Accuracy of paediatric echocardiographic transmission via telemedicine

We evaluated the accuracy of realtime echocardiography studies conducted via telemedicine (at 384 kbit/s) and prerecorded video studies, by comparing the results with subsequent in-person echocardiography examination performed at follow-up. Between January 2002 and December 2004, there were 769 paediatric echo studies of patients aged one day to 19 years by telemedicine. There were three cases (0.4%) in which the study could not be successfully transmitted in realtime due to ISDN line or equipment failure. These were recorded, then transmitted and reviewed within 24 h. A normal heart was demonstrated in 272 studies (35%). A non-urgent congenital heart defect (CHD) was detected in 311 studies (41%). A critical cardiac abnormality was identified in 25 studies (3% of the total). Sixty-seven studies were repeated in person at the tertiary centre. All but one of the follow-up echocardiograms confirmed our initial telemedicine diagnosis (99% sensitivity). There were 655 videorecorded studies delivered to the tertiary centre between February 2002 and November 2003. Thirty-two patients had a repeated echocardiogram performed there. There were either discrepancies or lack of diagnostic clarity in 12 studies (38%) compared with on-site echo evaluations. This suggests that echocardiography evaluation using videorecordings may be less accurate than via realtime telemedicine.