Methods: This single-center, quasi-experimental study assessed appropriateness of Clostridium difficile PCR testing pre- and post-implementation of a prospective, pharmacist-led, pre-authorization process. The antimicrobial stewardship team prospectively reviewed all adult CDI-PCR cases sent to the lab prior to specimen processing twice daily, 7 days a week to assess for clinical appropriateness based on guideline criteria. Bone marrow transplant and pediatric patients were excluded. If a patient lacked appropriate CDI clinical criteria, the provider was contacted to discontinue the PCR. CDI-PCR appropriateness was assessed for all patients with a CDI-PCR during the preceding year as a retrospective, comparative cohort. The primary objective was to assess appropriateness of the CDI-PCR pre- and post-intervention. Secondary objectives included intervention sensitivity, specificity, safety, total CDI-PCRs processed, and protocol adherence.
Results: A total of 714 patients were included (n = 360, pre-intervention; n = 354, post-intervention). There were significantly more hospital-onset CDI cases and antimicrobial use within the past 30 days in the pre-intervention group [(248 vs 133, respectively; p < 0.001) and (277 vs 197, respectively; p <0.0001)]. Appropriateness of the CDI-PCR significantly improved post-intervention [n=138 (38.3%) vs n=209 (59.1%), respectively; p <0.001]. Prospective pharmacist intervention was required for 146 inappropriate CDI-PCRs resulting in 79 discontinued and 66 processed CDI-PCRs (n = 1 positive; n = 65 negative). No patient with a cancelled CDI-PCR required additional testing or escalation of care. When compared to the pre-intervention, the CDI-PCRs with pharmacist intervention demonstrated a significant increase in the sensitivity (64.7% vs 98%; p < 0.0001) and decrease in specificity (66% vs 48.3%; p = 0.015) with an improved NPV (91.9% vs 99.3%; p = 0.035) and PPV (23.9% to 24.6%; p = 0.869).
Conclusion: Implementation of a pharmacist-led prospective CDI-PCR review improved PCR appropriateness and had no adverse clinical consequences although the PPV criteria remain low.
I. Almanaseer, None
J. Watson, None
R. Citronberg, None
L. Kelly, None
A. Whaley, None
S. M. Wieczorkiewicz, None