Here's a column by David Lazarus in the Los Angeles Times about Avandia:
Not being a regular reader of his column, I know little about Mr. Lazarus; but I assume from this specimen of his writing that he's a reasonably intelligent and perceptive individual. This makes it all the more discouraging how little he "gets it" with regard to what the real deal is about Avandia.
I'll discuss two assertions in his column. First, Lazarus seems to think it's a simple tradeoff of how much research you do pre-FDA-approval to discover potential adverse effects of drugs, vs. how much you hurt patients by delaying the approval of a potentially useful drug. That is, if you took long enough to do the pre-marketing research, you would find all major adverse reactions.
Experienced physicians know this ain't so, which is why there's the common folk wisdom among more conservative docs of waiting a year or two after a new drug is introduced before you start prescribing it for your own patients. Pre-marketing research always involves a limited number of patients compared to how many are exposed to a drug after it is being widely sold. A serious adverse effect that occurs 1 in 1000 times, for example, is very likely to be missed even in very intensive pre-approval research, yet once the drug is being sold, such a side effect might well be seen as fully sufficient reason to pull a drug off the market.
Even if the trade-off was as simple as Lazarus seems to think, he gets the politics all wrong. At the end of his column he advocates a policy shift that as soon as "credible" safety concerns are raised, the FDA should declare "an immediate moratorium on sales....If a drug becomes unavailable for a matter of months or even years, so be it."
Lazarus ignores the likelihood that neither the drug company nor patient advocacy groups (which in some cases will be joined at the hip financially) are likely to be in a "so be it" frame of mind. The industry will pull out all the stops to mobilize their "grass roots" to storm Congress and shriek that people are dying because this wonderful drug is suddenly unavailable. (As happened when new AIDS drugs were delayed in being released in the late 1980s, in that case by legitimate grass roots advocates and not by industry, which is why the FDA has its present policies in place that favor quick approvals.)
The other major issue that Lazarus doesn't get with regard to Avandia is why it is a bad drug and indeed could have been predicted to be a bad drug much sooner than it has been. Lazarus says that he's a Type 1 diabetic and so is not a candidate for Avandia in any case. That means he has a kind of diabetes where a drug that lowers your blood sugar provides positive health benefits. As I have tried to explain in numerous earlier posts, the main problem with Avandia is that it lowers your blood sugar but has never been shown to improve any other diabetes-related outcome. Type 2 diabetes has been now proven to the satisfaction of just about everyone except doctors who treat diabetes to be a disease whose outcomes are largely unaffected by tighter control of blood sugar levels. That means that blood sugar is what the evidence gurus call a "surrogate endpoint" and while a huge marketing tool for the drug company, ideally ought to be ignored by the medical community when deciding whether to recommend any new drug.
If a drug actually did something good for Type 2 diabetes, and as a downside it increased your risk of a heart attack by 43% (the figure Lazarus quotes for Avandia), then again you'd have a trade-off problem of the sort that Lazarus envisions. You'd have to decide if the benefit was worth the risk. The reason Avandia should have been yanked off the market the second there was even a whiff of a hint of increased heart risk was that in the face of no known (real patient-oriented) benefit, any risk is too high.
The short answer--the drug company bamboozled us twice about Avandia. First they conned us into thinking the drug was useful. Then they delayed us becoming aware of the downside risks of heart problems. Lazarus thinks he got un-bamboozled by figuring out the risk part of the equation. He's still bamboozled on the benefit side.
How would a person like Lazarus really figure it out with a complicated case like Avandia? Of the various books on the subject, the best I know on this particular matter is Jerry Avorn's Powerful Medicines. Avorn harps on how all judgments about pharmaceuticals necessarily hinge on judgments about benefits, risks, and costs, and all parts of the equation must be included. He also argues for a more flexible, sliding-scale FDA policy that can do a better job of approval than the simple "yes-no" the FDA now has as its only option. But, while Avorn's book is in fact quite correct and comprehensive, and while it offers a number of fine phrases here and there that I wish I had written, I have to say that overall it is a really tough slog. So you can see why even highly intelligent writers like Lazarus might not get it in the end.