Comparative effects of telemedicine and face-to-face warfarin management: A systematic review and network meta-analysis

Abstract

ObjectiveTo determine the comparative effects of different types of telemedicine on anticoagulation control and adverse outcomes in patients taking warfarin. Data sourcesA systematic search was conducted in PubMed, the Cochrane library, EMBASE, CINAHL, OpenGrey, and DART-Europe from inception to September 2019. Study selectionRandomized controlled trials, quasi-experimental and cohort studies were included if they evaluated the effect of telemedicine in patients using warfarin for 3 months or more. MethodsTwo reviewers extracted and assessed the quality of studies by means of the Cochrane Effective Practice and Organization of Care Group. Network meta-analyses were performed using a random-effects model. Surface under the cumulative ranking curve (SUCRA) methods were used to rank different telemedicine types. ResultsOf the 14,673 articles retrieved, 12 studies involving 11,478 patients met the inclusion criteria. The 5 types of telemedicine categorized were 1) laboratory draw with individual telephonic follow-up by health care (HC) providers (Lab/Tel/HC); 2) laboratory draw with individual telephonic follow-up with periodic, routine routine face-to-face (F2F) visit (Lab/Tel+F2F/HC); 3) laboratory draw with individual and group follow-up via online access (Lab/Online/HC); 4) patient self-test (PST) with telephonic or Web-based follow-up by automated management system (PST/Tel-Online/Auto); and 5) PST with telephonic follow-up by HC provider (PST/Tel/HC). PST/Tel-Online/Auto appears to be superior to F2F for TTR (mean difference 8.78%; 95% CI 0.06%–17.50%) and Lab/Online/HC for in-range INR (odds ratio 2.59; 95% CI 1.04–6.46). The SUCRAs suggested that PST/Tel-Online/Auto was preferred for both TTR and INR, at 84.2% and 93.9%, respectively, whereas Lab/Tel/HC was preferred to prevent major bleeding (74.1%) or thromboembolic event (70.7%). ConclusionAccording to the current evidence of uncertain quality, the best effects on anticoagulation control and adverse outcomes were achieved from different types of telemedicine. Variations in the effects of telemedicine on diverse outcomes should be considered before implementing telemedicine interventions in patients taking warfarin.