Nonmodal Clinical Decision Support and Antimicrobial Restriction Effects on Rates of Fluoroquinolone Use in Uncomplicated Infections.

Abstract

Medication alert overrides remain persistently high over the past decade, influenced by factors such as "alert fatigue" and lack of provider acceptance. We compared the aggregate rate of fluoroquinolone (FQ) prescribing for the treatment of acute sinusitis, acute bronchitis, and uncomplicated urinary tract infections (UTIs) in adult inpatients prior to (historical control group) and after (prospective intervention group) implementation of a program requiring indication when ordering FQ antibiotics in combination with a nonmodal best-practice alert regarding the latest U.S. Food and Drug Administration (FDA) recommendations. We then compared rates of prescribing among provider type, severity of infection, and patient age. Qualified orders were defined as new FQ orders for acute sinusitis, acute bronchitis, and uncomplicated UTI for adult inpatients between July 2016 through September 2016 (control) or November 2016 through January 2017 (intervention). The primary endpoint was a provider-initiated FQ order for a target indication. Secondary endpoints included FQ orders by provider type and patient age. Rates of FQ use among the target indications were compared between groups by chi-square test of independence with Yates' correction in the analysis of the primary endpoint and Fisher's exact test for secondary endpoints. FQ prescribing for acute bronchitis, and uncomplicated UTI occurred at a rate of 86/350 (24.6%) and 62/394 (15.7%) in the control and experimental groups, respectively (p = 0.0035). No patients receiving FQ qualified for a diagnosis of acute sinusitis. A program combining FQ restriction in combination with nonmodal messaging may have decreased the rate of prescribing for acute bronchitis and uncomplicated UTI, although the contributions of each individual element could not be rigorously assessed.