Critical comment on the accuracy of the new CV risk prediction calculator
Statins: new American guidelines for prevention of cardiovascular disease
Ridker PM en Cook NR.
The Lancet. Comment published online November 19, 2013 http://dx.doi.org/10.1016/S0140-6736(13)62388-01
The new guidelines of the American College of Cardiology (ACC) and the American Heart Association (AHA) or the management of cholesterol are a step in the right direction: they emphasise prevention of stroke and heart disease, focus appropriately on statin therapy rather than alternative unproven therapeutic agents and appreciate that for many patients more intensive treatment is better than milder treatment. It is also pointed out that the benefits with respect to myocardial infarction, stroke and cardiovascular mortality clearly outweigh the risks of developing diabetes and myopathy, in patients with a clear indication for statin therapy. By taking the focus away from LDL-treatment targets, as well as measuring creatinin kinase during follow-up, clinical practice is simplified specifically in primary care.
However, with regard to primary prevention the guidelines are possibly more controversial. The newly developed risk prediction algorithm is based on ‘hard’ atherosclerotic events. It recommends starting with statin treatment in a primary prevention population with a predicte 10-year risk of at least 7.5%, and between 5 and 7.5% statin treatment may be considered. In patients with type 1 or 2 diabetes mellitus, the threshold of 7.5% risk is used to prescribe either low or high intensity statins.
As described in the guidelines, these new criteria could lead to more than 45 million potential statin users, without any current cardiovascular disease.
The authors of this comment wonder whether a global risk prediction score is needed, since several large prevention studies have shown that statins are effective as primary prevention to reduce myocardial infarction and stroke in patients with high LDL-c (WOSCOPS, MEGA), low HDL-c (AFCAPS/TexCAPS), and high concentrations of C-reactieve protein (JUPITER), diabetes (CARDS) or hypertension (ASCOT-LLA).
There has never been a statin study that used a global risk prediction score as an inclusion criterium, thus prescription behaviour is difficult to defend on the basis of such a measure in an evidence-based climate. Moreover, trial data show that a high absolute risk does not always predict whether statins will be effective.
The patient’s situation should ultimately determine the decision of the physician, whereby the predicted treatment effect based on risk factors is more important than the absolute risk.
The authors further question whether the prediction algorithm correctly predicts vascular risk. They therefore validated this model in three large existing primary prevention patient cohorts (Women’s Health Study, Women’s Health Initiative, en Physician’s Health Study). They observe that the AHA/ACC prediction algorithm systematically overestimates the risk as compared to the observed event rates in these cohorts. The guideline also validated the model in two external cohorts, and saw an overestimation of risk, which is discussed in the guideline.
It is possible that the five validation cohorts used represent more contemporary populations than the cohorts used to develop the model. The authors of this comment therefore call for more validation in external cohorts, before these new prediction models are widely implemented. Based on the new risk calculator many primary prevention patients might receive statin treatment, although little trial evidence exists for benefit thereof.
You can find the comment of dr. Ridker en dr. Cook fulltext online in The Lancet News • 13-11-2013