[Computer-based interpretation of ECG--guiding or misleading?].

Computerised ECG diagnostic programs occasionally provide erratic diagnoses, and false diagnostic suggestions may mislead the physician. We wanted to investigate whether a diagnostic computer program guides or misleads the ECG interpretation in the emergency room. A panel of 20 first-line physicians from the Medical Department at Ullevål Hospital, Oslo, Norway each described sets of ten ECGs, composed from a selection of ten excellent and ten wrong computer interpretations, randomly with or without the print-out of this diagnosis. The presence of correct computer diagnosis resulted in 58% correct conclusions by the physicians, against 30% in the absence of the computers conclusions (p < 0.005). Whether an incorrect computer diagnosis was provided or not, did not significantly influence the physicians' conclusions. Among the physicians, the best performing third benefitted most from the presence of a good computer interpretation, whereas the poor performers did not even recognise the help provided. Computer-based ECG diagnoses seem to be helpful to emergency ward physicians, but a certain level of ECG experience is required to utilise the program.