Key points are not available for this paper at this time.
A random sample of all patients with a registered diagnosis of acute myocardial infarction (AMI) was selected from the hospitalisation register of Stockholm County in 1973. The sample consisted of 269 patients. The medical records of these patients were scrutinized concerning the fulfilment of criteria of a diagnosis of AMI. The criteria of AMI chosen were those established in a Sweden cooperative study (6). 33 patients did not fulfil the diagnosis of AMI, corresponding to an estimated proportion of 10%. In 4% this was due to more technical reasons such as the use of incorrect code numbers, and in 6% the diagnosis had been made without the criteria being fulfilled.
Ahlbom et al. (Fri,) studied this question.