Interrelation of individual risk and coronary atherosclerosis in coronary heart disease prognosis

Abstract

Probability estimation model for diagnostic of coronary atherosclerosis that needs invasive treatment, had been created for patients with various levels of individual risk. For multifactorial analysis with use of neural networks, 19 parameters were selected, including anamnestic data, instrumental examination results, pre-test CHD probability, and individual level of cardiovascular risk. Significant clinico-instrumental predictors were: stress test results, typical angina, myocardial infarction in anamnesis, ECG signs of left ventriular hypertrophy, pre-test CHD probability. A two-level scheme for coronary atherosclerosis diagnostic was developed. This scheme could be used in out-patient settings (screening, selection of candidates for invasive intervention). Its prognostic accuracy is 1. 5-3 times higher than that of traditional clinical decision making procedure.

Keywords

• coronary heart disease
• individual risk
• coronary atherosclerosis
• neural networks
• prognosis