Data-Driven Interventions for Reducing C. Difficile Incidence

Using AI model predictions to identify patients most at risk for C. Diff infection (CDI), the team used a pre-post study design to assess the effect of a bundle intervention (two BPAs) in reducing CDI incidence at Michigan Medicine. 

The team conducted a Quality Improvement Study in collaboration with the Pharmacy team at Michigan Medicine and with ten MM hospital units with high rates of CDI. The primary outcome in this project was CDI incidence rate. The secondary outcomes were antimicrobial use and qualitative assessments of the bundle intervention.

Publications

JAMA Network

Oxford Academic

Previous
Previous

Hospital Readmissions Risk Prediction and Prevention (HARPP)

Next
Next

PICTURE: Pediatric General Floor Deterioration Prediction