To the Editor:

We consider that the poor agreement (33.3%)in the high-risk estimations presented by Buitrago et al1 with use of the SCORE (Systematic Coronary Risk Evaluation) and REGICOR (Registre Gironí del Cor, Heart Register of Girona) assessment systems is a call for caution: the SCORE Function Chart is recommended by European societies and the Comité Español Interdisciplinario para la Prevención Cardiovascular (Interdisciplinary Spanish Committee for Cardiovascular Prevention), whereas the REGICOR Function Chart has only been validated in Spain.2

These charts are difficult to compare, since SCORE predicts vascular mortality (coronary death caused by cerebrovascular disease, peripheral artery disease, heart failure, dissecting aneurysm of the aorta, and others) in subjects between 40 and 65 years of age, excluding diabetics (with an indication that these should be treated directly as patients in secondary prevention, although they were not excluded from the cohort when adjusting the function), and does not consider high-density lipoprotein (HDL) concentrations. REGICOR predicts coronary morbidity and mortality (angina, fatal and nonfatal acute myocardial infarction) in patients between 35 and 74 years of age and includes individuals with diabetes and HDL.3

Each assessment system has its pros and cons. Perhaps the most important criticism of REGICOR is that 68.4% of the validation sample was from Catalonia, a region that contains approximately 16% (see http://www.almendron.com/ politica/ine/2006/np421.pdf) of the total population of Spain. It is worth noting, however, that the city and the state of Massachusetts where the Framingham study is being conducted account for less than 0.03% and 3% of the United States' population, respectively. Moreover, only 6% of the population from which the baseline risk was obtained for "low-risk" areas and the distribution of population-based risk factors in the SCORE function were Spanish, whereas the other 93.9% were from France, Italy, and Belgium, countries with a baseline risk approximately 30% higher than in Spain.3

The idea of predicting general vascular mortality in SCORE is based on one of the principles advocated by European societies, that is, the need to shift from coronary prevention toward cardiovascular prevention; however, according to percentages inferred from Figure 3 in the study by Buitrago et al,1 stratification based on high SCORE risk could exclude up to 28.2% of patients with an elevated probability of acute coronary syndrome (ACS), but would include up to 38.5% of subjects with "low" ACS risk according to REGICOR. The latter subgroup ("low" REGICOR risk and high SCORE) would theoretically be composed of individuals at high-risk of stroke who would benefit much more from treatment to reduce hypertension than from lipid-lowering therapies, unlike the "coronary" group.

Morbidity accounts for 75.1% of the morbidity and mortality in Spain, although the socioeconomic impact of morbidity is unquestionably greater. The mortality of coronary disease is, in fact, declining in most European countries, including Spain, even though its incidence remains stable. This fact indicates that mortality may be a poor indicator of morbidity.3 It has recently been shown that this type of risk function cannot be validated and compared in small, biased samples with few fatal cardiovascular events.4,5

In view of these arguments as well as the limitations of risk functions,5 we feel it is not necessary to wait for the complex validation of SCORE in Spain, which would require a cohort of more than 50 000 patients, to "tip the lead towards choosing one of them in the management of cardiovascular risk in Spain."1