Antibiotic Treatment Patterns and Outcomes Following Outpatient Treatment of Community-Acquired Bacterial Pneumonia: A US Cohort Study
Methods: CABP episodes were identified in the Truven database between January 1, 2007 and December 31, 2012 in patients 18 years of age or older with a pneumonia diagnosis and treated with monotherapy beta-lactam (BL), fluoroquinolone (FQ), or macrolide (M) preceded by both ≥30 days free of institutional care and ≥90 days free of antibiotic use. The effects of monotherapies BL v M and FQ v M on risk of adverse events, pneumonia-related hospital visits and treatment failure within 30 days of CABP episode were estimated after adjusting for pre-treatment demographics, health care resource use and costs using multivariate regression models. Treatment failure was defined as switch in drug class ≥ 6 days after initial CABP treatment.
Results: Out of a total of 441,820 CABP patients, 84% were initiated with the following monotherapies: BL (N=36,702), FQ (165,768) and M (N=169,335). Comparing the three cohorts revealed that patients receiving FQ therapy were the oldest while BL treated patients were the sickest. FQ v. M initiation was associated with an increased risk of an adverse event (OR=1.23, 95% CI: 1.20-1.25), increased 30-day odds of pneumonia-related hospital visits (OR=1.83, 95% CI: 1.69-1.98) and a reduced 30-days odds of treatment failure (OR=0.90, 95% CI: 0.87-0.94). BL v. M initiation was associated with an increased risk of an adverse event (OR=1.26, 95% CI: 1.23-1.31), increased 30-day odds of pneumonia-related hospital visits (OR=1.72, 95% CI: 1.53-1.93) and an increased 30-days odds of treatment failure (OR=1.27, 95% CI: 1.20-1.35).
Conclusion: Compared to initiating outpatient CABP treatment with M monotherapy, initiation with FQ or BL monotherapy was associated with a significantly increased risk of an adverse event and pneumonia-related hospital visit. The risk of treatment failure was reduced with FQ but increased with BL. Thus, treatments are needed that are both effective and have few side effects.
Cempra, Inc: Consultant, Research support
E. Tuttle, Cempra, Inc: Consultant, Research support