuroxime, clarithromycin, levofloxacin, and ofloxacin), with first-line agents being older traditional antibiotics and second-line agents being newer broad-spectrum antibiotics. We further divided the first-line antibiotics into subgroups on the basis of the agent's effectiveness against the organisms most commonly associated with respiratory tract infections, namely, Streptococcus pneumoniae, Haemophilus influenzae, and Moraxella catarrhalis. The first-line agents with partial coverage against these organisms were amoxicillin, cephalexin, and erythromycin, and the first-line agents with full coverage were doxycycline and SMX-TMP.

Descriptive statistics were used to report the frequencies of covariates (severity of disease as assessed by PFTs, supplemental oxygen dependence, current and past smoking status, age, race, sex, long-term use of oral steroids, and presence of comorbid conditions) among patient groups based on the use of the three groups of antibiotics. For multiple observations from one patient, logistic regression and Cox regression were used to adjust for dependence. For treatment failure rate, logistic regression with dummy variables coded for group effects was used to test the equality of proportion of failure across antibiotic groups. Any possible confounders identified by univariate comparisons were adjusted for in the model of the group effects. Time to relapse was plotted using Kaplan-Meier curves. Each group was plotted individually, and survival curves were compared using log-rank statistics. If significant differences were detected, the outcome was modeled using Cox regression, adjusting for appropriate covariates as necessary.

Results

Of 602 patients screened for potential inclusion in this study, 121 patients fulfilled the criteria. These patients had 170 documented AECB episodes, averaging 1.40 episodes per patient in 1999. Table 1 summarizes the reasons for excluding 431 patients from the study. The majority of exclusions resulted from lack of documentation of COPD or lack of documentation of antibiotic use in the medical records.

The overall mean age of the patients included in the study was 65.14 [+ or -] 9.06 years. All patients were male, and 77% were white. Table 2 summarizes patient demographics, clinical characteristics, and comorbidities by antibiotic group. There were no statistically significant differences between the groups.

A total of 135 AECB episodes were treated with first-line antibiotics; 48 episodes were treated with first-line antibiotics that have partial coverage against organisms commonly associated with respiratory tract infections, and 87 episodes were treated with first-line antibiotics that have full coverage. A total of 35 AECB episodes were treated with second-line antibiotics. Amoxicillin was the predominantly used first-line antibiotic with partial coverage, SMX-TMP was the most commonly prescribed first-line agent with full coverage, and azithromycin was the predominantly used second-line antibiotic (Table 3).

A total of 17 AECB episodes failed to respond to antibiotic therapy within 2 weeks of initiation. Among these failures, 14 episodes were treated with first-line antibiotics (10.3% failure rate), and 3 episodes were treated with second-line antibiotics (8.6% failure rate). This difference was not statistically significant (P = 0.84). Among the first-line antibiotic treatment failures, first-line antibiotics with partial coverage had been used in nine episodes (18.8% failure rate), and first-line antibiotics with full coverage had been used in five episodes (5.7% failure rate). There was still no significant difference among the groups when the first-line antibiotics were separated according to their effectiveness against the organisms most commonly associated with AECB (P = 0.13).

Of the AECB episodes that failed to respond to initial antibiotic therapy, eight resulted in hospitalization within 2 weeks of initiation. Of those episodes that resulted in hospitalization, seven were initially treated with first-line antibiotics (5.2% hospitalization rate), and one was initially treated with second-line antibiotics (2.9% hospitalization rate). This difference was not statistically significant (P = 0.57). Among patients hospitalized after initial treatment with first-line antibiotics, five were initially treated with first-line agents with partial coverage (10.4% hospitalization rate), and two were initially treated with first-line agents with full coverage (2.3% hospitalization rate). Further breakdown into three antibiotic groups still showed no significant difference among hospitalization rates (P = 0.14).

A subsequent AECB occurred an average of 33.4 [+ or -] 19.7 weeks after treatment with first-line antibiotics, with an average of 36.4 [+ or -] 19.3 weeks for first-line antibiotics that have partial coverage and an average of 31.9 [+ or -] 19.8 weeks for first-line agents that have full coverage. Among those episodes treated with second-line antibiotics, the next AECB occurred an average of 32.9 [+ or -] 19.6 weeks later. There was no significant difference in time to the next AECB between first-line antibiotics and second-line antibiotics (P = 0.73) or among the three antibiotic groups when analyzed separately (P = 0.44).

Table 4 summarizes the cost data. Mean pharmacy costs of the antibiotics were lowest with first-line antibiotics having full coverage (first-line with partial coverage, $3.18 [+ or -] 0.64; first-line with full coverage, $3.00 [+ or -] 0.48; second-line, $36.70 [+ or -] 16.29). The difference in pharmacy costs between first-line antibiotics and second-line antibiotics was significant (P < 0.0001).

Total costs included pharmacy costs, cost for physician or emergency room visits, laboratory and radiology costs, and average hospitalization cost. Average total costs associated with antibiotic treatment for each group were as follows: first-line with partial coverage, $1,024.14 [+ or -] 2,947.73; first-line with full coverage, $227.75 [+ or -] 1,442.21; and second-line, $328.21 [+ or -] 1,615.59. The difference between first-line antibiotics and second-line antibiotics was not significant (P = 0.58). When analyzed separately, the difference among all three antibiotic groups was still not significant (P = 0.22).

{PCO}Discussion

Despite several serious limitations, recent reviews (4 6) have promoted the use of newer broad-spectrum antibiotics for the treatment of AECB on the basis of the retrospective study by Destache et al. (7) Our study challenges this perception that newer broad-spectrum antibioties may be more cost-effective than appropriate traditional antibiotics in the treatment of AECB. However, our findings support the concept that use of antibiotics with only partial coverage against organisms associated with AECB, primarily amoxicillin, may be associated with poorer outcomes and, ultimately, higher overall costs. (9)

There is currently no consensus regarding which antibiotic should be used in the treatment of AECB, but choice of antibiotic has become an important consideration. Over the past decade, there have been increasing numbers of penicillin-resistant strains of Streptococcus pneumoniae and [beta]-lactamase-producing strains of both Haemophilus influenzae and Moraxella catarrhalis isolated from individuals with respiratory tract infections. (10 12) Because of the increasing development of resistance, particularly to older traditional antibiotics, attempts have been made to stratify patients by risk to optimize antibiotic treatment and decrease the risk of treatment failure. (13,14) In their study, Adams et al (9) evaluated 362 patient visits (173 patients) to assess factors predictive of relapse. They concluded that relapse from AECB was not related to the severity of the underlying disease or exacerbation. Patients treated with antibiotics had significantly lower relapse rates, strengthening the argument for the use of antibiotics in the treatment of AECB, even in those presenting with mild symptoms; however, the choice of antibiotic was important because patients treated with amoxicillin had the highest relapse rate. Despite the studies done in an attempt to identify risk factors for failure, whether patient outcomes improve on the basis of such stratification remains to be validated. In our study, we found that age, race, severity of COPD, and the presence of comorbid conditions did not affect primary outcomes such as treatment failure or need for hospitalization.

We sought to confirm the findings of Destache et al (7) in a similarly designed study and to adjust for the methodologic limitations that compromise their conclusions. (15) Namely, they did not stratify patients by severity of disease or PFT results, and they did not separately analyze antibiotics with partial coverage against the organisms most commonly associated with respiratory tract infections. In addition, although they mention that spirometry was not different among the groups, the FE[V.sub.1] was 1.27 L in the first-line antibiotic group compared with the FE[V.sub.1] of 1.71 L for the third-line antibiotic group.

Thus, although the differences were not statistically significant, they may have been clinically significant. In contrast to Destache et al, (7) we included important clinical confounders such as PFTs and severity of disease in our analysis. We also classified antibiotics by use at our facility. Unlike Destache et al, (7) however, we separated older traditional antibiotics into two subgroups on the basis of their effectiveness against organisms most commonly associated with respiratory tract infections. Finally, our study, although small, included more than twice as many patients as the study conducted by Destache et al. (7)

With the further separation of older traditional antibiotics according to their effectiveness against organisms most commonly associated with AECB, we found no difference in the efficacy between older traditional antibiotics with adequate coverage and newer broad-spectrum antibiotics. If we had validated the study by Destache et al, (7) there would have been an impetus to change treatment recommendations at our facility. Equally important, though not statistically significant, we found a trend toward greater failure rates and need for hospitalization with first-line antibiotics such as amoxicillin that have partial coverage against Haemophilus influenzae. Moraxella catarrhalis, and Streptococcus pneumoniae. Our findings support the results of previous studies that show greater failure rates for AECB episodes treated with amoxicillin. (9)

On the basis of their study results, Destache et al (7) suggested that newer broad-spectrum antibiotics may be more cost-effective despite their initial significant pharmacy costs. In contrast, we found no difference between older traditional antibiotics that have adequate coverage against organisms most commonly associated with respiratory tract infections and newer broad-spectrum antibiotics. Although the cost of treatment for all three groups was not significantly different because of large variations in cost and a relatively small sample size, the overall costs and antibiotic costs were lowest in first-line antibiotics with adequate, full coverage.

Our study has limitations. Primarily, it is a retrospective study, and therefore we relied on documentation in the clinical record for important variables such as severity of AECB. Thus, the clinical data were not collected in a standard fashion. Similarly, there may have been clinically significant differences among treatment groups. Although we did not find statistically significant differences in clinical characteristics among the groups, this was likely a result of our relatively small sample size. There may have been other important unmeasured factors, such as prescribing clinicians or specialty type, but we included many important clinical factors such as PFTs, comorbidities, and severity of AECB in our analysis of treatment failures, need for hospitalization, and time until subsequent AECB episode.

The small sample size may have masked significant differences among treatment groups relative to cost, particularly between older traditional antibiotics with partial coverage and the other two groups of antibiotics. Although overall costs among the groups were not significantly different because of large variations, the average costs for patients treated with older traditional antibiotics with partial coverage were more than three times that for all other patients. Another limitation was a lack of documentation of antibiotic use in some medical records, and several patients had a standing prescription for self-administered antibiotics; therefore, we excluded these patients from our analysis. However, it is not likely that one antibiotic was excluded more frequently than another. We did not examine whether patients received care at other institutions, but it is unlikely that any group was more likely to be treated outside our institution, and we doubt that this would have changed our results.

Because AECB is a common medical condition, antibiotic choice carries significant implications. We believe our study makes an important addition to the body of literature addressing antibiotic use in treating AECB. We found no difference in efficacy between older traditional antibiotics with effectiveness against organisms most commonly associated with respiratory tract infections and newer broad-spectrum antibiotics. Therefore, our study challenges the findings of Destache et al, (7) but we acknowledge that both studies have significant limitations. Most importantly, both studies were small retrospective reviews of patients who were treated for AECB in a nonrandomized fashion. Ultimately, the question of the most appropriate antibiotic for the treatment of AECB will require an appropriately powered, prospective, clinical trial.

That which you cannot let go of, you do not possess. It possesses you.

-Ivern Ball

Table 1. Reasons for excluding 431 patients from the study (a)

Exclusion criteria                           No. of patient

No documentation of antibiotic use                 101

No documented COPD                                  99

Antibiotic not used for an AECB                     59

Antibiotic initiated during hospitalization         49

AECB within 3 mo of initial episode in 1999         29

Radiographic evidence of pneumonia                  24

Active lung cancer                                  19

Standing prescription for self-administered
antibiotics                                         13

No PFTs within 3 yr                                  8

Other                                               30

(a) COPD, chronic obstructive pulmonary disease; AECB, acute
exacerbation of chronic bronchitis: PFTs, pulmonary function tests.

Table 2. Patient demographics and clinical characteristics (a)

                                         First-line antibiotics
                                         with partial
Characteristic                           coverage (n= 48)

Mean age (yr)                             65.1 [+ or -] 8.9
Race (% white)                            79
Sex (% male)                             100
Current smoker (%)                        29
Former smoker (%)                         94
Severity of COPD
   1(FE[V.sub.1] [greater than or equal   52
to]50%)
  2 (FE[V.sub.1]35-49%)                   31
  3 (FE[V.sub.1]<35%)                     17
Mean FE[V.sub.1] (L)                      1.56 [+ or -] 0.64
Mean % predicted                         51.1 [+ or -] 17.4%
Supplemental [O.sub.2] dependence         21
Chronic oral steroid use                  13
Comorbidity
  Diabetes                                19
  Coronary artery disease                 33
  Hypertension                            60
  Congestive heart failure                21
  Cancer                                   0
  Serum creatinine >2 mg/dl                2
  Liver disease                            8

                                         First-line antibiotics
                                         with full coverage
Characteristic                           (n = 87)

Mean age (yr)                            65.4 [+ or -] 8.3
Race (% white)                             79
Sex (% male)                              100
Current smoker (%)                         40
Former smoker (%)                          97
Severity of COPD
   1(FE[V.sub.1] [greater than or equal    47
to]50%)
  2 (FE[V.sub.1]35-49%)                    38
  3 (FE[V.sub.1]<35%)                      15
Mean FE[V.sub.1] (L)                      1.63 [+ or -] 0.72
Mean % predicted                          55.0 [+ or -] 22.4%
Supplemental [O.sub.2] dependence           29
Chronic oral steroid use                    16
Comorbidity
  Diabetes                                  22
  Coronary artery disease                   48
  Hypertension                              63
  Congestive heart failure                  31
  Cancer                                     7
  Serum creatinine >2 mg/dl                  3
  Liver disease                              6

                                         Second-line
                                         Antibiotics
Characteristic                           (n = 35)

Mean age (yr)                            63.0 [+ or -] 9.6
Race (% white)                             80
Sex (% male)                              100
Current smoker (%)                         40
Former smoker (%)                          94
Severity of COPD
   1(FE[V.sub.1] [greater than or equal    43
to]50%)
  2 (FE[V.sub.1]35-49%)                     0
  3 (FE[V.sub.1]<35%)                      17
Mean FE[V.sub.1] (L)                      1.45 [+ or -] 0.56
Mean % predicted                          47.9 [+ or -] 17.2%
Supplemental [O.sub.2] dependence           26
Chronic oral steroid use                    17
Comorbidity
  Diabetes                                  14
  Coronary artery disease                   43
  Hypertension                              46
  Congestive heart failure                  26
  Cancer                                     3
  Serum creatinine >2 mg/dl                  0
  Liver disease                              6

Characteristic                           P-value

Mean age (yr)                            0.43
Race (% white)                           0.99
Sex (% male)                             1.00
Current smoker (%)                       0.42
Former smoker (%)                        0.65
Severity of COPD                         0.82
   1(FE[V.sub.1] [greater than or equal
to]50%)
  2 (FE[V.sub.1]35-49%)
  3 (FE[V.sub.1]<35%)
Mean FE[V.sub.1] (L)                     0.35
Mean % predicted                         0.18
Supplemental [O.sub.2] dependence        0.67
Chronic oral steroid use                 0.83
Comorbidity
  Diabetes                               0.60
  Coronary artery disease                0.30
  Hypertension                           0.19
  Congestive heart failure               0.47
  Cancer                                 0.33
  Serum creatinine >2 mg/dl              0.68
  Liver disease                          0.81

 (a) COPD, chronic obstructive pulmonary disease; AECB, acute
exacerbation of chronic bronchitis; FE[V.sub.1], forced expiratory
volume in 1 second.

Table 3. Antibiotic use by group (a)

                                                       No.of
Group                                                  patients

First-line antibiotics with partial coverage (n = 48)
  Amoxicillin                                             42
  Cephalexin                                               1
  Erythromycin                                             5
First-line antibiotics with full coverage (n = 87)
  SMX-TMP                                                 75
  Doxycycline                                             12
Second-line antibiotics (n = 35)
  Amoxicillin-clavulanate                                  7
  Azithromycin                                            16
  Cefuroxime                                               2
  Clarithromycin                                           3
  Levofloxacin                                             5
  Ofloxacin                                                2

(a) SMX-TMP, sulfamethoxazole-trimethoprim.

Table 4. Cost associated with antibiotic treatment

                               First-line antibiotics
                               with partial
Cost                           coverage (n = 48)

Average pharmacy costs ($)       3.18 [+ or -] 0.64
Average total costs ($)      1,024.14 [+ or -] 2,947.73

                              First-line antibiotics
                             with full
Cost                         coverage (n = 87)

Average pharmacy costs ($)     3.00 [+ or -] 0.48
Average total costs ($)      227.75 [+ or -] 1,442.21

Cost                        Second-line
                          antibiotics (n = 35)      P value

Average pharmacy costs ($)   36.70 [+ or -] 16.29         <0.0001
Average total costs ($)     328.21 [+ or -] 1,615.59       0.22

Acknowledgment 

We thank Nancy Brucker, MPH, for assistance with data analysis.

From the Pharmacy Service, Department of Veterans Affairs, Pittsburgh Healthcare System, Pittsburgh, PA.

Reprint requests to Sherrie L. Aspinall, PharmD, BCPS, VA Pittsburgh Healthcare System, Pharmacy Service (132-MU), University Drive C. Pittsburgh, PA 15240. Email: [e-mail]

Accepted July 18, 2002.

Copyright [c] 2003 by The Southern Medical Association

0038-4348/03/9610-0986

References

1. Anthonisen NR, Manfreda J, Warren CP, et al. Antibiotic therapy in exacerbations of chronic obstructive pulmonary disease. Ann Intern Med 1987;106:196-204.

2. Saint S, Bent S, Vittinghoff E, et al. Antibiotics in chronic obstructive pulmonary disease exacerbations: A meta-analysis. JAMA 1995;273:957-960.

3. Niederman MS, McCombs JS, Unger AN, et al. Treatment cost of acute exacerbations of chronic bronchitis. Clin Ther 1999;21:576-591.

4. Adams SG, Anzueto A. Treating acute exacerbations of chronic bronchitis in the face of antibiotic resistance. Cleve Clin J Med 2000;67:625-628, 631-633.

5. Anzueto A, Angel L. Update in pulmonary disease. Ann Intern Med 2000;133:360-366.

6. Destache CJ. Optimizing economic outcomes in acute exacerbations of chronic bronchitis. Pharmacotherapy 2002;22:12S-17S, 30S-32S.

7. Destache CJ, Dewan N, O'Donohue WJ, et al. Clinical and economic considerations in the treatment of acute exacerbations of chronic bronchitis. J Antimicrob Chemother 1999;43(Suppl A):107-113.

8. American Thoracic Society. Standards for the diagnosis and care of patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1995;152;S77-S121.

9. Adams SG, Melo J, Luther M, et al. Antibiotics are associated with lower relapse rates in outpatients with acute exacerbations of COPD. Chest 2000;117:1345-1352.

10. Doern GV, Brueggemann AB, Pierce G, et al. Antibiotic resistance among clinical isolates of Haemophilus influenzae in the United States in 1994 and 1995 and detection of [beta]-lactamase-positive strains resistant to amoxicillin-clavulanate: Results of a national multicenter surveillance study. Antimicrob Agents Chemother 1997;41:292-297.

11. Doern GV, Brueggemann AB, Pierce G, et al. Prevalence of antimicrobial resistance among 723 outpatient clinical isolates of Moraxella calarrhalis in the United States in 1994 and 1995: Results of a 30-center national surveillance study. Antimicrob Agents Chemother 1996;40:2884-2886.

12. Doern GV, Brueggemann A, Holley HP Jr, et al. Antimicrobial resistance of Streptococcus pneumoniae recovered from outpatients in the United States during the winter months of 1994 to 1995: Results of a 30-center national surveillance study. Antimicrob Agents Chemother 1996;40:1208-1213.

13. Grossman RF, Cost-effective therapy for acute exacerbations of chronic bronchitis. Semin Respir Infect 2000;15:71-81.

14. Niederman MS, Antibiotic therapy of exacerbations of chronic bronchitis. Semin Respir Infect 2000;15:59-70.

15. Peng CC, Aspinall SL, Good CB. Update in pulmonary disease. Ann Intern Med 2001;135:301 (letter).

RELATED ARTICLE: Key Points

* Choice of antibiotics in the treatment of acute exacerbations of chronic bronchitis has become important because of cost and concern about the development of resistance.

* Although there was no difference in total costs among antibiotic groups in this study, pharmacy costs were significantly less with older traditional antibiotics used to treat acute exacerbations of chronic bronchitis than with newer broad-spectrum antibiotics.

* There was no difference in efficacy between older antibiotics with adequate coverage of organisms associated with acute exacerbations of chronic bronchitis and newer broad-spectrum antibiotics.

Catherine C. Peng, PHARMD, Sherrie L. Aspinall, PHARMD, BCPS, Chester B. Good, MD, MPH, Charles W. Atwood, Jr., MD, and Chung-Chou H. Chang, PHD