In a letter published in the British Medical Journal (BMJ), Michale Soljak, Chris Millett and I discuss the call by Krogsbøll to abandon the NHS Health Check Programme.As members of one of the research teams evaluating the national programme, we challenge their view that a randomised controlled trial (RCT) of the current programme should be the sole arbiter on which to base national health policy decisions. As noted in comments on the Cochrane review, the dearth of RCTs of screening for cardiovascular (CVD) risk since 1999 is unfortunate, but there were many RCTs of individual screening components over this period that have been used in subsequent NICE guidance.
There are other reasons why Krogsbøll et al’s views should be challenged. Firstly, RCTs are expensive and time-consuming, and as a result, a lack of evidence from RCTs may deprive millions of people of potential benefits from interventions. RCTs should therefore be complemented where possible by well-validated modelling studies. Screening for CVD risk in six European countries has been evaluated using the Archimedes simulation model, which has been extensively validated against many RCTs. The cost per quality adjusted life year of a universal check was €11,595 in Denmark and €2,426 in the United Kingdom (UK), and if targeted on the top quartile of risk costs would be €1,800 and cost-saving respectively. This is much better societal value than many healthcare interventions, and is very similar to the UK Department of Health’s original health economic impact assessment of the programme.
Secondly, Krogsbøll et al’s conclusion takes no account of the societal costs of waiting for RCT results. A UK analysis of current patterns of statin treatment showed substantial waste and inequity, with overuse in low CVD risk and underuse in high risk (600,000 and 850,000 patients respectively since 2007). There was wide variation between practices in statin prescribing to patients at high CVD risk. Perhaps this is not the case in Denmark, but the UK should not wait for the inverse care law to grind to its eventual outcome. Universal screening programmes can increase health inequalities if uptake varies, but the Scottish Health Check Programme targets deprived populations as a policy, and in England coverage is significantly higher in PCTs in the most deprived areas. RCTs and Cochrane reviews have little to tell us about health inequalities or the foregone costs and benefits.
Our local early evaluations of the English programme show large increases in appropriate statin prescribing after a Health Check, and significant reductions in mean CVD risk score, diastolic blood pressure, total cholesterol levels and lipid ratios. Uptake is lower than planned, but the equivalent New Zealand programme shows that 75% coverage can be achieved with effective implementation.
There are other reasons why Krogsbøll et al’s views should be challenged. Firstly, RCTs are expensive and time-consuming, and as a result, a lack of evidence from RCTs may deprive millions of people of potential benefits from interventions. RCTs should therefore be complemented where possible by well-validated modelling studies. Screening for CVD risk in six European countries has been evaluated using the Archimedes simulation model, which has been extensively validated against many RCTs. The cost per quality adjusted life year of a universal check was €11,595 in Denmark and €2,426 in the United Kingdom (UK), and if targeted on the top quartile of risk costs would be €1,800 and cost-saving respectively. This is much better societal value than many healthcare interventions, and is very similar to the UK Department of Health’s original health economic impact assessment of the programme.
Secondly, Krogsbøll et al’s conclusion takes no account of the societal costs of waiting for RCT results. A UK analysis of current patterns of statin treatment showed substantial waste and inequity, with overuse in low CVD risk and underuse in high risk (600,000 and 850,000 patients respectively since 2007). There was wide variation between practices in statin prescribing to patients at high CVD risk. Perhaps this is not the case in Denmark, but the UK should not wait for the inverse care law to grind to its eventual outcome. Universal screening programmes can increase health inequalities if uptake varies, but the Scottish Health Check Programme targets deprived populations as a policy, and in England coverage is significantly higher in PCTs in the most deprived areas. RCTs and Cochrane reviews have little to tell us about health inequalities or the foregone costs and benefits.
Our local early evaluations of the English programme show large increases in appropriate statin prescribing after a Health Check, and significant reductions in mean CVD risk score, diastolic blood pressure, total cholesterol levels and lipid ratios. Uptake is lower than planned, but the equivalent New Zealand programme shows that 75% coverage can be achieved with effective implementation.
Comments