Correlation of seasonal influenza rates with hospital-acquired Clostridioides difficile rates: Implications for research and interventions for prevention

Abstract

Introduction: Hospital-acquired Clostridioides difficile infection (HA-CDI) rate variability can create challenges for assessing interventions that might improve rates. We therefore sought to assess whether seasonal trends in HA-CDI could be related to influenza rates by examining the sensitivity and applicability of statistical methods that identify seasonal trend associations. Methods: We assessed HA-CDI rates per 10,000 patient-days and influenza hospitalizations per 1,000 admissions in four acute care facilities from January 2016 to December 2018. Four statistical approaches were used in R (4.5.1) including (1) autoregressive integrated moving average (ARIMA); (2) cross-correlation to assess if dependencies lagged over time; (3) Poisson harmonic regression to model cyclical trends; and (4) Poisson harmonic regression models including influenza rates. Results: ARIMA approaches did not detect seasonality in HA-CDI rates, but found strong seasonality in the influenza rates, whereas cross-correlation analysis found evidence of correlation between the series at a lag of zero (R = 0.41, 95% CI 0.08 to 0.65) and provided an indication of a seasonal relationship. Poisson harmonic regression for trends suggested that influenza rates predicted CDI rates ( p < 0.01), whilst Poisson harmonic regression identified evidence of seasonality in HA-CDI rates ( p < 0.05) and influenza rates ( p < 0.001). Conclusions: Poisson harmonic regression provided a sensitive means of identifying seasonality in HA-CDI rates with the seasonality effect being accounted for by influenza admission rates, which may be mediated by inappropriate antibiotic prescriptions. This observation may be useful in interpreting future interventions to reduce HA-CDI.