2070. Burden of Clostridium difficile infection (CDI) across hospital inpatient, nursing home, and home settings
Session: Poster Abstract Session: Clostridium difficile: Epidemiology
Saturday, October 29, 2016
Room: Poster Hall
  • IDWeek 2016_CDI transitions_Poster.pdf (779.7 kB)
  • Background: Clostridium difficile infection (CDI) is the main cause of healthcare related infectious diarrhea, with worse effects among the elderly. The settings of recurrent and index CDI may differ, posing challenges to CDI management and leading to greater burden. We followed CDI patients transitioning between hospital, nursing home, and home settings to identify the burden of CDI across the healthcare system.

    Methods: We conducted a retrospective cohort study in linked Medicare, Medicaid, and Minimum Data Set data among CDI patients >65 years old from Jan 2008 to Dec 2010. Index CDI was defined as ICD9-CM code 008.45 with no CDI in prior 84 days. Settings of interest were hospital inpatient (IP), nursing home – including skilled nursing and long term care facilities (NH), and Home – outpatient, home health, or other (H). Patients were followed for 12 months for recurrent CDI (rCDI, ≤ 84 days) and transitions between settings.

    Results: A cohort of 60,828 patients, that was 31% male with a mean age of 81 years, had index CDI distributed as: 45,465 (75%) IP; 2,959 (5%) NH; and 12,404 (20%) H. Home had a lower Charlson Comorbidity Score (3.2) compared to NH (4.3) or IP (4.4). Most patients were hospitalized in the previous year or 90 days (IP 81%, 69%; NH 95%, 93%; H 61%, 42%). Overall 29% had rCDI, observed in 29% of IP, 41% of NH, and 23% of H. The number of rCDI (1, 2, > 2) were: IP (18%, 6%, 5%); NH (22%, 8%, 11%); and H (14%, 5%, 5%). Mean number of transitions were 3.1 ± 3.5 for IP, 3.0 ± 3.1 for NH, and 1.8 ± 3.6 for H. Transitions from IP included 38% to NH by 10 days (median), with 58% of them returning to IP after 23 days at NH. Similarly, 50% of IP were discharged Home by day 8 with 51% readmitted by day 36 at Home. 56% of NH had a 1st transition to IP by day 31.  H transitioned to IP by day 65 (37%) or NH by day 12 (20%).

    Conclusion: Burden may be substantial for subsequent CDI across care settings. NH patients in particular were observed to have more prior hospitalization and rCDI, as well as more transition to IP. Over 50% of Home patients were admitted for care and more than half of IP discharges were readmitted at 2nd transition. Further work is needed to determine whether transitions were due to rCDI. Future studies should consider the possible role of rCDI across settings on overall healthcare system burden.

    Anita Sung, PhD1, Omneya Mohamed, PhD1, Li Wang, MA, PhD, MBA2 and Stephen Marcella, MD, MPH1, (1)Merck & Co., Inc., Kenilworth, NJ, (2)STATinMED Research, Plano, TX


    A. Sung, Merck: Employee , Salary

    O. Mohamed, Merck Sharp & Dhome: Employee , Salary

    L. Wang, STATinMED Research: Employee , Salary
    Merck: Research Contractor , Research support

    S. Marcella, Merck: Employee , Salary

    Findings in the abstracts are embargoed until 12:01 a.m. CDT, Wednesday Oct. 26th with the exception of research findings presented at the IDWeek press conferences.