Alongside growing discontent, the scientific community has answered by driving forward a raft of open science reforms. From initiatives to making research data publicly available, to ensuring that all published research can be read by the public, the aim of these reforms is simple: to make science more credible and accessible, for the benefit of other scientists and the public who fund scientific research.
Registered Reports: Peer review before results are known to align scientific values and practices
- Peer review occurs prior to observing the outcomes of the research.
- Manuscripts that survive pre-study peer review receive an in-principle acceptance that will not be revoked based on the outcomes, but only on failings of quality assurance, following through on the registered protocol, or unresolvable problems in reporting clarity or style.
The second problem with clinical trials is a form of cherry picking called hidden outcome switching. Back in the 1980s, the medical community decided that clinical trials should be registered in advance, pre-specifying the study design, outcome measures, and analysis plan. One of the reasons this was brought in was to stop cherry picking by researchers, either deliberately or unconsciously. In any study with a lot of variables, it is relatively easy to find a positive result by changing the main outcome measure after looking at the data or by using a different type of analysis to the one originally planned.
Publication bias places researchers under enormous pressure to engage in cherry picking – all the incentives in academia point researchers toward cutting corners and fooling themselves (and others) in the interests of publishing in prestigious journals. So it comes as little surprise that cherry picking is rife in clinical trials: a 2014 analysis found that around 1 in 3 trials change their primary outcome measure after the trial is complete,
The first step is to treat the publication process as we would treat a clinical trial itself and ensure that journals are kept blind to the results when they decide whether to accept or reject articles. This guarantees that positive and negative results are on equal footing in the published literature. And secondly, we need to ensure that researchers adhere to their registered trial protocols, or at least explain why deviations from protocol are required. For this, peer review of the protocol provides the answer.
Hooray some say, I say whoopy-do.
Will it mean the endless sea of rubbish underpowered clinical trials stop?...I doubt it.
Will it make clinical trials be published? I doubt it.
This is a vehicle for academics to get their dreary trials published.
At present there are Journals of clinical trials where you can publish your clinical trial design.
If they reviewed and said your clinical trial is pants...would it stop you?
We want to see the clinical trials done by companies published like the "inhibition of BAFF" So you get your trial design accepted but you decide not to publish the failed trial.
Now lets look at the journals who have signed up to this, we are missing the New England Journal of Medicine and Nature and Science...They care about the result and not the trial.
You have an easy way of ensuring companies publish results but it needs some guts...guts from the regulators, and guts from the journals, which are supposed to require trials to be registered as part of publication.
Let people registered with clinical trials.gov have a 24 month window from the end of the trial and if they do not publish the trial, refuse to allow future registrations..hence publications.
Companies would loose the advertising medium.
Yeah......it ain't going to happen. People don't have the jajce!
However, we can't talk on this can we....ProfG?
Where is the INSPIRE trial publication?....Time to Pull your finger out:-)