Letters NHS Health Check

Risk stratification could reduce costs in primary prevention of cardiovascular disease

BMJ 2011; 343 doi: http://dx.doi.org/10.1136/bmj.d4913 (Published 10 August 2011) Cite this as: BMJ 2011;343:d4913
  1. Andrew R H Dalton, doctoral student1,
  2. Richard J Bull, senior public health strategist2
  1. 1Department of Primary Care and Public Health, Imperial College, Charing Cross Hospital, London W6 8RP, UK
  2. 2NHS City and Hackney, London EC2A 2EJ, UK
  1. andrew.dalton07{at}imperial.ac.uk

England is rolling out a major primary prevention programme for cardiovascular disease, the NHS Health Check.1 2 Only patients at high risk of diabetes need a blood glucose test, not all need serum creatinine tests, but all must have lipids assessed. We investigated, using data from 73 853 medical records in NHS Ealing,3 the need to measure lipids in low risk patients.

We applied the QRISK2 algorithm4 in patients aged 40-74, firstly replacing lipid data with age and sex estimates from the Health Survey for England, and secondly using complete data. Using survey data, 27 682 (37.5%) (table) patients were estimated at <5% risk, and none was at high risk of cardiovascular disease when complete patient data were substituted. Another 13 170 (17.8%) were at 5-10% risk, and only 11 became >20% risk. Around 15 million people are eligible for the programme.5 Given our estimates and a cost of £4.20 (€4.8; $7) per lipid test,4 £24m will be spent every five years on lipid tests in patients at the lowest (<5%) risk.

Cardiovascular disease risk scores using lipid data or estimated lipid data

View this table:

We question the usefulness of universal lipid testing during a health check. Many already have lipids recorded.3 For patients at the lowest risk, lipid values add limited information to risk profiles, with risk scores unable to discriminate between low levels of risk. Familial hyperlipidaemia is an important mediator of cardiovascular risk, but family history of coronary heart disease should be the important driver in diagnosis, not population screening. The premise of lipid testing as a “hook” to promote attendance has no supporting evidence. With increased strain on NHS spending, risk stratification within the programme could reduce costs.


Cite this as: BMJ 2011;343:d4913


  • Competing interests: None declared.