To evaluate changing Clostridioides difficile infection (CDI) testing among inpatients with indeterminate enzyme immunoassay (EIA) results (antigen+/toxin−) from reflexive polymerase chain reaction (PCR) testing to clinician-ordered PCR testing.

Multicenter, before-and-after, quasi-experimental study.

Four large urban tertiary-care hospitals.

We evaluated two 6-month periods before and after an intervention. The primary study outcome was the change in the number of CDI diagnoses between periods. Secondary outcomes included the number of PCR tests performed, adverse events, and healthcare cost savings.

In total, 500 EIA-indeterminate C. difficile test results were evaluated: 281 before the intervention and 219 thereafter. CDI was diagnosed by PCR among EIA-indeterminate cases in 182 in the preintervention period versus 94 patients in the postintervention period (48% reduction; P < .01). PCR testing was performed in 99.6% of indeterminate cases (280 of 281; 1 not performed due to an inhibitor) in the preintervention period versus 66% (144 of 219) in the postintervention period (34% reduction; P < .01). We observed no differences between study periods in 30-day all-cause (P = .96), GI-related (P = .93), or C. difficile (P = .47) readmissions, nor in 30-day C. difficile infections (P > .99). No patient without a PCR test in the postintervention period and not treated was later diagnosed with CDI. Each reflexive PCR test not performed led to a cost savings of $4,498 per patient.

Applying diagnostic stewardship to C. difficile PCR testing in the inpatient setting led to significant reductions in both testing and cases. Changing the C. difficile PCR testing algorithm for EIA-indeterminate cases from reflexive to clinician-required ordering resulted in valuable cost savings without associated adverse events.