914. Statin Use and Hospital Length of Stay in Adults Hospitalized with Community-Acquired Pneumonia (CAP) in the CDC Etiology of Pneumonia in the Community (EPIC) Study
Session: Poster Abstract Session: Respiratory Infections
Friday, October 4, 2013
Room: The Moscone Center: Poster Hall C
  • EPIC Statin IDWeek poster final.pdf (582.0 kB)

    Background: Statins have anti-inflammatory and immunomodulatory effects, and some retrospective studies suggest that statins might be beneficial for treatment of severe infections, including CAP. The effect of statins on CAP remains unclear, as statin users often have different co-morbid conditions than non-users (confounding by indication), and  may be more likely to engage in health-promoting behaviors than non-users of similar health status (healthy-user bias).

    Methods: Adults ≥ 18 years-old admitted with clinical and radiographic CAP were prospectively enrolled at five hospitals in Chicago and Nashville between January 2010 and June 2012. Adults on statins prior to admission and who continued on statins during hospitalization (statin group) were compared with those who never took statins (control group). Adults who started or stopped statins in the hospital were excluded from this analysis. We used a Cox proportional hazards regression (PH) model to examine the association between statin use and hospital length of stay (LOS), and adjusted for potential confounders, including demographics and co-morbidities. We also compared outcomes between propensity score (PS; propensity to receive statins) matched groups.

    Results:  Of 2,192 adults (88% of all enrolled) included in this analysis, 571 (26%) were in the statin group and 1621 (74%) were in the control group. Statin users were older, had more co-morbid conditions, were better educated, and more likely to have health insurance than controls. Statin users had a longer median LOS compared with controls (92 hours (inter-quartile range (IQR): 55, 166) vs. 75 hours (IQR: 48, 127), p-value < 0.01)). Among 1,974 adults with complete co-variate data for the PH model, statin users and controls had similar LOS (adjusted hazard ratio [aHR]: 0.98; 95% confidence interval (CI): 0.86, 1.1). PS matching retained 346 patients in each group and showed comparable LOS (aHR: 0.93; 95% CI: 0.8, 1.1) (Figure).

    Conclusion:   Substantial confounding affected LOS due to baseline differences between statin users and controls. After adjusting for potential confounders with either a PH or PS model, among adults hospitalized with CAP, we found no difference in LOS between patients with and without statin exposure. 


    Fiona Havers, MD, MHS1, Anna M. Bramley, MPH1, Carrie Reed, DSc, MPH1, Lyn Finelli, DrPH, MS1, Wesley H. Self, MD, MPH2, Christopher Trabue, MD3, Sherene Fakhran, MD, MPH4, Robert Balk, MD5, Mark Courtney, MD6, Timothy Girard, MD, MSCI2, Evan J. Anderson, MD6, Carlos G. Grijalva, MD, MPH2, Kathryn Edwards, MD, FIDSA2, Richard G. Wunderink, MD6 and Seema Jain, MD1, (1)Centers for Disease Control and Prevention, Atlanta, GA, (2)Vanderbilt University School of Medicine, Nashville, TN, (3)University of Tennessee Health Sciences Center, Baptist Hospital, Nashville, TN, (4)John H. Stroger, Jr. Hospital of Cook County, Chicago, IL, (5)Rush University Medical Center, Chicago, IL, (6)Northwestern University Feinberg School of Medicine, Chicago, IL


    F. Havers, None

    A. M. Bramley, None

    C. Reed, None

    L. Finelli, None

    W. H. Self, None

    C. Trabue, None

    S. Fakhran, None

    R. Balk, None

    M. Courtney, None

    T. Girard, None

    E. J. Anderson, None

    C. G. Grijalva, None

    K. Edwards, None

    R. G. Wunderink, None

    S. Jain, None

    Findings in the abstracts are embargoed until 12:01 a.m. PST, Oct. 2nd with the exception of research findings presented at the IDWeek press conferences.