Meeting of the Food and Drug Administration’s Anti-Infective Drugs Advisory Committee, College Park, MD
We like to think that medicine is science, that science is based on numbers, and that numbers don’t lie. Mark Twain once famously observed though that there are “lies, damned lies, and statistics.” I was reminded of Mr. Clemens’ quip this morning sitting in a hotel conference room, listening to a drug manufacter’s team and the members of an FDA advisory panel draw very different conclusions from the same trial data.
Through some miracle, I never had to take statistics to get my chemistry degree. I thank my lucky stars for having dodged that bullet. Today, it was abundantly clear to me that by adjusting confidence intervals and increasing margins and some other statistical hocus pocus that I don’t really understand, you can more or less get numbers to say pretty much anything you want them to.
When I read the background material on the trial data, my common-sense thought was ‘I don’t think this drug should be approved.’ I was reassured about how seriously FDA advisors are taking their role in the wake of the uproar over the safety of the coxibs, when they agreed with me today.