Merck & Co corrects APPROVe trial analysis

USA-based Merck & Co has released a correction to one of the statistical analysis methods used in the APPROVe (Adenomatous Polyp Prevention on Vioxx) study, which was originally published in the New England Journal of Medicine in 2005. The firm said that it identified the error while reviewing the preliminary analyses of the off-drug extension data for the study, but added that it believes that the overall results of the program are unaffected.

Specifically, the analysis focussed on the relative risk of confirmed thrombotic events caused by Vioxx (rofecoxib), compared with placebo, requiring the use of numerous statistical methods. The defined variable, logarithm of time, which was published in the description of methods in the NEJM article and was submitted to regulatory authorities, was used in error. The reported result (p-value = 0.01) came from a statistical model which used linear time, not logarithm of time, which has a p-value of 0.07.

The findings from the APPROVe study led to the withdrawal of Vioxx from the market, and caused the firm's share price to plummet 23%, when they were originally announced (Marketletters passim).