Screening for Clostridium difficile at Admission Reduces Outbreak Risk in the Surgical Unit

Katherine Linsenmeyer, MD, VA Boston Healthcare System

Katherine Linsenmeyer, MD

A study in Clinical Infectious Diseases reports that a ward-based isolate and screening strategy for Clostridium difficile, consisting of perirectal swabbing and analysis by polymerase chain reaction (PCR) for toxin B gene at admission to the surgical care unit, can successfully identify asymptomatic patients at risk progression to symptomatic C. difficile.

Study investigators, led by Katherine Linsenmeyer, MD, of the VA Boston Healthcare System in Massachusetts, found that an autoregressive integrated moving average (ARIMA) model may be valuable for predicting the number of patients needed to screen and isolate, resulting in greater targeted identification and further reductions in outbreak risk.

“This is the first initiative in the United States to address the potential role of screening and isolation of asymptomatic C. difficile carriers in an outbreak setting as an adjunct to standard infection prevention measures,” wrote the study investigators. “We demonstrate a definitive decline in hospital-acquired C. difficile infection rates in the 10-month period following the implementation of this initiative. This decrease is not attributable only to reclassification of carriers, as the reduction in HA-CDI rates persisted even when correcting for cases which were misclassified.”

Patients admitted to the surgical intensive care, surgical step-down unit, or acute care surgical floor at the VA Boston Healthcare System were included in this prospective ward-based C. difficile screening intervention (n = 1250). Investigators screened patients for infection using a perirectal swab and PCR analysis for the toxin B gene. The patients who tested positive but were asymptomatic were placed under contact isolation, and subsequent assessment for symptomatic infection was performed if diarrhea developed.

Throughout the 1 year study period, during which time patinets were followed for 90 days post-discharge, investigators sought to identify whether the screening strategy on admission reduced the rate of symptomatic C. difficile and hospital-acquired C. difficile. Symptomatic infection was defined as ≥3 loose stools within a 24-hour period with no known cause. Additionally, researchers defined hospital-acquired C. difficile as symptomatic infection during hospitalization of ≥2 calendar days following admission. The ARIMA mathematical model was used to estimate the rates of hospital-acquired C. difficile infection per 10,000 bed days of care during the study.

Among the 773 unique patients who had patient-level data available for analysis, a total of 24 (3.1%) patients were found to be asymptomatic carriers of C. difficile. Variables univariately associated with asymptomatic infection included prior surgery-related hospitalization, non-white race, a history of C. difficile infection, and recent antibiotic use. The multivariate analysis found that a history of infection (odds ratio [OR], 8.9; 95% CI 2.5-31.7), as well as recent surgery-related hospitalization (OR, 6.2; 95% CI 1.7-22.3), were the strongest predictors of being an asymptomatic carrier of the infection upon admission.

A greater proportion of patients who were positive asymptomatic carriers of C. difficile at screening progressed to symptomatic infection within 90 days of hospital admission compared with those who were negative (29% vs 1%, respectively; P <.05).

Using data on asymptomatic carriers obtained from the screening test at hospital admission, the investigators calculated rates of hospital-acquired C. difficile infection per 10,000 bed days of care (10.9, 10.2, and 3.0 in 2015, 2016, and 2017, respectively). In the absence of screening data, there was a 25% misclassification of hospital-acquired infection rates in 2016 and a 29% misclassification rate in 2017.

A limitation of the study is the use of a perirectal swab versus a rectal swab protocol, which may have biased the findings toward the null hypothesis “since carriers would go undetected and continue to potentially contribute to transmission and also be counted as hospital-acquired cases if they became symptomatic.”

“More traditional variables such as prior antibiotic exposure, proton pump inhibitor use, or prior history of C. difficile infection were not significantly associated with active C. difficile infection development, likely because they were overpowered by the carrier variable, suggesting that carrier status is an overwhelming factor in C. difficile infection development,” the investigators explained. “It is unclear if other interventions in addition to isolation could be useful in this population as this was not addressed in the current study. Treatment of carriers in older studies failed to show benefit in eradication of disease or reduction of hospital-acquired C. difficile infection.”

REFERENCE

Linsenmeyer K, O'Brien W, Brecher SM, et al. Clostridium Difficile Screening for Colonization During an Outbreak Setting [published online May 26, 2018]. Clin Infect Dis. doi: 10.1093/cid/ciy455.