Figure 10.1 is taken from Nakamura et al. (2006) who reported a large placebo-controlled randomised trial evaluating the effect of pravastatin in preventing cardiovascular disease. The overall treatment effect was positive, with a hazard ratio of 0.67 (p = 0.01). We will cover hazard ratios and their use in survival analysis in Chapter 13; for the moment simply note that, like the odds ratio and the relative risk, a value of one corresponds to equal treatments. The homogeneity of the treatment effect was assessed by looking at the p-value for the treatment-by-covariate interaction and also by calculating the hazard ratio separately in various subgroups defined by baseline factors of interest as seen in Figure 10.1.
Was this article helpful?