Effectiveness of interventions targeting stewardship of Clostridium difficile testing

From the 2018 HVPAA National Conference

Natalie Kappus (NYU Langone Health), Alexis Reed (NYU Langone Health), Preston Kramer (NYU Langone Health), Stephanie Sterling (NYU Langone Health), Sarah Hochman (NYU Langone Health)

Background

Clostridium difficile infection (CDI) is a common healthcare-associated infection with significant morbidity and mortality. Diagnosis causes increased length of stay and costs from additional medical and surgical management. Because PCR-based C. difficile testing does not differentiate between colonization and disease, it is crucial to test only when appropriate.

Objectives

To reduce the number of inpatients tested and treated for CDI who do not have significant diarrhea or who have an alternate cause for diarrhea.

Methods

We reviewed 100 consecutive C. difficile PCR orders spanning two hospitals of NYU Langone Health to assess appropriateness of ordering practices and identify targets for change. We defined appropriate orders as meeting the following criteria: patient had ≥3 liquid bowel movements within a 24-hour period; had not recently received laxatives, oral contrast or new enteral feeds within the preceding 24 hours; and had no recent C. difficile testing (negative PCR within previous 7 days or positive PCR within 30 days). Interventions included implementation of health system-wide C. difficile testing guidelines, face-to-face education of licensed providers regarding new guidelines and just-in-time education in the form of alerts embedded in the electronic medical record (EMR) C. difficile PCR order. Post-intervention analysis included prospective auditing of provider responses to PCR order alerts and chart review of 100 C. difficile PCR orders originating from the two hospitals. We assessed health system-wide hospital-onset CDI rates before and after the intervention bundle as well as CDI-related mortality.

Results

Pre-intervention, 78% of C. difficile PCRs from the Brooklyn campus did not meet testing criteria, which fell to 66% post-intervention. On the Manhattan campus, 44% of C. difficile PCRs did not meet testing criteria prior to the intervention, compared with 28% of orders after the intervention. The majority of orders came from Medicine and the Emergency Department. The most common factors making orders not appropriate on both campuses were receipt of laxatives or <3 episodes of diarrhea. Half as many C. difficile PCRs were ordered post-intervention across both campuses. EMR alerts embedded within the PCR order resulted in provider cancellation 40% of the time. The most common cause for an alert to be triggered was concomitant presence of laxatives. Hospital-onset CDI rates for both campuses combined fell from 0.75 to 0.36 cases per 1000 patient days. There were no deaths attributable to CDI after the intervention.

Conclusion

The majority of tests performed to diagnose CDI in our health system were ordered inappropriately. Health system interventions that incorporated guidelines, face-to-face education and EMR-embedded decision support resulted in an overall reduction of the number of C. difficile PCRs ordered and an increase in the proportion of tests performed for appropriate indications. There were no mortalities associated with delayed CDI diagnosis.

Implications for the Patient

Guidelines focused on testing stewardship for C. difficile can reduce inappropriate diagnosis, treatment and isolation of patients for CDI by decreasing testing of asymptomatically colonized patients, and do not result in CDI-associated mortality from delayed diagnosis.