Utilization, Safety, and Technical Performance of a Telemedicine System for Prehospital Emergency Care: Observational Study (Preprint)
BACKGROUND As a consequence of increasing emergency medical service (EMS) missions requiring an EMS physician on site, we had implemented a unique prehospital telemedical emergency service as a new structural component to the conventional physician-based EMS in Germany. OBJECTIVE We sought to assess the utilization, safety, and technical performance of this telemedical emergency service. METHODS We conducted a retrospective analysis of all primary emergency missions with telemedical consultation of an EMS physician in the City of Aachen (250,000 inhabitants) during the first 3 operational years of our tele-EMS system. Main outcome measures were the number of teleconsultations, number of complications, and number of transmission malfunctions during teleconsultations. RESULTS The data of 6265 patients were analyzed. The number of teleconsultations increased during the run-in period of four quarters toward full routine operation from 152 to 420 missions per quarter. When fully operational, around the clock, and providing teleconsultations to 11 mobile ambulances, the number of teleconsultations further increased by 25.9 per quarter (95% CI 9.1-42.6; <italic>P</italic>=.009). Only 6 of 6265 patients (0.10%; 95% CI 0.04%-0.21%) experienced adverse events, all of them not inherent in the system of teleconsultations. Technical malfunctions of single transmission components occurred from as low as 0.3% (95% CI 0.2%-0.5%) during two-way voice communications to as high as 1.9% (95% CI 1.6%-2.3%) during real-time vital data transmissions. Complete system failures occurred in only 0.3% (95% CI 0.2%-0.6%) of all teleconsultations. CONCLUSIONS The Aachen prehospital EMS is a frequently used, safe, and technically reliable system to provide medical care for emergency patients without an EMS physician physically present. Noninferiority of the tele-EMS physician compared with an on-site EMS physician needs to be demonstrated in a randomized trial.