Clinical perspective on risk stratification.

Abstract

Surveys show that subjective risk assessment or risk stratification is often widely inaccurate. Objective data from large observational studies or from clinical trials identifies a persons absolute risk of an event in a given time in order to assess the risk/benefit ratio of a given treatment. In general, the higher the risk the better the risk/benefit ratio. Relative risk reduction by a given treatment is often similar across subgroups divided by sex, age, blood pressure etc.; however if the absolute risk is low it may not e worth taking a treatment with serious side effects (e.g., cerebral haemorrhage). Risk stratification is also important in assessing the cost effectiveness of treatment (e.g,. cholesterol reduction by statins for different groups of the population). Inappropriate surrogate end points should be avoided in cost benefit analysis (e.g., suppression of ventricular ectopic beats by antiarrhythmic drugs). A Bayesian approach should be adopted.