Using Power BI to Inform Clostridioides difficile Ordering Practices at an Acute Care Hospital in Central Florida

Abstract

<h3>Background</h3> Positive Clostridioides difficile (C. diff) laboratory results likely overestimate the true incidence of C. diff cases. In 2015, a multi-campus intervention that reduced hospital-onset C. diff consisted of diverting automatic reflex polymerase chain reaction test (PCR) testing for negative enzyme immunoassay toxin tests (EIA) to requiring a separate order. Rates initially decreased, but have since increased, suggesting a need to improve C. diff surveillance and track ordering patterns of providers and C. diff trends across our campuses. <h3>Methods</h3> A C. diff dashboard was developed using Microsoft Power BI, consisting of information on test type, test order counts, and rates per 100 patient (pt) days by month/year and campus/unit. Data analysis focused on one facility comprising 237 beds. A root-cause database in Microsoft Access was used to track risk factors that may have contributed to C. diff infection. <h3>Results</h3> From January 2017 to November 2019, C. diff test (EIA and PCR) order rate per 100 pt days was 1.24 in 2017, 1.42 in 2018, and 1.09 in 2019, correlating with C. diff rates per 100 pt days, 1.76 in 2017, 3.80 in 2018, and 1.42 in 2019 respectively (r=0.94). PCR ordering rates increased throughout the years. Review of risk factors in the root-cause database did not appear to significantly contribute to the increase in C. diff rate seen in 2018. <h3>Conclusions</h3> C. diff testing rates highly correlate with positive C. diff results. The construction of a C.diff dashboard and the root-cause database allows us to analyze diagnostic ordering trends in real time in order to identify gaps to reduce unnecessary testing.This can be accomplished through feedback, education, and changes in processes.