Computer-unaided diagnosis of acute abdominal pain in an accident and emergency department.

The initial accident and emergency and final in-patient diagnoses of 200 patients with acute abdominal pain made without computer aid or structured data sheet were compared. Sixty-five per cent were correctly diagnosed and 5% had normal laparotomies. These results compare favourably with those obtained using computer aid and structured history-taking forms. It is suggested that the spotlight should be on the training and experience of doctors making the initial diagnosis rather than on computer aid.