Deprivation-based risk scores: the re-emergence of postcode prescribing in the UK?

Socioeconomic gradients exist in the prevalence of cardiovascular disease. This has prompted the development of risk scores such as ASSIGN and QRISK, which incorporate measures of deprivation, to address the issue of underprescribing of primary preventive medicines in the socially disadvantaged. The scores use area-based measures of deprivation rather than the socioeconomic status of the individual. We examined to what extent the decision to treat a patient might be influenced by where that individual lived.

On the basis of individual patient risk factor data from the Scottish Health Survey, we compared the theoretical level of deprivation [Scottish Index of Multiple Deprivation (SIMD)] required to give a person an ASSIGN risk of 20% (the treatment cut-off), with the person's actual SIMD quintile. We assumed that patients are more likely to move between areas of similar deprivation (i.e. the same SIMD quintile). If the theoretical SIMD value for that individual fell within their actual SIMD quintile, we assumed that prescribing decisions could be influenced by the area of residence.

If the ASSIGN risk score was implemented, the area of residence would affect the decision to initiate statins in the case of 15.7% of the population (aged 30-74 years), and for borderline hypertension, in 3.0%. This corresponds to 407 000 and 15 000 people, respectively, in Scotland and 4.6 million and 169 000 in the entire UK.

These findings demonstrate that by using cardiovascular risk scores based on area deprivation, primary prevention treatment decisions will be affected by the area of residence in a large number of individuals.