How Rule-Based AI Helped One Hospital Cut C. diff Infections by 33%

"Almost 80% of our C. diff infections were because of late testing. Nurses are having a hard time identifying patients with early onset symptoms. They're supposed to test between hospital day one, two, and three – but sometimes they document late, identify late, test late."

This is how Peggy Leung, Infection Preventionist at PIH Health, describes the challenge her team faced in 2023: rising C. diff rates driven by missed testing windows. If a patient isn't tested within the first three days, a community-onset infection becomes classified as hospital-onset. The issue wasn't lack of effort, it was the reality of nursing workflows where documentation happens inconsistently, and critical symptoms get flagged too late.

The solution was rule-based AI that detects early symptoms, generates actionable reports, and prompts charge nurses to remind staff when testing windows are closing. The result: PIH Health reduced their C. diff infection rate from 1.077 to 0.722 – dropping below the national benchmark and preventing outbreaks before they start.

Get Leung's perspective on improving outcomes through analytics and why hospitals must trust their data to prevent disease.