Radhika Prakash-Asrani, Kayoko Shioda, Chris W Bower, James P Steinberg, Scott K Fridkin, Jesse T Jacob, Lindsey B Gottlieb
Despite an improvement in hand hygiene adherence, we did not find a resultant impact on MDRO+ detection, likely owing to the complex interaction between the two variables and unmeasured confounders.
OBJECTIVE: To assess the association between hand hygiene adherence, as measured by an automated hand hygiene monitoring system (AHHMS), and multidrug-resistant organisms and other pathogens of epidemiologic significance (MDRO+).
DESIGN: Retrospective cohort study.
SETTING: Acute care hospitals within an academic healthcare system.
PARTICIPANTS: Hand hygiene data was collected from all badged staff January 2021 through September 2022. Adult inpatients on non-maternity units during this period were eligible for inclusion.
METHODS: After implementing an AHHMS, we analyzed correlations between monthly unit-level hand hygiene adherence (proportion of room entry-exit opportunities during which hand hygiene solution was dispensed) and hospital-onset MDRO+ rates from all clinical cultures and Clostridioides difficile stool tests. Mixed-effect negative binomial regression models evaluated associations between monthly unit-level hand hygiene adherence and MDRO+ rates while adjusting for the number of clinical tests ordered, unit type, and average number of inpatient COVID-19 cases.
RESULTS: During the study period, there were 22,767,036 hand hygiene opportunities. Mean hand hygiene adherence improved from 41% (January 2021) to 57% (September 2022). 1,847 MDRO+ detections were observed in 752,341 patient days with a median composite MDRO+ rate of 1.7 per 1,000 patient days (IQR: 0-3.6). The mixed-effect negative binomial regression model found a null association between hand hygiene adherence and MDRO+ rates (rate ratio 1.01; 95% confidence interval: 1.00-1.01).
CONCLUSIONS: Despite an improvement in hand hygiene adherence, we did not find a resultant impact on MDRO+ detection, likely owing to the complex interaction between the two variables and unmeasured confounders.