C. diff Testing Stewardship Liesel Nelson, MLS Infection Prevention St. Luke’s Hospital
We are not the only ones struggling…
Challenges Correctly diagnose and appropriately treat those with true C. diff disease Navigate testing and diagnosis in the setting of the current reporting case definition (or lack thereof) Encourage clinical evaluation of pre-test probability of disease
Stewardship Interventions Best Practice Advisories (BPAs) Real-time provider feedback Laboratory rejection criteria
BPAs and Direct Feedback: Let’s talk laxatives!
Results
Conclusions Following implementation of a BPA demonstrating specific instances and receipt of a laxative/diarrheal agent in our academic institution, 29% of intended orders for Clostridium difficile NAAT testing were cancelled. Conversely, MOST BPAs were disregarded. But still… The BPA was not associated with a delay in Clostridium difficile testing
Conclusions Despite the implementation of the clinical decision support alert, a majority were overridden (75.2%). Common reasons for C. difficile testing in patients receiving laxatives were related to 1) the clinicians’ concern for the risk of CDI, 2) concern for severe CDI, and 3) instructions by an attending physician to test. Frequently, clinicians initially accepted the alert but then reinitiated ordering C. difficile testing. The rate of positive C. difficile test results was the same between the total study population and the subset of patients whose clinicians were called (NS). Real-time phone calls were highly effective in changing clinicians’ behaviors regarding accepting the clinical decision support alert. If implemented longitudinally, real-time phone calls could have a significant impact in reducing unnecessary C. difficile testing.
In May 2017, the hospital infection prevention team began giving direct feedback and education to physician teams in the form of emails and live conversations whenever there was a positive HO C. difficile test, especially following laxative use within 72 hours. What kind of feedback given in instances with no laxative use?? In March 2018, we implemented an EHR alert for providers ordering C. difficile tests on patients who had received laxatives within the past 72 hours.
Results
Conclusions There was a 43% decrease in HOCDI after implementation of direct feedback and a 57% decrease from baseline with both EHR alert and direct feedback, without an increase in community onset CDI. There was a 28% decrease in HO CDI within 72 hours of laxative use with direct feedback, but no improvement when the EHR alert was added. Overall, education of staff regarding CDI lab identified module criteria and importance of timely testing of diarrhea on admission is crucial. Criteria for proper stool specimens (consistency and patient age) continually need to be enforced.
All stool samples sent to lab for C All stool samples sent to lab for C. diff testing (CDT) are held pending IP review. The IP review tool provides a guide for EMR review of contributing factors in diarrhea, including patient history, medications (e.g., laxatives, stool softeners), nutrition (e.g., tube feeds), symptoms (abdominal pain), and labs (e.g., serum creatinine, WBC count). Inappropriate CDT was defined as test of patients receiving pro-motility treatment, without signs of infection.
Results
Disease vs. Colonization
Between August 30, 2015 and November 30, 2018 47,048 C. diff screens were performed on admission PCR testing on perirectal swabs
Results 2,010 (4.2%) were positive LabID events per 10,000 patient days decreased from 12.2 to 10.2 C. difficile colonization varied significantly based on type of unit Those who screened positive were significantly older than those who screened negative (mean age 60.1 vs. 58.1 years, p<0.005)
Definitive CDAD = 1 Probable CDAD = 94 Possible CDAD = 62 Unlikely CDAD = 85 Asymptomatic Colonization = 111
Patients with healthcare-associated Clostridium difficile infection are often colonized with the infecting strain on admission by whole genome sequence Gonzalez-Orta M1,2, Saldana C1,2, Ng-Wong Y1, Cadnum J1, Jencson A1, Jinadatha C3,4, Donskey C5,6. Conclusion: 4 of the 13 (31%) episodes of healthcare-associated CDI occurred in patients colonized with a genetically related strain on admission. Standard control measures focused on preventing transmission will not be effective in reducing infections in such patients. Antimicrobial stewardship and CDI test stewardship interventions are needed to reduce the potential for diagnosis of healthcare-associated CDI in asymptomatically colonized patients.
Other themes Reducing antimicrobial days for high-risk agents Toxin vs. PCR testing and provider interpretation Laboratory rejection criteria
What can learn from all of this? Reinforce expectation of laboratory rejection criteria Patient <1 yo or formed stool Consider implementing laxative-related BPA in Epic Not perfect, but some impact Roll out provider-to-provider education Targeted vs. large scale? Set the stage for IP/nursing/lab-led feedback to providers Discuss methods for real-time feedback to providers Must be scalable Targeted: after dismissal of BPA?
Questions?