Comment by jdietrich | Hacker News Reader

Comment by jdietrich | Hacker News Reader

Parent Full thread

Read Dr Ben Goldacre's book "Bad Pharma". Scientific misconduct is endemic in the pharmaceutical industry, most importantly in the form of publication bias. Vast swathes of research goes MIA, with the statistical analysis showing clearly that positive results are vastly more likely to be published than negative results. Industry-funded research is more than twice as likely to return a positive result than other research. This all takes place with the implicit endorsement of the FDA - the FDA have far greater access to trial data than any physician.

The pharma industry produces a lot of useful and important drugs, but they also do a lot of seriously shady shit. GSK habitually lied about the risks of Paroxetine and hid data showing serious safety issues. Merck committed fraud on a grand scale by falsifying and concealing data related to Rofecoxib, which led directly to around 60,000 deaths. It's quite clear that the ethical standards of the pharmaceutical industry are close to nil - most large companies are happy to kill patients with useless drugs as long as it's profitable.

That might be. But reducing placebo effect is not shady shit. Publication bias is a difficult phenomenon to fight, but when a pharmaceutical company invests millions into a drug that does not significantly outperform placebo then that product has failed. You can't keep repeating the research over and over again until it succeeds if every research project costs millions.

You make it sound like a drug that performs as good as a placebo is useless. Maybe you haven't read about the placebo effect, it is pretty damn effective.

"You make it sound like a drug that performs as good as a placebo is useless."

But it is useless, is it not? The placebo effect is effective, cheap and safe. If the new drug has the same effectiveness, it's probably just another placebo, just far more expensive and probably nowhere near as safe.

Not necessarily. A placebo might perform as good as a regular drug in clinical trials, where there is a 50% chance of it being the real thing and there is significant personal interaction with researchers. Also for some reason people always assume that if there is a clinical trial with humans involved the real drug will probably work very well.

In real world scenarios however drugs might be bought with _no_ personal interaction with professionals at all. Furthermore people are very sceptical of the long-term effects of any psychology related drug. In this case the placebo's might perform a lot worse than the actual drug, which may have an actual noticeable effect.

The placebo effect is effective, cheap and safe, and might be present in both the placebo and the tested drug. But the placebo can't be sold as medicine. And the tested drug might still work without the placebo effect.

Ultimately, I understand it as this.

Especially in terms of psychological effecting drugs

If the people getting the placebo get better (in the case of psychological drugs.. feel better) as often as those on the actual drug, then it's quite likely the drug is doing nothing and both sets of patients are getting better because the belief of taking a drug that makes them better is causing them to get better.

(apologizes for the confusing paragraph)

And hence, the drug is deemed worthless and thrown out, time to get on with a trial of another drug.

"it is quite likely the drug is doing nothing"

The only way to be sure that the drug is doing nothing is to repress the placebo effect as much as possible.

There are people dying or suffering that could be helped with drugs that failed the placebo test. (Some of them could even have been helped with the placebos themselves)

But if the placebo effect is caused by human interaction, would this suggest that the best therapy is simply one-on-one counselling rather than a new chemical? In other words, we need to help people change their way of thinking more so than mess with the chemistry of their brain.

Yes, but one-on-one counselling is expensive and error-prone. In an ideal world everyone had access to a doctor that has the time and spirit to treat all of his patients with dedication and positivity.

In the real world enrollment to medical schools is heavily throttled and doctors in cities have many more patients then they can cope with.

There might be enough psychiatrists and other therapists, but they are not believed to be able to cure physical pain (even though they are), which renders them less effective. This is why these therapists rely so much on these medications that seem to have little effect over placebos.

What is problematic is when you set out to develop a statistical approach purely designed to increase the rates at which experiments deliver positive outcomes. You start letting the tail wag the dog when you start from the assumption you are testing stuff which must have a significant benefit, and ask how you can better demonstrate that benefit - rather than accepting that the results appear to show that the benefit is not particularly statistically significant.

Drug trials are already done in circumstances incredibly favorable to the drug companies. If even in such a situation a potential drug is struggling to show benefits the solution isn't to keep changing the way you're testing until you get the result you seek.

You start letting the tail wag the dog when you start from the assumption you are testing stuff which must have a significant benefit, and ask how you can better demonstrate that benefit - rather than accepting that the results appear to show that the benefit is not particularly statistically significant.

There is no such bias in this method. It could just as easily show a drug to be worse than a placebo (harmful) than better: there is no prior assumption in either direction. The method merely increases the accuracy of this result.

The fact that it increases the number of positive drug outcomes is a good thing: we want more working drugs. Hiding drug effects with weaker tests, giving more inconclusive results, is not an advantage.

Reply on news.ycombinator.com