The Interplay Between Acute and Long-Term Care C. difficile Incidence in the United States Veterans Health Administration: A Retrospective Cohort Study of 169 Facilities

Background: Although most acute and long-term care facilities in the same region are coupled by patient sharing and many experience high rates of C. difficile infection (CDI), the inter-facility spread of C. difficileis understudied. Our objective was to consider the determinants of CDI incidence across acute care (AC) and long term care (LTC) facilities, with a specific interest in the role of patient sharing.

Methods: We conducted a retrospective cohort study of CDI from 2006-2012 across the Veterans Affairs Healthcare System. Our outcome was defined as a health-care-associated lab-identified C. difficileevent, defined as a positive lab test in a patient having at least 3 days of facility exposure in the prior 8 weeks, and occurring at least 8 weeks from a previous positive test. Individual-level risk variables included (1) age, (2) antibiotic use, and history of (3) acute or (4) long-term care stay in the prior 56 days. Facility-level predictors included (1) antibiotic use (days with therapy per 1,000 person-days), the proportion of persons with an (2) acute or (3) long-term care stay in prior 56 days, and importation of CDI cases from (4) acute or (5) long-term care per 10 000 person-days.

Results: 87 LTC and 82 AC facilities met our inclusion criteria. The incidence of CDI in LTC was 3.8 per 10,000 patient-days (n=6,766 cases) and was 17.9 per 10,000 patient-days (n=26,113 cases) in AC. LTC patients with a recent AC stay were more likely to develop CDI than those without a recent AC stay (IRR=4.81, 95%CI: 4.56, 5.07). Similarly, AC patients with a recent LTC stay were also more likely to develop CDI than those without, but to a lesser degree (IRR=1.88, 95%CI: 1.79, 1.98). Imported CDI cases were more prevalent in LTC (median=75 per 10,000 patient-days, range: 1.5-355) compared to AC (median=39, range: 0-115, p<0.001). In bivariate weighted linear regression models, importation of AC cases was a strong predictor of increased CDI incidence in LTC (R²=0.62), but importation of cases from LTC was not as strong a predictor of AC rates (R²=0.21).

Conclusion: Acute care and long-term care facilities are both impacted by importation of CDI. This research suggests that improved regional communication, and an inter-facility, coordinated approach to infection control, could help reduce CDI spread.