287 Emergency Department Telemedicine Consults Consistently Decrease Time to Head Computed Tomography Interpretation in a Multi-Network Stroke Cohort

Abstract

The objectives of this study were to (1) identify the overall impact of telemedicine on emergency department (ED) stroke care measures of time to head computed tomography (CT) interpretation and tissue plasminogen activator (tPA) administration and (2) describe the variation in these process outcomes across EDs and telemedicine networks. A retrospective matched cohort study identified stroke patients in 23 EDs participating in four telemedicine networks between October 2015 and December 2017. Primary exposure was defined as a telemedicine consultation during the ED evaluation. Outcomes included two existing quality measures for stroke care: (1) head CT interpretation within 45 minutes for subjects with a last known well within two hours of ED arrival and (2) for eligible patients, time to tPA administration. To estimate variation in the impact of telemedicine across hospitals and networks, the rate of head CT interpretation completed within 45 minutes of ED arrival was estimated for each hospital. Rates were risk- and reliability- adjusted for each hospital. Estimates were adjusted for primary payer, race, emergency severity index, and ED visit arrival day of week. Of the 932 stroke patients identified, 36.4% (n=338) received telemedicine consults and 259 subjects had a last known well within two hours. Subjects receiving telemedicine were, on average, younger, presenting during non-business hours, and higher initial severity. For subjects with a last known well time within two hours of ED arrival (n=259, 27.9%), the recommended CT Interpretation within 45 minutes was met for 66.8% of subjects. Subjects receiving telemedicine had a higher odds of head CT interpretation compared to those that did not receive telemedicine consults (aOR: 3.03 [95% CI: 1.69 – 5.46], p<0.001). Telemedicine consultation in the ED was not associated with a higher odd of receiving tPA within three hours of last known well compared to those that did not receive telemedicine consults (aOR: 2.13 [95% CI: 0.55 – 8.24], p= 0.271). Variation among hospitals did not significantly differ from average after risk adjustment for case severity and reliability adjustment due to small numbers of cases in some EDs (Figure 1). Telemedicine consultation during the ED encounter decreased time to head CT interpretation among stroke patients, and the rate of adherence with recommended head CT times in telemedicine versus non-telemedicine patients did not differ by hospital. While previous single-center studies have indicated telemedicine improves process outcomes in emergency stroke care, this is the first study of multiple telemedicine networks and suggests the positive effects of telemedicine on stroke quality of care indicators are generalizable to many ED settings and geographic regions.