Screening for Cardiovascular Risk: Cost-Benefit Considerations in a Comparison of Total Cholesterol Measurements and Two Compound Blood Lipid Indices

Abstract

Background: Serum total cholesterol measurements have been shown to differentiate between patients with angiographically confirmed coronary artery disease and controls less well than compound indices of cardiovascular risk. Details of employees (n = 229) nominated by an occupational health service in a non-manufacturing firm were used as a starting point for calculations to compare the costs and benefits of using compound indices of cardiovascular risk with those of total cholesterol measurements alone.

Methods: Healthy employees were defined as having a low or a high risk of cardiovascular disease according to either total cholesterol level or two compound indices of blood lipid components. The compound indices were the ratio of total to high-density lipoprotein (HDL) cholesterol (the TC:HDLC ratio) and an ‘atherogenic index’ defined as ([total cholesterol – HDL cholesterol] × [apolipoprotein B])/([HDL cholesterol] × [apolipoprotein A]). If compound indices discriminate better between people at low and high risk, both the number of people given unnecessary advice on lifestyle changes or urged to take cholesterol-reducing medication and the number of people not treated because of their ‘normal’ cholesterol levels would be reduced. In our calculations, we assumed as ‘gains’ that (1) the disclosure that a total cholesterol test result is false-positive is equal to treatment costs, consultation fee and consumption foregone (i.e. resources already used on medication, services etc.) (8909 Nkr [US$1 =7 Nkr]), and (2) the disclosure that a test result is false-negative is equal to consultation fee plus loss of 2 h wages (288 Nkr).

Results: The screening of 100 000 men and 100 000 women would incur a cost of 99 and 710 Nkr, respectively, per person assumed to benefit from extended screening using two different compound indices. Net gain would be 438 and 192 million Nkr, respectively, for the two compound indices. However, the lack of prospective data on compound indices suggests the need for cautious interpretation.

Conclusion: Although prospective studies are needed to confirm our findings, the changes in number of false-positive and false-negative values achievable using different indices suggests a need for greater caution when using single lipid measurements as predictors of risk. The calculations of this non-prospective study indicated an increased benefit-cost ratio in assessing cardiovascular risk by using compound indices of cardiovascular risk compared with total cholesterol measurements alone.