Assessing the risks of meds

Dr. Maurice Bernstein responds to my post about Celebrex and drug safety. Dr. Bernstein writes the excellent Bioethics Discussion Blog. Here's his comment:

Just as relative benefits of drugs are displayed to the public in drug ads to get them to request from the physician one drug over another, it seems, unfortunately, that relative risk is presented to the public by the news media without providing understanding the statistical significance. Look..in a drug study, if two patients in a million patients who take the drug at this dose for this length of time gets a "heart attack" and only one patient in a million who were given an inert placebo gets a "heart attack" then the news media will report that there is TWO TIMES THE RISK of a "heart attack" in those who took the drug than those who took only a placebo. Where is the disclosure of the absolute risk which is two patients in a million compared with one patient in a million? Wouldn't the absolute risk be important to know if one is going to weigh the risk of taking or not taking the drug compared to the absolute benefit it may provide...such as 500 of 1000 patients taking the drug get complete relief of their chronic "aches and pains"? I think that the relative risk as reported is meaningless if the absolute risk is left unreported to the public. But I guess that is how it goes in news reporting.. get the "juicey" part out quickly or maybe I am too cynical. Somebody knows the absolute risks of the current COX 1 and COX 2 drug studies. Let's hear about those numbers. After all, full statistical disclosure makes good ethics...Maurice.