Implementation of guidelines for the prevention of cardiovascular diseases in Europe: Problems and challenges. Is the SCORE prediction system useful, or does it do more harm than good?

Primum nil nocere

There has been an ongoing discussion in European countries regarding the primary prevention of cardiovascular disease. Clinicians in the front line—especially general practitioners—argue that thresholds for risk intervention should be higher and treatment goals less ambitious than those advocated by cardiologists and in various guidelines [1]. Geoffrey Rose's well-known statement that most cardiovascular deaths occur among people with low or moderate risk, because they are so numerous, is a strong argument for the population strategy [2]. On the other hand, Rose's arguments have been questioned in a recent study from Canada, which finds that applying a multiple-factor high-risk strategy is more effective in preventing cardiovascular deaths [3].

The population strategy for risk prevention is the responsibility of societies. Doctors deal with individuals, and our first priority is, and should be, the sick and suffering patient. The prevention of possible future disease must come second, and for clinicians the high-risk strategy is the only possible risk intervention strategy.

It is time for a shift in focus: we need to inform and educate general practitioners in the importance of preventing future cardiovascular disease in the proportion of the population with the highest risk level. General practitioners need practical guidelines and working tools in order to give priority to the right part of their patient population.

So where are we?

The SCORE prediction system for cardiovascular death, presented by Graham in this issue of European Journal of General Practice, is an important achievement. Containing data from 12 European countries, it is based on a larger population data set than the Framingham algorithm. It was adopted in the Third Joint European Societies’ Guidelines published in 2003, in which WONCA took part as a representative for GPs [4].

The SCORE algorithm and the European guidelines strongly focus on multiple-factor risk-factor prediction: a doctor should not only look at blood pressure or cholesterol level, but include smoking habits, obesity, physical activity and genetic predisposition. However, many GPs feel that intervention thresholds and treatment goals are disproportionate, and have argued against the European guidelines. Based on this, in 2004, the WONCA European Council decided to withdraw its support for the 2003 guidelines.

Getz et al. [5] have shown that implementation of the SCORE prediction system, as used in the guidelines, would label the vast majority of the Norwegian population as being at high risk. There are two main reasons for this. First, the intervention threshold is set to ≥5% total cardiovascular death risk. Second, the risk for younger patients is extrapolated to age 60 years. Another reason may be that the 2003 European guidelines do not take into account patients’ own assessment of health status, which is an important risk factor. The overestimation of risk by the SCORE algorithm is serious, and, if not corrected in later versions, will undermine the good intentions behind the guidelines. We need modelling studies where the risk algorithm and cut-off points for interventions are applied to populations representative of general practice. We should assure that parts of the population defined by the risk algorithm as a target for intervention are in concordance with a sustainable and manageable high-risk strategy. At the same time, this strategy must ensure the primum nil nocere principle, avoiding the possible hazards of overtreatment and medicalization of large parts of the population.

Agenda

Table

2007
10–13 May 2007	Nijmegen (The Netherlands) EGPRN spring meeting	Informationhttp://www.egprn.org/meetings.htm
24–27 July 2007	Singapore 18^th WONCA World Conference	Information www.wonca2007.com