Nearly 1 in 3 drugs have a significant safety issue after FDA approval

Yup, hence Obamacare with mandatory enrollment and the removal of all the tools they used to it. Then people complain about prices going up. :/

So pretty much everyone needs an 18 months of savings + COBRA costs. Even then its quite likely not enough unfortunately.

FWIW I've been prescribing medications for a number of decades and pretty well understand the general benefits and limitations of pharmocotherapy as a treatment method.

To save you a lot of time and trouble studying this very complex field, let me boil it down to these few words: any medication can cause any side-effect at any time.

All drugs cause side multiple effects. The ones we like we call "benefits" or therapeutic effects, those we don't like are "adverse effects" (AEs). Useful drugs will have few troublesome AEs vs. good benefits, but it's a relative matter, a judgement call about what's good, bad or ugly.

It partly depends on the condition the drug is used for. In life-threatening situations we'll accept bad AEs, for example, cancer chemotherapy agents. For merely annoying, self-limiting problems, like a common cold, we wouldn't take big risks.

However, no drug is absolutely safe. All are capable of causing serious problems in at least a few people. Therefore we must use all drugs carefully and with thorough knowledge of the risk/benefit trade-offs, but there are never any guarantees of outcomes.

Another aspect is that often severe AEs are quite uncommon. Think of a side-effect that occurs in 1/100000 recipients. What are the odds this will be detected in drug registration trials? Most drugs are tested on a perhaps a few thousand people before approval, and often fewer subjects than that. The chances of encountering these outcomes is remote until drugs have been around for years, even decades before the harmful effects can be gleaned.

One drug that became notorious for incidents of fatality in the late 90's was nefazodone (Serzone), an antidepressant. I was peripherally involved with research on the matter. Basically what came to light was that in people with rare (like 1 in a million) mutations in certain liver enzymes, the drug was metabolized along a pathway that produced compounds causing fatal liver necrosis. After 18 deaths around the world became known, the association was made and the drug was withdrawn from the market. There was no way to predict this would be an effect of the drug when it was approved.

Of course AEs are very seldom fatal. The point here is that AEs (serious or not) are quirky, inevitable and randomly distributed. The benefits are always intertwined with finite risks. It's not surprising in the least that a high proportion of drugs are associated with troublesome AEs (with varying definitions of "troublesome").

I can be swayed , but i do subscribe to the concept that there is an availability bias in FDA approval: the fda has the incentive to be overly cautions, because punishment hurts fda, but banning potentially good medicine does not cause FDA any problems.

That, and the potential of regulatory capture make me dubious of any positive claim about the existence of the FDA.