Teaching physicians about different measures of risk reduction may alter their treatment preference.

We explored during a postgraduate workshop whether basic teaching about absolute and relative effect measures changed physicians' perceptions of the benefit to be derived from modifying particular cardiovascular risk factors. Before and after instruction physicians were asked about the priority they would give to interventions to reduce four risk factors of coronary heart disease in two male patients, aged 35 and 65 years with multiple risk factors. They were given information about the relative risk (RR), absolute risk reduction (ARR) and the number of patients who need to be treated (NNT) to prevent one event associated with the modification of each risk factor. Ratings of 48 of the 67 participating physicians (71.6%) were evaluated. About half did not change their choices regarding the benefit from a particular intervention. Among those who changed, the new choice was in favor of the patient with the higher ARR for three risk factors (hypertension, p = 0.01; smoking, p = 0.002; non-insulin-dependent diabetes, p = 0.05) but not the fourth (left ventricular hypertrophy, p = 0.82). Teaching basic principles of clinical epidemiology to physicians can have an impact on their perception of treatment effects. However, this will not suffice in itself to guarantee that this new knowledge will become part of their clinical practice.