Association of eConsult Implementation with Access to Specialist Care in a Large Urban Safety-Net System

Abstract

Importance: Accessing specialty care continues to be a persistent problem for patients who use safety-net health systems. To address this access barrier, hospital systems have begun to implement electronic referral systems using eConsults, which allow clinicians to submit referral requests to specialty clinics electronically and enable specialty reviewers to resolve referrals, if appropriate, through electronic dialogue without an in-person visit. Objective: Measure the effect of implementing an eConsult program on access to specialty care. Design, Setting, and Participants: Using an interrupted time series design with data from 2016 to 2020, this study analyzed 50260 referral requests submitted during the year before and the year after eConsult implementation at 19 New York City Health + Hospitals (NYC H+H) specialty clinics that spanned 7 NYC H+H hospital facilities and 6 unique specialties. Exposures: Referral request was submitted to a specialty clinic in the year following eConsult implementation. Main Outcomes and Measures: Main outcomes included the fraction of referral requests resolved without an in-person visit following eConsult implementation; and, among requests triaged to have an in-person visit, the fraction of referrals with a successfully scheduled appointment, mean wait time to a specialty appointment, and the fraction of referral requests with a completed specialty visit. Changes associated with eConsult implementation were estimated using multivariate linear regression adjusting for patient age, gender, and specialty clinic fixed effects. Results: Across 19 NYC H+H specialty clinics, 26731 referral requests were submitted in the year before and 23529 referrals were submitted in the year after eConsult implementation. Following eConsult implementation, 13% of all requests were resolved electronically. Among requests requiring a follow-up visit, the fraction with an appointment successfully scheduled increased by 15.8%, from 66.5% to 82.3% (P