Table 3

 Multilevel linear regression analyses (outcome: CERs)

VariablesUnivariate analysis Effect in % (95% CI)Interaction effects with absolute risk of CHD*Effect in % (95% CI)
CERs, cost effectiveness ratios; CI, confidence interval; NS, not significant effect; CHD, coronary heart disease. *Effects on the cost effectiveness ratios, per 1% rise in absolute risk of coronary heart disease. The original (univariate) effect of 1% increase in absolute risk (21.8 % decrease in CERs) is compared in each variable category with account for potential interactions. †p Value <0.05. ‡Primary: based on findings from a randomised trial. Secondary: based on expectations from a theoretical model. §Secondary prevention is among patients with clinical cardiovascular disease, primary prevention among apparently healthy persons.
Absolute risk of CHD−21.8 (−27.6, −15.6)†
Age in years
    <45Reference−53.3 (−62.2, −42.4)†
    45–6523 (−20, 91)−26.4 (−32.1, −20.2)†
    >6576 (0, 211)−24.3 (−36.6, −9.8)†
Treatment durationNS
    <5 yearsReference
    5 to 10 years−22 (−69, 103)
    Lifetime (>10 years)−42 (−73, 24)
Cost country
    USAReference−27.9 (−35.6, −19.3)†
    Canada165 (−59, 1594)−4.4 (−16, 8.8)
    Europe207 (−48, 1764)−31 (−40.7, −19.9)†
Effect modelledNS
    Intermediate effectsReference
    Risk reduction69 (−58, 581)
Type of modelling‡NS
    PrimaryReference
    Secondary−27 (−71, 87)
    Prevention Categories§
    PrimaryReference−31.5 (−40.6, −21.1)
    Secondary−62 (−74, −45)−5.6 (−16.8, 7)
Economical perspectiveNS
    SocietalReference
    Third party payer−53 (−89, 96)
Funding source
    Pharmaceutical companiesReference−13.9 (−22, −5)
    Others−44 (−86, 129)−31.3 (−38.8, −23)
Publication year
    <1996Reference−40.2 (−49, −29.8)
    ⩾1996147 (−55, 1255)−16.5 (−23.3, −9.2)
Discount factor
    > = 5%Reference−28.9 (−35.1, −22.2)