In drug management of hypertension and dyslipidaemia for primary prevention of cardiovascular disease, the most recent guidelines advocate an overall risk approach as opposed to previously a risk factor approach.^1,2 However, the evidence, on which both guidelines are based, is essentially the same. What is the basis for the new guidelines? What are the possible implications of the changes in the new guidelines?

In the risk-factor approach, management of blood pressure and cholesterol is dealt with independent of other risk factors. The presence of hypertension and hyperlipidaemia is judged by comparing the levels of blood pressure and blood cholesterol in a person with a threshold that is arbitrarily determined. Elevated blood pressure or blood cholesterol is all that is needed in order to decide whether drug treatment should start or not. In contrast, the new guidelines recommend that drug treatment for primary prevention should be limited to those who are at a sufficiently high risk of cardiovascular disease in addition to elevated blood pressure or cholesterol. This risk is determined by the overall profile of coronary risk factors in a person and can be projected by using the Framingham risk equation or associated risk assessment tables and charts.^2,3 For a long time, drug treatments for primary prevention were a transitional approach which was between the risk factor approach and the overall risk approach. For example, an old man with hypertension was more likely to be treated than a young woman with similar level of blood pressure.

The evolution of the primary prevention strategy reflects a change in the perception of what constitutes a benefit. Different perceptions of risk may lead to different decisions and emphasise the benefit to different people. Two hypothetical examples illustrate the point. Treatment of patients with disease A reduces the risk of death from 5% to 0.1%, representing a relative risk of 0.02 or a relative risk reduction of 98%. Treatment of disease B reduces the risk of death from 50% to 40%, representing a relative risk of 0.8 or a relative risk reduction of 20%. Here the effect of treatment is expressed in relative terms. Treatment of disease A is more effective than treatment of disease B.

The effect of treatment can also be expressed in absolute terms. To continue with the above examples, the difference in risk between the two compared groups or absolute risk reduction will be -4.9% for disease A and -10.0% for disease B. The absolute risk reduction can also be converted into the number needed to treat (NNT) in order to avoid one death event, which is the reciprocal of the absolute risk reduction. The smaller the NNT is, the more effective the intervention is. The NNT is 25 for disease A and 10 for disease B. By comparing the absolute risk reduction, treatment of disease B is more effective than treatment of disease A.

Given limited healthcare resources, decision makers have to prioritise healthcare services to maximise the benefit. In the above simplified example, a higher priority would be given to disease A if we use the relative risk reduction and to disease B if we use the absolute risk reduction. Our conclusion is dependent upon how the treatment effect is analysed and interpreted.⁴ Generally, the relative risk reduction does not discriminate between interventions that have a common effect or a rare effect. In contrast, the absolute risk reduction tends to emphasise more on the benefit of the total population and downgrade prevention measures for rare diseases and low-risk patients of common diseases. Thus, those who seek to maximise the benefit of the total population, such as healthcare planners, will often use the absolute risk reduction in decision making. Those who make decisions for individuals would have to balance the messages from both effect measures. In addition, decision making regarding either a population or an individual will also need to take into account possible harms and the cost-effectiveness or cost-utility of a treatment, which is essentially a function of the cost and the absolute benefit.

Randomised controlled trials show that the relative risk reduction in cardiovascular risk through lowering blood pressure and blood cholesterol varies to a much smaller degree than the absolute risk reduction as the number of other risk factors increases.² The new guidelines, which are based on the absolute risk reduction and targeted at high-risk persons, represent a shift of emphasis in primary prevention from individuals to populations. Doctors who make decisions for individuals rather than populations may feel obliged to treat patients as well when a small risk may be reduced or avoided, particularly when patients are willing to pay for the treatment by themselves. However, in low-risk persons a much larger number of persons would have to be treated in order to prevent one cardiovascular event and a much larger proportion of the treated would bear the costs and experience the adverse effects but not the benefit. Thus, in making clinical decisions doctors should take into account not only the benefit to some treated individuals, but also the interest of many who are treated and experience the adverse effects of treatment but will not benefit and of those who are not treated but may benefit if the resources are allocated differently.