519. Longer Length of Antibiotic Therapy for Community-Acquired Pneumonia and Risk of Clostridium difficile Infection
Session: Poster Abstract Session: Healthcare Epidemiology: Updates in C. difficile
Thursday, October 4, 2018
Room: S Poster Hall
  • SarahYi_CAPCDI_poster_20180928.pdf (973.3 kB)
  • Background: We previously observed a median 9.5 days length of antibiotic therapy (LOT) among patients with community-acquired pneumonia (CAP) requiring hospitalization (Clin Infect Dis. 2018;66:1333-41). Treatment guidelines for CAP, however, suggest LOT >7 days is rarely necessary. In this study, we evaluated the risk of Clostridium difficile infection (CDI) as a potential harm of longer LOT.

    Methods: This retrospective cohort study included Medicare beneficiaries with parts A, B, and D coverage hospitalized for uncomplicated CAP in 2012-2013 for 2-10 days, home discharge, and no hospitalizations 30 days before or 3 days after index hospitalization. The main exposure was total LOT, represented by the sum of estimated inpatient and observed outpatient LOT, and defined as “longer” if >9.5 days and “shorter” if ≤9.5 days. The outcome, post-discharge CDI, was defined using ICD-9-CM diagnosis code 008.45 in inpatient, skilled nursing, or outpatient claims within 6 months after index hospitalization. CDI 12 months before or during index hospitalization was excluded. CDI risk was assessed through a multivariable logistic model stratified by outpatient antibiotic class and adjusted for confounders including comorbidities, severity via ICU status, demographics, and hospital characteristics.

    Results: The cohort consisted of 99,883 patients. Median total LOT was 9.5 days (IQR: 7.4-11.4). Antibiotics filled at discharge included quinolones (40%), none (20%), multiple (14%), cephalosporins (10%), macrolides (7%), and β-lactam/β-lactamase inhibitor combinations (5%). CDI risk was 1.2%. Overall adjusted risk among those with longer LOT was 1.2 (95% CI: 1.1-1.4) times that of those with shorter LOT. Increased risk was observed among those prescribed quinolones at discharge, for whom adjusted CDI risk for longer LOT was 1.4 (95% CI: 1.2-1.7) times the risk of those with shorter LOT. We observed no difference in risk between longer and shorter LOTs for other antibiotic categories.

    Conclusion: These findings suggest decreased LOT, which can be achieved with better adherence to current treatment guidelines, could reduce risk of subsequent CDI among patients hospitalized with CAP, particularly among those treated with fluoroquinolones at discharge.

    Sarah H. Yi, PhD, Sujan C. Reddy, MD, Sophia V. Kazakova, MD, MPH, PhD, Kelly M. Hatfield, MSPH, James Baggs, PhD, Alice Y. Guh, MD, MPH, Preeta K. Kutty, MD, Lauri A. Hicks, DO, Arjun Srinivasan, MD, L. Clifford McDonald, MD and John A. Jernigan, MD, MS, Division of Healthcare Quality Promotion, Centers for Disease Control and Prevention, Atlanta, GA


    S. H. Yi, None

    S. C. Reddy, None

    S. V. Kazakova, None

    K. M. Hatfield, None

    J. Baggs, None

    A. Y. Guh, None

    P. K. Kutty, None

    L. A. Hicks, None

    A. Srinivasan, None

    L. C. McDonald, None

    J. A. Jernigan, None

    Findings in the abstracts are embargoed until 12:01 a.m. PDT, Wednesday Oct. 3rd with the exception of research findings presented at the IDWeek press conferences.