Authors analysed the possibilities of in-hospital prognosis made of patients with acute myocardial infarction by using information obtained during admission. Eighty eight patients with transmural myocardial infarction (not older than 70 years and the prehospital delay shorter than 24 hours) were analysed. On the basis of the hospital events they were divided into 3 classes: (a) uneventful (16), (b) complicated (55) and (c) lethal (17). The initial 26 data available in the first hour were analysed by PRIMA pattern recognition method adapted to ROSY--80B microcomputer. 3 days later the 3 class distances of the initial PRIMA analysis together with the 21 new data of the remaining 79 patients were analysed again. The third analysis (66 patients) was made between the 4th and 6th days after a successful mobilization. The last analysis (44 patients) was performed just before the discharge of the patients, who were able to carry out the low-level and submaximal ergometric tests. The average effectiveness of the method gradually improved during hospitalization from 80% to 91%. The recognition ability of each class respectively was 57% for (a) uneventful, 82% for (b) complicated, and 95% for (c) lethal initially, and it has improved finally: 82% for uneventful, 94% for complicated cases.
|Pages (from-to)||683-686, 689|
|Publication status||Published - Apr 1 1990|
ASJC Scopus subject areas