Non-invasive risk stratification within 48 h of hospital admission in patients with unstable coronary disease.

Abstract

In this study we evaluated the prognostic value of three methods of early risk estimation in patients with unstable coronary disease. The methods evaluated were: clinical risk estimation at hospital admission, continuous ST analysis with computerized vectorcardiography for 24 h and serial measurements of creatinine kinase-MB for 48 h. Twenty-seven (14%) of the 195 patients died or had a non-fatal infarction within one year. Clinical risk evaluation correctly identified a subgroup of patients with low risk but did not otherwise predict outcome. Fifty-six (29%) patients had ST vector magnitude episodes on vectorcardiography, 70 (38%) had three or more episodes of ST change vector magnitude and 74 (38%) had a peak creatinine kinase-MB value of 6 microgram.l-1 or more. The even rate for patients with ST vector magnitude episodes (23%) was significantly higher than for those without (10%; P < 0.05). For patients with and without three or more episodes of ST change vector magnitude the event rate was 23% and 9% respectively (P < 0.05) and for patients with and without creatinine kinase-MB > or = 6 microgram.l-1 the event rate was 23% and 8% respectively (P < 0.01). The positive predictive value of having none, either one or both of the ST or creatinine kinase-MB markers positive was incremental. Continuous vectorcardiographic monitoring of ischaemia in combination with serial creatinine kinase-MB measurement considerably improves risk stratification in unstable coronary disease.