Evaluating Facility Characteristics and Connectivity Metrics as Predictors of Clostridioides difficile Rates in Nursing Homes, Atlanta, GA. (October 2020)

• Record Type: Journal Article
• Authors: Sefton, Samantha; Goodenough, Dana; Saiyed, Sahebi; Fridkin, Scott
• Is Part Of: Infection control and hospital epidemiology. Volume 41(2020)Supplement 1
• Page Start: s35
• Page End: s36
• Publication Date: 2020-10
• Subjects: Nosocomial infections -- Epidemiology -- Periodicals; Health facilities -- Sanitation -- Periodicals; Hospital buildings -- Sanitation -- Periodicals; Cross Infection -- Periodicals; Epidemiology -- Periodicals; Hospitals -- Periodicals; Infection Control -- Periodicals; 614.44
• DOI: 10.1017/ice.2020.515
• Languages: English
• ISSNs: 0899-823X

Abstract

Abstract : Background: Nursing home (NH) residents are at high risk for Clostridioides difficile infection (CDI) due to older age, frequent antibiotic exposure, and previous healthcare exposure. Incidence of CDI attributed to NHs is not well established, but it is hypothesized to be related to the magnitude of transfers. We evaluated the relationship between NH CDI incidence and facility characteristics to explain variability in rates in Atlanta, Georgia. Methods: Incident C. difficile cases from 2016 to 2018 were identified through the Georgia Emerging Infections Program (funded by the Centers for Disease Control and Prevention), which conducts active population-based surveillance in the 8-county metro Atlanta area. An incident case was defined as an NH resident with a toxin-positive stool specimen (without a positive test within 8 weeks). Sampled (1 to 3 on age and gender) incident cases were attributed to a NH if a patient was an NH resident within 4 days of specimen collection. Facility characteristics (beds, resident days, admissions, and average length of stay [ALOS]) were obtained from NH cost reports, and facility-specific connectivity metrics were calculated (indegree and betweenness) from 2016 Medicare claims data. Case counts were aggregated to estimate yearly incidence and correlated with facility characteristics and location within the healthcare network using the Spearman correlation. A negative binomial model was used to assess residual variability in NH CDI incidence. Results: In total, 386 incident CDI cases were attributed to 64 NHs (range, 0–27). Approximately half (54.7%) resided in the NH at the time of specimen collection; however, 33.7% were in inpatient units (≤4 days of admission), and 10.9% were in an emergency room (ER). The frequency of NH CDI cases correlated strongly with admissions (r = .70; P < .01), inversely with ALOS (r = −0.53; P < .01), and moderately with resident days (r = .38; P < .01). After accounting for admissions, incidence (per 1, 000 admissions) still varied (Fig. 1) (median 14; range, 0–34). The inverse association with ALOS decreased and incidence no longer correlated with the remaining facility characteristics or location within the healthcare transfer network ( P > .05, all comparisons). However, there was residual correlation with connectivity metrics (indegree r = 0.26; P = .04). Conclusions: Our data suggest that attributing CDI to NHs requires the inclusion of hospital and ER-based specimen collection. NH CDI incidence appears highest among facilities with a low ALOS and a high number of admissions; incidence rates calculated per 1, 000 admissions may best account for infection risk inherent early in a resident's stay. Residual variability attributed to connectivity to the healthcare network was of borderline significance and should be further explored in the NH setting. Funding: None Disclosures: Scott Fridkin, consulting fee, vaccine industry (spouse)