A Multimodal Intervention to Reduce C. difficile Infections and Stool Testing

Abstract

Background and objectives: The introduction of multiplex gastrointestinal panels at our institution resulted in increased Clostridioides difficile (C. difficile) detection and stool test utilization. We aimed to reduce hospital-onset C. difficile infections (HO-CDIs), C. difficile detection, and overall stool testing by 20% within 1 year.

Methods: We conducted a quality improvement project from 2018 to 2020 at a large children's hospital. Interventions included development of a C. difficile testing and treatment clinical care pathway, new options for gastrointestinal panel testing with or without C. difficile (results were suppressed if not ordered), clinical decision support tool to restrict testing, and targeted prevention efforts. Outcomes included the rate of HO-CDI (primary), C. difficile detection, and overall stool testing. All measures were evaluated monthly among hospitalized children per 10 000 patient-days (PDs) using statistical process-control charts. For balancing measures, we tracked suppressed C. difficile results that were released during real-time monitoring because of concern for true infection and C. difficile-related adverse events.

Results: HO-CDI decreased by 55%, from 11 to 5 per 10 000 PDs. C. difficile detection decreased by 44%, from 18 to 10 per 10 000 PDs, and overall test utilization decreased by 29%, from 99 to 70 per 10 000 PDs. The decrease in stool tests resulted in annual savings of $55 649. Only 2.3% of initially suppressed positive C. difficile results were released, and no patients had adverse events.

Conclusions: Diagnostic stewardship strategies, coupled with an evidence-based clinical care pathway, can be used to decrease C. difficile and improve overall test utilization.

Figure 1.. Key Driver Diagram

Figure 2.. Rate of Hospital-Wide Hospital-Onset C. difficile Infections (HO-CDI) Per 10,000 Patient-Days (PDs).

In this U-chart, the blue line demonstrates the rate per month, red line represents the mean, and the dashed lines represent the upper and lower control limits. There was a shift in April 2019 based on having a run of eight points in a row below the mean centerline. In April 2020, the HO-CDI rate was above the upper control limit; we evaluated causes of this outlier - it was thought to potentially be related to the beginning of the pandemic and resulting disruption in prevention and testing protocols as well as changes in clinician behavior around testing for infectious diseases.

Figure 3.. C. difficile Testing Rates per 10,000 Patient-Days (PDs)

In this U-chart, the blue line demonstrates the rate per month, red line represents the mean, and the dashed lines represent the upper and lower control limits. There was a shift in February 2019 based on having a run of more than eight points in a row below the mean centerline. There was one point (March 2019) that was above the newly shifted upper control limit; we believe this is because the decrease C. difficile testing was not fully in control until EMR testing changes in April 2019.

Figure 4.. Stool Tests Performed Per 10,000 Patient-Days (PDs)

In this U-chart, the blue line demonstrates the rate per month, red line represents the mean, and the dashed lines represent the upper and lower control limits. There was a shift in February 2019 based on having a run of more than eight points in a row below the mean centerline.