3 ms·

The solution to the problem is simple. No more grant money until all results are published. If you run a trial with reporting requirements and fail to do so the

The solution to the problem is simple. No more grant money until all results are published. If you run a trial with reporting requirements and fail to do so then you are banned from getting any further funding. A credible threat to cut off funding will get the results out ASAP. If you want to make it even more effective ban the entire university. Peer pressure will work wonders. To solve the same problem in the private pharma industry just refuse to approve any new drugs from any company unless all trials are reported within x time period.

gruez 11y ago
>solve the same problem in the private pharma industry just refuse to approve any new drugs from any company unless all trials are reported within x time period. I'm assuming that they would also have to register what their trials will be ahead of time? Otherwise there's no way for the regulators to know what trials need to be reported.
- danieltillett 11y ago
  Yes all human trials have to registered in advance - this is why we know how many have published results or not.
Gatsky 11y ago
It can take a while for results to get published - it's often up to the whims of the journal/reviewers. Just make them release the raw data. That will definitely encourage people to publish. Although these stats look bad for the researchers, the other side of the story is that publishing anything these days is a pretty painful and not exactly transparent process.
- danieltillett 11y ago
  Yes this is true - I think you have are supposed to have two years after the trial ends to publish the results. This study was more than 5 years later so I don’t think we can blame reviewers or editors for the delay.
personjerry 11y ago
That's not a solution at all. That might stop just seriously stem corporate money into research. It seems like to me a better solution would be a "reverse copyright" of sorts, where the results of any study MUST be published after X amount of years.
- danieltillett 11y ago
  The studies are already supposed to be published - the problem is that there has been no enforcement. Until there is some real penalty for not publishing then nothing will change. Having said this if the publication(s) are not well done then this can be almost as bad as not publishing. No data is often better than bad data.

gruez 11y ago

>solve the same problem in the private pharma industry just refuse to approve any new drugs from any company unless all trials are reported within x time period. I'm assuming that they would also have to register what their trials will be ahead of time? Otherwise there's no way for the regulators to know what trials need to be reported.

danieltillett 11y ago
Yes all human trials have to registered in advance - this is why we know how many have published results or not.

Gatsky 11y ago

It can take a while for results to get published - it's often up to the whims of the journal/reviewers. Just make them release the raw data. That will definitely encourage people to publish. Although these stats look bad for the researchers, the other side of the story is that publishing anything these days is a pretty painful and not exactly transparent process.

danieltillett 11y ago
Yes this is true - I think you have are supposed to have two years after the trial ends to publish the results. This study was more than 5 years later so I don’t think we can blame reviewers or editors for the delay.

personjerry 11y ago

That's not a solution at all. That might stop just seriously stem corporate money into research. It seems like to me a better solution would be a "reverse copyright" of sorts, where the results of any study MUST be published after X amount of years.

danieltillett 11y ago
The studies are already supposed to be published - the problem is that there has been no enforcement. Until there is some real penalty for not publishing then nothing will change. Having said this if the publication(s) are not well done then this can be almost as bad as not publishing. No data is often better than bad data.