A study of diagnostic accuracy in suspected acute appendicitis.

A prospective study was conducted of 276 patients with suspected acute appendicitis. Two hundred and eighteen patients had operations and 165 had acute appendicitis. Data from a previous retrospective study were used to develop an eight-item statistical program for a pocket calculator and this was made available to half the surgical registrars as an additional aid to clinical decision making. The other registrars involved in the study used only their clinical skills. There was no difference in diagnostic accuracy or frequency of operations or complications between patients managed by the two groups. Forty items of clinical data on each patient were collected prospectively. These were subsequently analysed on a computer program for the diagnosis of the acute abdomen which was developed in the United Kingdom. The accuracy of diagnosis of the clinicians was superior to the accuracy of the computer diagnosis.