OBJECTIVES: Telemedicine is increasingly viewed as a tool to provide a wide range of health services with high expectations regarding its cost effectiveness; however a cross sectional analysis of digital health trials found a 27% non-publication rate at 5 years and postulated this was due to either technology failure or negative results. We present lessons drawn from failed telemedicine experiments conducted in the Paris region.

METHODS:

We used a mixed method design combining data analysis and interviews with stakeholders to study eight evaluations (RCT or case-control studies) of telemedicine projects commissioned by the Paris Regional Health Agency between 2013-17:

• Teleconsultation for dependent polymorbid nursing home residents without on-site access to care (2 studies);
• Teleconsultation for autistic children and adolescents and teleconsultation for children and adolescents with multiple handicaps living in institutions;
• Tele-expertise for newborns hospitalized in a neonatal intensive care unit with severe brain disorders;
• Tele-expertise for pathology;
• Medical tele-assistance for intraoperative pathology during a surgical procedure
• Telemonitoring of heart failure patients

RESULTS:

We identified the following reasons for disappointing outcomes: assessment requested too early because payers required information for budgetary decisions; long learning curve, technical problems, diversion of use, insufficient number of inclusions and a lack of adherence prevented the demonstration of successful outcomes of the projects. Telemedicine often generated a significant work overload for the remote specialists and consequently additional human resources were requested. Telemedicine should be considered at least in part as an addition and not a substitute for current care pathways.

CONCLUSIONS:

Evaluation of telemedicine should be undertaken after sufficient uptake to ensure barriers to implementation are overcome and to obtain the sample size necessary for statistical power and reduce the average cost for one telemedicine request. Randomized controlled trials should be encouraged with an appropriate funding and the follow-up period should be extended.