Prediction of lethal or life-threatening cardio-circulatory events in patients who apparently are not at risk : a preliminary retrospective echocardiographic study.

Major cardio-circulatory events, defined as circulatory death, myocardial infarction, unstable angina, or stroke, sometimes occur unexpectedly in patients who apparently have no evident increase in risk (absence of overt heart failure, hypertrophy, uncontrolled or severe hypertension, previous or present myocardial infarction, angina, myocarditis, infectious or any other pericardial, valvular or great vessel disease, heart malformation, significant arrhythmia or conduction disturbances). To investigate whether 2D-guided M-mode echocardiographic variables have predictive value in such patients, a retrospective analysis of 1,965 cases was performed. Twenty-one patients were found who on the day of echocardiographic examination fulfilled the above criteria, but suffered major cardio-circulatory events during the first following year (1 yr group), 12 during the second year (2 yr group), and 16 during the third year (3 yr group). Twenty-eight patients who fulfilled the same criteria, but were followed-up free of major cardio-circulatory events for 935 +/- 144 days constituted the control group. Multivariate analysis of variance (MANOVA) of echocardiographic data was used to select the final set of 11 variables from 30 measurements and calculations which enabled satisfactory discrimination between the four groups (Hotelling T2 = 3.979, Fisher F = 7.596 > Ftab = 1.585). Extension of MANOVA with the leave-one-out method revealed that none of 28 control patients was predicted to be at risk of major cardio-circulatory events in the next year, and only one of 21 patients from the 1 yr group was misdiagnosed as not being at risk. Patients at risk were older, had slightly greater body size (particularly weight), and slightly increased diastolic diameter and volume of the left ventricle. The left ventricular mass, mean wall thickness, and estimated cross-sectional area indexes were also slightly increased. The peak systolic stress was slightly increased and contractility index (BPS/ESVI) was slightly decreased. Our preliminary results suggest that easily obtained echocardiographic measurements and calculations contain clinically useful predictive information.