Tuesday, 19 May 2015

Bad Pharma Its Not All bad but still trying to hide stuff?

There is a move to try and get all trials registered with an undertaking to make the trial data public, so the data can be re-analysed and failed trials will be published so prevent others following in the foot steps. In the UK it seems that a contract research organisation has made a legal challenge against this action according to Newspapers.
The Health Research Authority (HRA) is a UK body set up to protect and promote the interests of patients and the public in health research in the UK. It has been working to make clinical trials more transparent: In September 2013 the HRA made registration of a clinical trial a condition of approval to run that trial and in April this year they started asking researchers applying to run a clinical trial to declare that all past clinical trials they ran have been registered.

However it has to be said that some companies have embraced this, and do make their trial data available, so they are not all bad apples.

How quick should we expect that data to be released, it takes time to get papers published.

What do you do with thousands of pages of trial data?..Should you publish stuff that hasn't been published from the data or is that bad form?

9 comments:

  1. That this challenge to openness in clinical trial data is being made by a UK company is depressing.
    I would hope that potential clients would vote with their feet.

    ReplyDelete
    Replies
    1. I suspect it is their clients that are making them do it. The CRO could fold with a golden handshake and resurface under a different name and the clients still smell of roses rather than mud...However this can backfire as the white knights may reveal

      Delete
  2. Say for instance that all the data was available to peruse through, how much more confused will we all be?
    Don't get me wrong, Ben Goldacre makes a valid point for the shifty types, but in everyday instances it's the interpretation which leaves you with the exclamation 'you what??!' - at times from respected scientists/clinicians whom seem to have lost a widget or two...
    Peer review you say? more like pal review...

    ReplyDelete
    Replies
    1. PAL review I like it......Some reviewers make a real dog's breakfast of it. 😀

      Delete
  3. Failed clinical trials and their data should not just slip off the radar, never to be seen again. However, the effort of creating manuscripts for journal publication is labor intensive and a process skewed by the PAL reviews. Yes, I know the pharmas have plenty of skilled writers available to put these together but would we really want to read it?
    I would propose a formulaic method of reporting fails through a registry - simply something that says name of the trial, sponsor, the drug, the cohort information and why the trial was stopped, including the adverse effects information. And then do something revolutionary and list contact information if someone really wants to ferret out the details they can do so by asking for more information directly from the source. This is my layperson view - simplistic but gets the information into the public arena and requires sharing.

    ReplyDelete
    Replies
    1. I'm not sure I entirely agree here

      There are already places for reporting clinical trials and one such trial is clinical trials.gov and so there is a registry. There are European Registries also.

      It is a condition of many top journals that trials are registered before any person is treated otherwise they refuse to publish the results. This is to stop p-hacking. The companies and PIs doing the trials are often indicated, but as you say there is no requirement to say what the result is or if their was no result why no result.

      It could be possible that companies are enforced to report the results within a certain time. However if the companies are liquidated,which often happens after a failure this may be hard. You will need a stick rather than a carrot for this to happen

      Companies sit on immense amounts of science data that they keep quiet and this will pale the unseen trial data into insignificance. They try and reproduce a lot of the academic stuff
      and know it is unreproducible pants but the data never see the light of day.Some examples have been published but very few.

      Furthermore, whilst I agree it is a lot of work to get stuff ready for publication, the data will be analysed and presented in a publication ready format because those people analysing the work are not going to be the same senior people making the decisions about whether to can or keep the project going and therefore there will be a written report or a data set of graphs, maybe not publication quality nut there will be graphs etc.

      Likewise pharma collects data in the form of written reports because they need this for regulatory submission and much of the work is outsourced because (a) they dont need to pay to learn the skill set (b) it is independent and so transparent and (c) they can can the company or go elsewhere.. However, companies contracted to do the work don't get their final payments until the reports are done. So there are always reports in a virtual "data room" Furthermore because most trials are on multiple sites the data is stored electronically so the data is there.

      The difficulty is extracting it in a usable format.

      For example You request the data and you get it and then are faced with twenty thousand pages of data it is a daunting task to know where to start.

      Delete
    2. thanks for the thorough response - this is exactly why I normally stick to the liberal arts side of the discussions - there is so much I don't know about these are handled. I did take a quick look at clinicaltrial.gov (my go to source, quite often) and see all the different varieties. Wouldn't it be nice if I could look at a trial that is marked 'terminated' and see why... I'm know you and other researchers would agree that information might be useful in your own work and also save you potential heartbreak from failed ideas.

      What type of incentives need to be created/funded to encourage the sponsoring agencies to play nice and share? (sorry, this is a rhetorical question).

      I look forward to sitting in on your talk at CMSC

      Delete
    3. You have me worried I don't now about presenting at CMSC next week I guess it could be profG.

      The incentives are a big stick beating them if they don't. The only incentive would be if they were not allowed to submit something else it they hadn't closed the study off. Although I am being hypocritcal as we have data that is still to see the light of day, we are hoping to submit some stuff that we started about 14 years ago and has some stuff older than that, we let some it surface from time to time.

      Delete
    4. No need to check your calendar or your passport - you are so right it is Prof. G. Too bad you aren't coming along to this side of the pond. It would be lovely to sit and chat about this and so much more.

      In a perfect world.... well to start with there wouldn't be MS so we wouldn't need to discuss sharing ideas and data and working in unison for a cure, would we. :-)

      Delete

Please note that all comments are moderated and any personal or marketing-related submissions will not be shown.