A lot people it seems have been singing and dancing in the street at news last week that AbbVie the pharmaceutical company that make Humira have dropped a legal action against the European Medicine’s Agency (EMA) that was aimed at blocking any access by researchers or others to their clinical trial data.
Stacy of Arc
Not everyone will be pleased though. Stacy London might wonder whether she should fire her advisers. She could have featured as a Joan of Arc leading a RxISK campaign against oppression had she responded to our Open Letter to Stacy. She might be sitting pretty now had she adopted a position that Humira can be a tremendous drug but hiding clinical trial data is just not acceptable. But we got no response, and it didn’t seem fair to harass her.
AbbVie’s action was set in train several years ago when one of medicine’s most extraordinary figures, Peter Gøtzsche from the Danish Cochrane Center took a case for access to clinical trial data to the European ombudsman and won a ruling that the European Medicines’ Agency should make the data from clinical trials submitted to them for regulatory purposes available more generally. This was a wonderful breakthrough. In November 2012 EMA convened meetings to work out the ground rules for what was expected to be the new way of doing things.
But on April 30th 2013 just as the post Brand Fascism was posted on dhblog.wpengine.com saying that we might all be better off if company efforts to hide their data were successful, a legal action was taken in the European Court against EMA’s data access by two companies. AbbVie and InterMune.
Rights to privacy
AbbVie applied to block access to two clinical study reports that focused on the safety and efficacy of their blockbuster Humira, responsible for 70% of the company’s profits. The reports were on its use for the treatment of Crohn’s disease. Another clinical study report concerned safety risks linked to combinations of Humira and corticosteroids.
InterMune tried to block disclosure of data on its only drug Esbriet, an anti-inflammatory drug it hopes to push for the treatment of a lethal – largely drug induced condition – idiopathic pulmonary fibrosis. Esbriet might weirdly become the first drug to become a blockbuster out of managing an adverse effect caused by other drugs.
AbbVie and InterMune claimed that the decision to provide access to all clinical data violates their’ ‘fundamental right to the protection of confidential commercial information’, and their ‘legitimate expectations’, and copyrights they hold. The bottom line is company rights to privacy are being breached.
RxISK petition
This legal action led to an enormous brouhaha. RxISK got involved and over 6000 of you from over 120 countries signed our petition calling on AbbVie and InterMune to drop their action – See Extreme Petitioning.
There is a wonderful moral case to be made for transparency of clinical trial data and a compelling case to have a petition to raise awareness of the issues more generally but as we made clear in a number of posts almost anything in this area is hazardous.
Welcoming AbbVie’s apparent dropping of their action feels terribly like the Trojans welcoming the apparent departure of the Greeks and opening a hole in the City Wall to haul in a Wooden Horse. We scream at them not to.
Trojan horse
The press coverage of AbbVie’s withdrawn legal action suggests that most major companies have now embraced an option for transparency pioneered by GSK.
All Trials are among those taking credit for pressuring AbbVie into submission. They have aggressively welcomed the offer by GSK to make clinical trial data available with no questioning of the terms on which the data is being produced.
But GSK’s offer is a manoeuvre worthy of Ulysses himself. You’d never guess from company self-congratulation that it was forced on GSK by a New York Court as part of the resolution of a Fraud action. The Fraud Action happened because GSK had written up a positive portrayal of a trial when in fact the company itself thought the trial had shown their antidepressant, Paxil, did not work.
The trial – Study 329 – was test of Paxil in children. As mandated by the Courts in the wake of Study 329, GSK put up company study reports for all of their trials, Paxil and other drugs including the diabetes drug Avandia. These could be downloaded. Steve Nissen of the Cleveland Clinic did just this for the Avandia reports and was able to show that Avandia killed. A company blockbuster was stopped in its tracks.
Poacher turned game-keeper?
Putting study results up on the web must have seemed like a very bad idea to GSK. So why are they now championing data access?
GSK and other companies are reaping kudos for apparent transparency. And they can say with a straight face to any TV anchorman or Congressional committee that they are making data available.
But in fact here is what is happening. Having seen what happened with the Avandia study reports GSK now know what to do when writing a Study Report to avoid a repeat. Suitably Doctored Study Reports for other drugs will go up on their website.
The Study Reports however do not contain the data. A first approximation to the data in the case of Study 329 comes in a series of 7 appendices to the 329 Study Report – something GSK did not put up on the company website until the omission was spotted nearly 10 years later by Peter Doshi and New York State required them to do so.
In the case of Study 329, the ghostwritten article that led to GSK being sued for fraud, is 11 pages. The Study Report is over 700 pages. There are then 7 appendices that between them come close to 5500 pages. Even this however is not the raw data. The raw data lies in Clinical Report Forms.
You can look but you cannot see
As things stood before GSK’s offer of transparency, the 5,500 page of appendices and 700+ pages of the Study Report could be downloaded and printed off. Finding what went on in a clinical trial from paperwork like this is a bit like playing Memory – where there a bunch of cards with faces or plants or whatever turned face down and if you turn one up you have to remember where the matching face you turned up before is. It can be done with 5,500 pages printed off.
But playing Memory is much harder to do now with the new improved access GSK is offering.
If you apply to access a GSK trial now you are forced to submit an analytic plan which essentially stops an applicant from accessing any adverse events on the drug. Adverse events are the material the company tries hardest to hide.
Should you get access to the full set of appendices that contain company listings of adverse events, there is almost no way to play the Memory Game because access is through a remote desktop. It may be that a younger generation used to playing Digital Memory will be able to work the system, but it’s not easy. It takes multiple passwords to access the desktop. You are logged out regularly. And while on the desktop, GSK can monitor your every keystroke.
Nightmare in Harlow
But here’s the rub. To really nail down what’s going on, you need access to the approximately 70,000 pages of patient level data. Through a remote desktop this becomes a nightmare.
This scheme to deliver frustration cloaked in the appearances of transparency was devised several years ago. The history of the idea was outlined two years ago in May Fool’s Day. Last year the details of GSK’s scheme were outlined in April Fool in Harlow.
Selling a message to the media and public that the appearances of more transparency are being used to hide the data is not an easy sell. Telling them that companies can devise trials that use a problem their drug causes to hide a problem their drug causes is something that makes the machinations of House of Cards seem innocent. A nice company like GSK or AbbVie wouldn’t do this.
Let’s do the AbbVie again
But we are not faced with an insoluble problem. Company efforts reveal their Achilles Heel. They are vulnerable to adverse event data. And this is where the original RxISK campaign to get people to do the AbbVie comes in. The more people who report adverse events in good quality reports and get their doctors on board in helping them grapple with the issues, the more between us we can restore sanity.
This led to the creation last year of an AbbVie – the opposite to a Boycott.
We asked you to pay close attention to these drugs – shine a spotlight on them – and communicate with them and each other rather than shun them. If you are part of any of the patient organization dealing with Crohn’s Disease or Ulcerative Colitis, such as the Great Bowel Movement, or arthritis you may be able to spread the word rapidly and mobilize many people. If you are Stacey London, Phil Mickelson or Mick Jagger you could make a big difference – Mother’s Little Poisoner.
The beauty of this that unlike the nineteenth century Irish tenant farmers who had to endure the hardship of not earning a living off Boycott’s farms, doing the AbbVie will not hamper your treatment or access to treatment in any way. Just the opposite – it will improve your treatment.
A good medicine is a chemical plus good information. GSK, Lilly, Pfizer and AbbVie want to degrade the quality of your medicine in particular and all our medicines in general by restricting access to key information on the full range of things these drugs do. We need to tell them this is unacceptable. The best way to do this is to share unexpected events on treatment in a forum they can’t block and to shout louder when told to keep quiet.
Illustrations: © 2013 created by Billiam James
I’m a bit confused. Does AbbVie’s agreement to withdraw the lawsuit itself contain a Trojan horse? If so, what is the exact nature of the beast they are smuggling in? AbbVie itself has put out a press release making a claim something like this … “we weren’t defeated, we simply negotiated a satisfactory settlement.” Apparently the EMA did agree to redact a few bits of those controversial study reports before releasing them. However, the EMA has flatly rejected this idea of a “settlement”, saying the redactions were minimal. My impression was that AbbVie was just doing what all corporations do when they lose in court – spin it so it looks like they got what they wanted, for a reasonable price. But maybe they took more ground than I realize?
Or are you just saying, like Malcolm X, that we can’t chase away the wolf howling at the front door, only to let the fox sneak in the back door? (To Malcolm the wolf was the reactionary politicians in DC, and the fox was their so-called liberal colleagues. In this case, AbbVie has been playing the wolf while the leading fox is GSK with its “responsible data sharing” schemes.)
In that case, let’s be on guard – but those who are celebrating are not necessarily fools. We won one battle – just one, and probably not the Big One – in a long war. But I can’t see that we’d have been better off if we had just let AbbVie gobble up whatever it wanted.
I missed any mention of the recent announcement in the New York Times by Dr. Harlan M. Krumholz: “LAST week, Johnson & Johnson announced that it was making all of its clinical trial data available to scientists around the world. It has hired my group, Yale University Open Data Access Project, or YODA, to fully oversee the release of the data.” It is not clear whether the raw data of all J&J clinical trials will be made available. Will they be? Or will they be subject to the kind of restricted access that you describe in this post about the GSK “release of data”?
I have struggled a bit with this. I was celebrating AllTrials and Goldacres Bad Pharma as significant moves in the right direction. I never really understood how any of this could be a bad thing when it comes to the safety of our medicines. And a year ago I would have been celebrating the AbbVie outcome too. But that has changed now and I realise my support for all trials was for my own selfish reasons.
I was sick of being called a conspiracy theorist, stupid, paranoid and all the rest. All trials and Bad Pharma made my views more mainstream, which helped me a lot with my family, friends and colleagues. All of a sudden I wasn’t nuts in their eyes, I was ahead of the game. It also helped when I next visited my Doctor, as the facts I put to her some 4 years previous (facts she denied) were now out in the open and appeared to be widely accepted.
But now the whole thing just seems like a case of history repeating. A way of reinforcing or reinvigorating the role of RCTs in decision making, keeping control firmly in the hands of the people who have enough money to run these trials. All these public cases are being used to appease the growing body of dissenters while simultaneously ensuring that industry keep control over the all important data. Meanwhile the position of RCT’s being the gold standard of evidence gets a massive boost with all the new transparency going about. Sorted!
I see the whole pharma transparency play as similar to what’s happening with the long awaited report on Iraq in the UK. Why is Tony Blair so eager to see a report that should be damning of his actions? That might be because the report keeps the real information secret, with only this independent panel being able to see it on our behalf. They will no doubt water that down for the public so it causes no more than a bit of mild outrage, and we all go away happy thinking the truth is out. Our government was a little bit naughty but no more than that. Nothing criminal. Job done, move along please, nothing more to see here.
My take on this whole transparency situation is a little more ‘Hero’ than Trojan Horse. As in the 2002 film with Jet Li. With Big Pharma as the Qin Dynasty, Andrew Witty as King Qin Shi Huang, Goldacre and all trials as the Namesless warrior and the BMJ, Cochrane and the EMA as the assassins Long Sky, Snow and Sword.
The question is, was the nameless warrior a genuine assassin intent on revenge for the other territories of China, who in the end was convinced by the wise King Qin that uniting under Qin was best for everyone and would bring peace and prosperity to all the land? Or was the namesless warrior part of the Qin establishment from the beginning and just played his part an ingenious plot to bring down the dangerous assassins and crush the dissent once and for all?
Is Ben Goldacre the modern version of this story – is our hero a hero?
AllTrials have done a huge amount to bring these issues to wider awareness and this is wonderful But now is the time of greatest peril. Having done so much good, is there a risk of being outmaneuvered by GSK and other companies. It would be more reassuring to feel that Alltrials were aware of this risk.
In the story Hero – the protagonist is a Hero one way or the other. Not such a good story for Naivety to land us all in a mess.
Naivety or something else…
The end result is the same…
Glad that I’m not the only one who has noticed …
It appears that the drug companies have had the luxury of distorting clinical trial data for such a long time that they think it’s a right guaranteed in the Magna Carta. One can’t help but conclude that AbbVie withdrew their suit after extracting this agreement with the EMA to keep Clinical Trial data in the fog. And the only reason to do that is to be able to continue to distort it ad infinitum…
Dear Dr. Healy,
– About Humira: https://www.psoriasis.com/ – “Get the right dermatologist”
Re: here is data – (anyone can look and see) – undisclosed liaisons between industry (pharmacological & non-pharmacological technologies) and specific medical professions and careers (e.g., click URL below).
http://www.propublica.org/search/search.php?qss=dermatologist&x=0&y=0&csrf_token=de026ba69ff6da50b3dc4560cd73620045fd8ade3a42d55c95ce491f0aeb3400
There are other types of data (i.e., we can look and see)
http://www.bmj.com/content/347/bmj.f1880/rr/763200
The number of unpublished trials is larger than 50% in most of my estimates (ClinicalTrials.gov & WHO ICTRP) – raw data is available for independent scrutiny (anyone can look and see, and perform corroborative analysis).
http://chaoticpharmacology.com/?s=tamsulosin
http://chaoticpharmacology.com/?s=paroxetine
http://www.bmj.com/content/347/bmj.f4386/rr/763130
http://www.carlheneghan.com/2015/723/evidence-live-2015-analysis-of-us-phase-3-clinicaltrials-gov/
The exact percentage it’s also uncertain.
– “around half” is probably underestimation according to data:
Tamsulosin >75% unpublished
Dabigatran >80% unpublished
Agomelatine >80% unpublished
Liraglutide: 68.5% – three times as likely to be disseminated (3.1 to be specific) & sample sizes doesn’t make any sense.
But one thing I am almost certain: it is above 50% of unpublished trials.
We can actually look and see deep inside some clinical trial registries (e.g., RE-LY clinical trial – JUPITER trial with rosuvastatin – antidepressants).
http://www.bmj.com/content/349/bmj.g4747/rr/778288
http://figshare.com/articles/Database_of_human_studies_evaluating_antidepressant_drugs/1168926
Probably (almost certain about this) the terms “drug regulatory agencies” (e.g., INVIMA – Colombia) and some centers for the evaluation of healthcare technologies (e.g., IETS – Colombia) are a euphemism.
http://chaoticpharmacology.com/2015/04/21/about-the-self-regulation-of-the-medical-profession/
“self-regulation is a myth not likely to be realized outside utopia” – Dr. Charles D. May, 1961.
Probably (almost certain about this too) the real regulators are those who expose practices of the wrongdoings and do it despite the potential consequences of facing attacks aimed to suppress ideas threatening vested interests.
Sincerely,
Jorge R.
P.S. I read with interest the title of your post “Welcome to Troy”, I say: welcome to chaos (there are other types of data).
https://clinicaltrials.gov/ct2/results?term=&recr=&rslt=&type=&cond=&intr=&titles=&outc=&spons=GlaxoSmithKline&lead=&id=&state1=&cntry1=&state2=&cntry2=&state3=&cntry3=&locn=&gndr=&rcv_s=&rcv_e=&lup_s=&lup_e=
https://clinicaltrials.gov/ct2/results?term=&recr=&rslt=&type=&cond=&intr=&titles=&outc=&spons=GSK&lead=&id=&state1=&cntry1=&state2=&cntry2=&state3=&cntry3=&locn=&gndr=&rcv_s=&rcv_e=&lup_s=&lup_e=
–
Jorge
Many thanks for this – I hope many will explore these links and would encourage you to add them as a comment to another more recent post also just to increase the chances of people becoming aware of them.
The only point I’m confused about is the access to data point – there is no access to clinical trial data. Published articles represent something about the data but without access to the data and in the case of the SSRIs and suicide in children the published claims were diametrically opposite to what the data showed.
David
Dear David,
Thank you too, for all of your writings, as well as for replying to my comment.
Regarding the question of clinical trial data access:
Clinical trial data have variable levels of accessibility:
– Study results (but not their raw data – data of industry-sponsored studies is not open at all: it’s locked up behind a lot of bureaucratic requirements and excuses).
– Historical changes in the study registry.
– Publications (sometimes interesting things can be found comparing their results and methods).
– Competing interests of authors.
– Other possibilities…
It mainly depends on the condition and the intervention.
For example,
1. Antidepressants and suicide in children.
A. ClinicalTrials.gov = 10 studies with results (outcome: suicide; intervention: antidepressants; age group: children) – the results of these studies can be also downloaded (*.xml).
We can look and see – data can be also matched with the results of publications – inconsistencies can be documented.
https://www.clinicaltrials.gov/ct2/results/displayOpt?flds=a&flds=b&flds=f&flds=c&flds=h&flds=t&flds=g&submit_fld_opt=on&rslt=With&intr=antidepressants&outc=suicide&age=0&show_flds=Y
—
Database: https://docs.google.com/spreadsheets/d/13qRAv_QaJTYWQFYWoThO1AQwZMlyIuAI8qvrjg7yny4/edit#gid=2033799350
…
Largest patient enrollment (n=2272): NCT00605865 – observational – sertraline – unpublished.
Completed suicide † 1
# participants affected / at risk 1/2157 (0.05%)
“Other disclosure agreement that restricts the right of the PI to discuss or publish trial results after the trial is completed.
Restriction Description: Pfizer has the right to review disclosures, requesting a delay of less than 60 days. Investigator will postpone single center publications until after disclosure of pooled data (all sites), less than 12 months from study completion/termination at all participating sites. Investigator may not disclose previously undisclosed confidential information other than study results.”
—
Study results (NCT00605865) database – please look and see.
URL: https://clinicaltrials.gov/ct2/show/NCT00605865
Database: results can be downloaded and archived in a database
Study History: https://clinicaltrials.gov/archive/NCT00605865
PubMed matches = zero – http://www.ncbi.nlm.nih.gov/pubmed/?term=NCT00605865
Google Scholar matches = zero — https://scholar.google.com.co/scholar?hl=en&q=NCT00605865&btnG=&lr=lang_en&as_sdt=1%2C5&as_sdtp=
Notes:
– We can see on “the surface” but depth scrutiny is not possible because patient anonymized raw data is not available – I believe that the BMJ Open Data campaign is more coherent in that aspect than AllTrials – but both (I think) are mistakenly relying on clinical trials as the best level of evidence that we can get.
– I believe (almost certain about this – I am working on the paper for peer-review) that clinical trials are not the highest level of evidence.
http://chaoticpharmacology.com/2015/03/28/evidence-based-medicine-ebm-not-evidence-based-medicine-is-broken-2/
http://chaoticpharmacology.com/2015/04/17/hierarchical-levels-of-evidence-based-medicine-are-incorrect-evidencelive-2015-poster/
– Other sources of data are also very valuable (e.g., I used the data of quetiapine on RxISK.org in this post – quetiapine / Mark Markingson case – URL below)
http://chaoticpharmacology.com/2015/04/04/psychiatric-research-misconduct-at-the-university-of-minnesota-cafe-study-nct00034892-more-information-about-quetiapine-via-rxisk-org/
– It’s possible to get some valuable information scrutinizing all of the remaining nine studies (antidepressants, children, suicide, available results) and comparing their data with those of the publications or documents reviewed by drug regulatory agencies (e.g., FDA and EMA).
B. We can look – but we cannot see:
Two (n=2) studies without results:
https://www.clinicaltrials.gov/ct2/results?recr=Closed&rslt=Without&intr=antidepressants&outc=suicide&age=0&show_flds=Y
—
C. There is still pending to take a look to WHO ICTRP (more data there – as well in the individual registries – eg. EudraCT).
D. It’s like being a little more certain about the uncertainties related to unpublished studies of clinical trials (I also believe that data of observational studies and reports of adverse events – pharmacovigilance – could be equally important to trials). Nonetheless, it’s impossible to know all the hidden data (trials, observational studies, and pharmacovigilance reports) there is evidence of studies unregistered but published (violating the requirements of the ICJME) – http://www.bmj.com/content/348/bmj.g1888/rr/762419 –
but the question should be: what about unregistered and unpublished studies?
It’s like we cannot look and neither see – it’s like the dark matter in the space.
http://chaoticpharmacology.com/2015/03/19/dark-matter/
Finally, it’s also possible analyze the quality of reporting of these studies, using criteria such as the CONSORT for trials or the STROBE for observational studies, this information could be also very valuable.
I am aware of the Maudsley debate about psychiatric drugs, unfortunately, I cannot be there physically (but hopefully I will be there virtually) http://www.kcl.ac.uk/ioppn/news/events/2015/may/52nd-Maudsley-Debate.aspx
Sincerely,
Jorge R.