The Antibiotic Likelihood Index: A New Measure of Antimicrobial Stewardship for Pediatric Primary Care Based on Reason for Visit

Abstract

Abstract Background Quality metrics for antibiotic prescribing by pediatricians are limited. We sought to define a novel measure that assesses clinicians’ overall antibiotic prescribing. Methods Using electronic health record (EHR) data from 2018 to 2019 for children 3 months to 17 years of age from 53 practices within a large pediatric network, we grouped encounters into Reason for Visit categories using the classification system of the National Ambulatory Medical Care Survey and analyzed the proportion of encounters with an antibiotic prescription. Categories were sorted according to the attributable proportion of encounters with an antibiotic prescribed. The proposed metric—the Antibiotic Likelihood Index (ALI)—was defined as the proportion of encounters with an antibiotic prescribed among categories that accounted for >80% of all encounters with an antibiotic prescribed. The ALI was calculated for the entire network and for individual prescribers, and the distribution among prescribers was described. Results Six Reason for Visit categories—cough, ear complaints, fever, sore throat, rash, and congestion/upper respiratory infection—accounted for 82.4% of all antibiotics prescribed. Among the 222 682 encounters for the top 6 categories combined, 67 368 (30.3%) had an antibiotic prescribed, defined as the ALI for the entire sample. The index among individual prescribers ranged from 7.5% to 57.2% (interquartile range 24.3% to 34.9%). The correlation for individual prescribers between 2018 and 2019 was high (R2 = 0.80). Conclusions The ALI, a proposed new metric of pediatric antibiotic prescribing, can be readily calculated from EHR data and captures the range of antibiotic prescribing among pediatricians for common clinical scenarios.