Hacker News Reader

Comment by angry_octet | Hacker News Reader

What you’re saying is true, and completely tangential to the problem. When you are concerned that validation done by a subcontractor might be weak, it is entirely sensible that you subdivide the work. What we need is a statistical reanalysis that considers that, e.g., 10% of each step is done wrongly. And a regulatory regime that doesn’t hand wave away complaints, has rigorous protection for whistleblowers, and serious financial penalties for misconduct.

The BMJ might be right, but also have constructed an article that will be wildly misunderstood... and not sensible content for facebook.

View on HN

It will not be misunderstood by intended readers of BMJ. As for facebook it has all kinds of readers and groups including those that are meant for professionals.

I think I completely understand the article but still think that it's editorializing facts. Why exactly would professionals feel differently? If they had done some investigations themselves and found a series of poor contractors that all reported similar issues, I think we could take this as much more concerning. This looks like a director from a single company and a few anonymous researchers at the same company came forward with bad practices for 1000 trials.

Also, as the article points out, investigations were performed at other contractors. Wild speculation, but that could be that Pfizer didn't even need a full 44k and they just ignored 1k? (For example, they concluded the trial with 41k getting a second dose.)

The issue is that proper investigation was not done. We need more transparency and this is not what is happening. What happened with the data from these researchers? There are no indications that they were removed from the study analysis. If the problems with this one researcher is swept under carpet, how do we know that there are not number of other researchers with different problems?

Without this vigilance how do you distinguish between ivermectin studies that showed a positive effect and vaccine studies? Some of ivermectin studies but not all were discovered to be total frauds. At some point when you see too many problems you just have to distrust them by default until proved otherwise.

In fact, I think that there is possibility that the Pfizer vaccine trial results turned out quite different from effectiveness in real life in part due to issues like this.

The BMJ article, near the end, reports that the FDA actually did investigate and did find issues with the trial.

> An FDA review memorandum released in August this year states that across the full trial swabs were not taken from 477 people with suspected cases of symptomatic covid-19.

Of course, this comes after insinuating earlier that there is a complete lack of oversight.

> “There’s just a complete lack of oversight of contract research organisations and independent clinical research facilities,” says Jill Fisher, professor of social medicine at the University of North Carolina School of Medicine and author of Medical Research for Hire: The Political Economy of Pharmaceutical Clinical Trials.

I guess since you yourself were misled according to their own inconsistencies, maybe that is enough to convince you that the article is misleading.

I don't know what FDA review memorandum involves but it doesn't sound that the FDA did investigation. It could be simply that the FDA checked submitted documentation from that site and concluded this because the CRO hadn't provided any test results.

Sorry, I don't see any inconsistencies. The box clearly explains what they mean by complete lack of oversight – doing inspections many months after the trial and even then just checking the paperwork.

Not necessarily. If there's an instance of someone blowing the whistle, and there's no investigation it's completely reasonable to suspect that the whole integrity of the process might be iffy.

Blocking the article (instead of opening relevant books to show that in fact it was just exception to otherwise solid prices makes) makes even stronger case that the prices is not honest.

Reply on news.ycombinator.com