UK Government does what I tell them – and – how would you write the legislation?

March 6th, 2008 by Ben Goldacre in regulating research | 19 Comments »

Just a brief post on how gratifying it is to see the government obediently doing exactly what I told them to and announcing plans to ensure that all drug company trial data is registered and disclosed. The MHRA press release is below so that you can bathe in unmediated news.

One thing still troubles me: covering up harms from medication is tangibly evil, and eye catching; but it’s a shame that it takes this to trigger concerns, because the issue of covering up results which simply show your drug performing poorly against placebo is equally important, although less eyecatching. In fact I would say it’s more important, because it’s probably more prevalent, and has been known about for so many decades that it’s also more of an embarrassment that we’re only sorting it out now.

No trial should be done without being registered, no trial should be passed by an ethics committee without a commitment to publish, and no collaborative contract should exist between a company and an academic department – on any kind of treatment – without a clear clause in the contract stating that the researchers will have freedom to write up as they wish.

The interesting thing about epidemiology and big numbers is that it’s easy to forget what they mean. Bad data costs lives, and this piece of legislation, whilst nerdy, will probably have a greater impact on important things like death and pain than anything about foxes or tinkered health spending.


Press release

Date: 6th March 2008
Time: 0930 hours
Subject: GSK investigation concludes

Contact: Press Office 020 7084 3535 / 3564
Out of hours 07770 446 189



GSK investigation concludes

The MHRA has concluded its four year investigation into Glaxosmithkline and its anti-depressant drug Seroxat. The investigation focused on whether GSK had failed to inform the MHRA of information it had on the safety of Seroxat in under 18’s in a timely manner. The investigation was undertaken with a view to a potential criminal prosecution for breach of drug safety legislation. It was the largest investigation of its kind in the UK, and included the scrutiny of over 1 million pages of evidence.

The decision taken by Government Prosecutors, based on the investigation findings and legal advice, is that there is no realistic prospect of a conviction in this case, and that the case should not proceed to criminal prosecution. The legislation in force at the time was not sufficiently strong or comprehensive as to require companies to inform the regulator of safety information when the drug was being used for, or tested outside its licensed indications.

GSK provided the MHRA with data from clinical trials in May 2003, confirming that patients under 18 had a higher risk of suicidal behaviour if they were treated with Seroxat than if they received a placebo; and that Seroxat was ineffective in treating depressive illness in under 18’s. Acting upon this information the MHRA immediately reviewed the data and published advice to all doctors that Seroxat should not be used in under 18’s. The investigation arose from concern that GSK had held the information for some time before this and failed to disclose it.

Professor Kent Woods, MHRA Chief Executive, said: “I remain concerned that GSK could and should have reported this information earlier than they did. All companies have a responsibility to patients, and should report any adverse data signals to us as soon as they discover them. This investigation has revealed important weaknesses in the drug safety legislation in force at the time. Subsequent legislation has partially addressed the problem, but we will take immediate steps to ensure the law is strengthened further, so that there can be no doubt as to companies’ obligations to report safety issues.”


1. See website homepage to download full report and related


Next up, I’d like the government to buy into my other unloopy idea (not really just mine) of randomising every single treatment episode in the NHS, wherever there is genuine uncertainty about which is the better of two therapeutic options. It would answer the important questions about huge numbers of treatments in a trice, with admirable statistical power, it wouldn’t add much to the costs of running the NHS, it would be a great gift to humanity, and nobody would suffer, because we’d only be randomising between two treatments where nobody knew which of the two was better anyway.

Hopefully it’ll only take them a week or so to pop that one onto the legislative calendar.

But more importantly, back to the task in hand. Guiding blog discussion is like herding cats, but try this on, imagine you have unimaginable power….

….how would you write this new legislation to govern the pharmaceutical industry and make them hand over data?

If you like what I do, and you want me to do more, you can: buy my books Bad Science and Bad Pharma, give them to your friends, put them on your reading list, employ me to do a talk, or tweet this article to your friends. Thanks! ++++++++++++++++++++++++++++++++++++++++++

19 Responses

  1. Ben Goldacre said,

    March 6, 2008 at 9:01 pm

    yeah i already tried that.

    they’ll only regret ignoring me.

  2. Michael Power said,

    March 6, 2008 at 9:16 pm

    Shouldn’t we rebrand EBM as Transparent Health Care (THC)???

  3. Geeb said,

    March 6, 2008 at 9:20 pm

    So what exactly are they going to legislate?

    Will it be an offence to…
    – do any kind of medical trial on humans without registering it first?
    – change the design of a trial while in progress?
    – cancel a trial without publishing all of the raw data obtained so far?
    – conduct a trial without publishing the results *within a specified timeframe*?
    – withold the raw data, and only publish the “post-processing” results?
    – publish results from trials not conducted according to these rules?

    Unless they’re pretty heavy-handed with some of this stuff, I can’t help worrying that the only real effect will be the creation of a whole new industry of clinical lawyers specialising in finding regulatory loopholes.

  4. Ben Goldacre said,

    March 6, 2008 at 9:27 pm

    yeah, my opening gambit would be, alongside mandatory registering protocols etc

    1 demand publication of processed data,

    2 ask for raw data escrow, promising not to use it for anything too embarrassing, although it’ll make them squirm.

    be good for transparency of stats methods used, poss for exposing data mining, handy for subsequent metaanalysis, etc.

  5. Anthony said,

    March 6, 2008 at 10:10 pm

    “I can’t help worrying that the only real effect will be the creation of a whole new industry of clinical lawyers specialising in finding regulatory loopholes.”

    A whole _new_ industry? Are you kidding?

  6. frontierpsychiatrist said,

    March 6, 2008 at 10:50 pm

    @Maarten Van Hemelen

    In short, lots of things – thing is, the British government’s scheme is not simply for an ID card, but more for an identity register. You should visit for a full discussion about this, but essentially:

    – the government is proposing a national identity register i.e. individual checking and numbering of the population — marking many personal details as “registrable facts” to be disclosed and constantly updated and kept in a central database with provision for use across the private and public sectors.

    – Overseas ID cards are not comparable: Many western countries that have ID cards do not have a shared register. Mostly ID cards have been limited in use, with strong legal privacy protections. the records are not linked. Belgium has made use of modern encryption methods and local storage to protect privacy and prevent data-sharing, an approach opposite to the Home Office’s.


    – the Government has not made a case that the cards will actually will help with any of the areas one might hope such a scheme might do so for example terrorist attacks. It is likely that the scheme will actually make us less safe – we have seen that the UK government is rather clumsy when it come to data retention.

    – the scheme is massively expensive and relies on unproven technology

    – the scheme also raises questions of unchecked executive powers for instance the government could essentially ‘unperson’ a citizen by removing their card

    – there are examples of identity cards being instrumental for the purposes of discrimination (and worse) against minorities. For example during the Rwandian genocide identity cards, originally issued by the Belgians, were used to ascertain Hutu from Tutsi.

    No2id is encouraging British citizens to ‘take a pledge’ against ID cards – I encourage anyone who feels stongly about this to do so.

    On the other point of this posting – how about divorcing the conducting of drug trials completely away from the pharmaceutical companies? Trials could be conducted by a non-commerical organisation whom the pharmaceutical companies would be obliged to fund, thus removing the bias that creeps in to trials which are conducted ‘in-house’ by people who have a vested interest in their success.

  7. Mongrel said,

    March 7, 2008 at 8:42 am

    In another (baby)step forward the MHRA is implementing a “traditional herbal medicine registration scheme”. It’s aimed at safety and quality so still no efficacy requirement

    Summary here –

  8. elizabeth said,

    March 7, 2008 at 10:10 am

    “Demand publication of processed data”
    Let’s not get hung up on the word “publication”, because then the Elseviers will want in on the act. Certainly, the processed data should be posted on public access internet sites. Certainly, the raw data should be lodged, in full, with MHRA. And I mean raw, not recoded, since a very small do-file can hide a multitude of sins. But even this does not address two equally important issues. Firstly, the deliberate manipulation of study designs to ensure the best outcome for a drug. Secondly, the results of research that is of public health significance but of no interest to Big Pharma. British taxpayers have poured millions into HIV surveillance systems in other countries, for example, but even people trying to plan prevention and care programmes that will also be paid for by the British taxpayer often don’t have access to results. And God forbid that the results should be shared with prostitutes or drug injectors, the people who could actually do something about them.

    So please, when you are issuing your next commands to government, put in a plea for access to more than just drug trial data.

  9. Michael Power said,

    March 7, 2008 at 10:13 am

    Geeb asked a penetrating question “So what exactly are they going to legislate?”

    The MHRA want “absolute clarity in the legislation as to the information that must be supplied to the regulator”

    As the MHRA report begins with the disclaimer that “Legal constraints make it impossible to disclose any information … except … already in the public domain” it seems hard to imagine that the MHRA are pressing for research data to be made available to anyone othe than themselves.

    Seems like they want more opportunities for opacity.

  10. Squander Two said,

    March 7, 2008 at 12:14 pm

    I’m sick of mainland Europeans saying that they have compulsory ID cards and there are no problems with them. Yes, you do have problems, but you don’t think of them as such. We have a tradition in Britain that the state is answerable to the people. ID cards turn that on its head, which is why their introduction here is only part of a general legislative move against the citizen: getting rid of presumption of innocence in some cases, trial by jury in others, etc. In the UK, the police aren’t supposed to stop you unless they have genuine reason to suspect you of a crime. In France and Belgium, they can stop you to check your papers, i.e. to force you to prove that you have the state’s permission to walk down the street. Belgians may see no problem with that sort of thing, but, if they had any sense, they’d be fighting against it.

    I have a friend from mainland Europe whose life has been made enormously difficult for the last few years because people hostile to her have effectively brought about the cessation of her official existence. This is only possible when you need the state’s paperwork to prove your existence in the first place.

  11. Maarten Van Hemelen said,

    March 7, 2008 at 12:27 pm

    Eeehm… I don’t know if you’ve ever been to Belgium before, but the state is answerable to the people here as well. You do know about the whole democracy thing right, that you’re not the only ones that have it?

    And I can’t see how ID is linked to getting rid of the presumption of innocence. We don’t do that, it would turn the very principles of our idea of justice upside down.

    And noone will ever stop you to see if you have the state’s permission to walk down the street. They can ask for your papers, yes, but there are other reasons for that. For example, you do something wrong, they check you ID to make sure they fine you, and not your neighbour.

    Furthermore, it’s hardly polite to say we don’t have any sense for not fighting it. Ad hominems don’t make for good arguments, you know.

  12. tom-p said,

    March 7, 2008 at 1:44 pm

    Geeb @ #4 asked
    “Will it be an offence to…
    – do any kind of medical trial on humans without registering it first?”
    It already is, and has been since 1 May 2004, when directive 2001/20/EC entered into force. the database where they have to register is called EudraCT, the CT standing for clinical trials. This does not cover observational studies, where you’re simply looking at what’s happening, but does cover interventional clinical trials, where you randomise people or use blinded medication.

    Ben has previously written about (and alluded to suc here) the need for all sorts of approval before a doctor could perform a study on his patients whereby he randomised them to one of two groups where we don’t know which medicine is better.
    Now, aside from this probably being a very small study, and thus rather underpowered, it’s also an unnecessary protocol. If a GP knows that he has patients on a variety of different medicines for the same condition, he can go back and look at the records (and indeed proactively look at the records) to see which is working best (both for safety and efficacy), or he could farm out anonymised data to a friendly statistician who could then do the same without knowing which group referred to which product. The GPRD is already a tool for this, covering about 10% of the patients in the country.

  13. tom-p said,

    March 7, 2008 at 1:47 pm

    Martin, a cop stopping you to ask for your papers is being stopped to see if you have the stat’s permission to walk down the street.
    I don’t know about the police in Belgium, but over here the police have a long track record of harassing racial minorities and the power to stop someone on this spurious grounds is, many of us fear, likely to be used for just this purpose.

  14. lordluvaduck said,

    March 7, 2008 at 4:59 pm

    For any well-intended legislation, the devil is always in the details. And one new law will, as Geeb points out, bring a massive effort from Pharma to circumvent it.

    I’m optimistic though, that such legislation may be helpful in changing attitudes in the research community as well. We still have a ways to go in persuading reviewers and editors of juried journals that negative and no-difference results are important.

  15. thom said,

    March 7, 2008 at 8:16 pm

    Maarten: I’m not that strongly against ID cards. I think the case for ID cards is finely balanced – though being required to carry one or provide one on request is something I’d feel uncomfortable with (but might be persuaded on).

    My real object is to the governments proposed ID card scheme.

    1) It is _insanely_ expensive and every available piece of evidence I have seen suggests that it will be much more expensive to implement than the government estimates. (By insanely expensive I mean that any proposed benefits of of the scheme would be much more cost-effectively met by spending the money in other ways such as increased police, customs and intelligence service manpower).

    2) The technical difficulties of the scheme – as proposed – are immense. I’m not sure that it is technically possible to achieve the required results in the time frame they are proposing. I’m not an expert on this – but I have (in the past) done research on face processing (one of the proposed biometrics) and I still try and keep up with the literature in the area. Most likely the technical difficulties with be fudged – diluting any potential benefits but still retaining many of the risks (including lots of wasted investment and many of the potential errors).

    3) The dangers of the system have been massively underestimated. These include threats to civil liberties, data protection issues (for which this particular government has an appalling recent record) and the likely rate of errors (due to the aforementioned technical difficulties). A particular worry is that if they did get the system to work anyone who already has a false identity (or has stolen someone’s identity) could easily get on the system when it starts up.

  16. ControlFreak said,

    March 9, 2008 at 11:04 am

    Working in an the automotive industry, which is heavily regulated, I find it hard to believe that someting as fundamental as drug efficacy and safety are not judged by government specific tests, or at the very least by independent testing bodies. Vehicle safety and emissions all have to pass independently verifiable tests and in Europe the government(s) effectively ‘own’ the test result. In the UK, type approval tests are done by the “ministry” with their own staff and equipment.

    Some aspects are down to a manufacturer data submission but they risk a europe wide recall if they get it wrong so there is good incentive to be sure of your result.

    Of course things will always slip through and some products are designed just to pass the test and exploit loop holes. However manufacturers are now wary of not complying with the “spirit” of the regulations after some hefty fines from the EPA in the US.

  17. eingram said,

    March 21, 2008 at 4:55 pm

    I’d write the new legislationas follows.
    “Don’t do no bad stuff just because your’re greedy. Remember the golden rule and practice it.”

  18. eingram said,

    March 21, 2008 at 4:58 pm

    P.S. Just kidding in case anyone thought I was serious. I wouldn’t want to get flamed.

  19. Robert Carnegie said,

    March 24, 2008 at 2:05 am

    “Don’t be evil” would be an interesting constitution. And the constitutional court would be awfully busy.