• Cardiovascular disease
• prevention
• risk scores
• CHD/coronorary heart
• epidemiology
• prevention
• public health policy

Predictive risk algorithms are embedded in the daily practice of clinicians; they often represent a crucial tool to estimate a particular individual's risk of future disease. And, clearly, these individuals, en masse, will comprise the entire population.

However, do the same risk principles apply en masse? Geoffrey Rose was famously one of the first scientists to consider risk in entire populations. He suggested that the determinants of risk at the population level and at individual level might be crucially different.

The interesting paper by Manuel et al1 starts by highlighting that this risk is often assumed to be widely diffused across a population. However, this assumption might be wrong and might, therefore, have implications when evaluating prevention policies aimed at populations or at individuals. Manuel et al suggest that for cardiovascular disease (CVD), risk may actually be concentrated in a relatively small number of individuals, particularly when using risk functions that summarise the effect of multiple risk factors along with gender and age.1 The population thus categorised, would then show a concentration of events on the ‘riskier’ quantiles of the distribution. Manuel et al further suggest that routinely collected population data might therefore be used to develop robust risk algorithms which offer accurate, discriminating risk estimates in a whole population. Finally, they propose that ‘risk algorithms developed for the population setting offer a range of opportunities to support health planning including: the projection of disease cases; description of population risk; and evaluation of health risks and preventive interventions.’

The paper by Manuel et al, therefore, raises several interesting questions about how diffuse that risk might be, risk functions as …