[Risk stratification in acute coronary syndromes: an unresolved issue]

Risk stratification in patients with acute coronary syndrome has several objectives: to inform and provide advice to patients and their families; to identify patients at a high risk of death or myocardial infarction whose prognosis can be improved by appropriate treatment; to identify those with a very low risk who do not need invasive studies, thereby avoiding unnecessary cost and the risk associated with these procedures; and to formulate cardiac rehabilitation and secondary prevention measures following an acute event. During the last decade, a number of predictive models, or algorithms, have been developed in an attempt to predict death or myocardial infarction in patients admitted for acute coronary syndrome, with or without ST-segment elevation. The aim was to enable treatment to be adapted to the individual's risk. The best-known models are: GRACE, PREDICT, TIMI, In-TIME, CPP, GUSTO and PURSUIT. Most of these models suffer from limitations that derive from the particular population group from which they were extrapolated or from the limited number of clinical variables used. Consequently, their application to the general unselected population with a high rate of comorbid conditions is not helpful for predicting risk in individual patients. Moreover, these algorithms have not taken into account other important factors that could influence prognosis, such as atherosclerotic disease burden or the degree of vascular inflammation. By taking all these factors into consideration and by employing the most up-to-date statistical techniques, more comprehensive and more clinically applicable models could be devised in the future.