Scott Alexander argues that the 33% rate of post-marketing safety events for FDA-approved drugs is not necessarily concerning and explains why post-marketing surveillance is a normal part of drug safety monitoring.
Longer summary
Scott Alexander discusses a study showing that 33% of FDA-approved drugs in the past decade have faced post-marketing safety events. He argues that this doesn't necessarily mean the FDA is too lax, explaining that post-marketing surveillance is a normal and necessary part of drug safety monitoring. He points out that only 1.3% of drugs were actually withdrawn from the market, which he considers a good success rate. Scott explains why it's impossible to catch all potential side effects in pre-approval studies and gives examples of safety communications that range from important to seemingly trivial. He concludes that the 33% figure alone is meaningless without a broader cost-benefit analysis of FDA approval standards.
Shorter summary