Diagnostic Accuracy in Primary Care E-Visits: Evaluation of a Large Integrated Health Care Delivery System's Experience.

Abstract

To compare diagnostic accuracy between primary care E-Visit and face-to-face (F2F) encounters for low-acuity illnesses. This cross-sectional retrospective analysis of electronic health records in a large not-for-profit integrated delivery system included patients covered by the health care system's employee health plan with an established affiliated physician-patient relationship and an F2F encounter in the past 12 months who had an E-Visit (n=490) or an F2F (n=2201) primary care encounter for a low-acuity illness from July 1, 2015, through December 22, 2016. Patients with a related follow-up visit within 10 days resulting in a revised diagnosis, as determined by 2 physician reviewers, were compared (1) including only the first encounter for each patient and (2) including all encounters more than 10 days apart for included patients. In both analyses, a follow-up visit occurred within 10 days more than 40% of the time in both groups. However, follow-up visits related to the initial diagnosis occurred only 9% to 12% of the time. Only 2.1% to 2.4% of initial diagnoses were identified by both physician reviewers as revised, whereas 3.8% to 5.5% were so identified by at least 1 reviewer. The only significant difference observed between the E-Visit and F2F groups was in the rate of related follow-up visits when only each patient's first encounter was considered, which was higher for E-Visits (12% vs 9%; P=.04). Diagnostic accuracy for low-acuity illnesses in this population was equivalent between E-Visit and F2F encounters.