Hard to swallow
Ben Goldacre's new book charts the pharma industry's poor research, missing data and excessive secrecy that amount to a 'murderous disaster'. By Kevin Gopal
Ben Goldacre's new book charts the pharma industry's poor research, missing data and excessive secrecy that amount to a 'murderous disaster'. By Kevin Gopal
Not a problem, insisted health minister Lord Howe early one morning last month on Radio Four’s Today programme. Clinical trials data wasn’t going missing. Drug companies had to disclose it by law.
His comments may not have shaken listeners out of their slumbers but by the time fellow health minister Norman Lamb admitted in parliament only a few hours later that more must be done to ensure publication of trials, the government had finally awoken to the fact that this was no obscure, technical question but one in which thousand of patients’ lives are at stake.
The charge against the pharmaceutical industry is that its research is poorly designed, that it buries negative results and that it has ensnared regulators, academics and the medical establishment in a web of complicity or neglect that has resulted in products that are ineffective or downright dangerous.
“Good science has been perverted on an industrial scale.”
The heart attack patients given heart drugs that killed them, the anti-depressants that make young people think about committing suicide, and the estimated £500 million bill handed to taxpayers to stockpile a flu treatment that’s claimed to be possibly no more effective than paracetamol – these are not new allegations but they’ve been forced on to the public agenda because they’ve been so thoroughly laid out in Dr Ben Goldacre’s new book Bad Pharma.
“The true scale of this murderous disaster only reveals itself when the details are untangled,” he writes. “Good science has been perverted on an industrial scale.”
Goldacre is quick to credit the researchers and campaigners who have long tried to prise out missing research but the physician, academic and self-confessed nerd – previously author of the popular Bad Science column in the Guardian and bestselling book of the same name – has made a complex subject more accessible.
The day of Howe’s comments, Goldacre had written a comment piece in the Times warning that “there is poison in the well of medicine” and would go on to co-sign a letter to the same paper, along with leading members of the scientific and medical establishment, saying that until the problem was tackled “patients will continue to suffer unnecessarily”.
It’s not all down to him but suddenly questions are being asked in parliament, the British Medical Journal is forcefully petitioning Roche to release data on Tamiflu, and British firm GlaxoSmithKline, the world’s second biggest drug maker, is promising to publish the raw data that sits behind the results of its trials.
Why has the scandal taken so long to be noticed? “How do you get the public to pay attention when everyone we rely on has failed – the public’s on a sticky wicket,” Goldacre, a research fellow in epidemiology at the London School of Hygiene and Tropical Medicine, tells The Big Issue in the North. “Regulators and medical bodies have done nothing. There’s this fascinating interlocking eco-system of failures that reinforce each other.
“I know this, but even I learnt something as I wrote it all down. And it’s nobody’s day job to do this – I wrote the book in the evenings.”
Bad Pharma reveals the problem to be systematic, not an unfortunate series of unconnected events. Goldacre cites research showing that half of all completed trials on medicines go unpublished, and that trials with positive results are about twice as likely to be published as those with negative ones.
Then the disasters, including reboxetine, the anti-depressant Goldacre had himself prescribed until previously unpublished data revealed it to be no better than placebo (a dummy pill), with worse side-effects than other anti-depressants. The manufacturer Pfizer had published one study showing it was better than placebo – and sat on six saying it wasn’t.
“In reality, it was no better than a sugar pill, and worse, it does more harm than good,” he writes. “As a doctor, I did something which, on the balance of all the evidence, harmed my patient simply because unflattering data was left unpublished.”
Six volunteers agreed to test a new drug, TGN1412, in humans for the first time in London in 2006. Within a day all developed horrific symptoms, including fluid on their lungs, failing kidneys and uncontrollable blood clotting, and were ventilated on intensive care. The volunteers just escaped with their lives but might have been spared their ordeal if research on a similar drug a decade earlier, which found similar problems, had not languished unpublished.
Global pharmaceutical industry sales neared a trillion dollars in 2011
Most seriously, 100,000 post-heart attack patients are estimated to have died after being prescribed anti-arrhythmic drugs in the 1980s. It was assumed to be a sensible treatment, since heart attack patients are known to have dangerous abnormal heart rhythms. But when three anti-arrhythmics were finally fully tested the increased risk of death they caused for heart attack patients was so great the trial had to be halted.
Global pharmaceutical industry sales neared a trillion dollars in 2011, according to IMS Health. Annual sales of the world’s best selling drug, Nexium – a treatment for gastrointestinal disorders – are an estimated £900 million. And the industry is finding it much harder to come up with new drugs than in the golden years its executives fondly recall. No wonder they are so keen to get their products on to the market.
But Goldacre reserves plenty of his ire for others – academic institutions that allow pharma industry clients to tie researchers up in confidentiality agreements, and regulators who feel their obligation is to manufacturers, not patients, and so are excruciatingly tight in releasing information – and calls it “an extraordinarily shameful situation”.
One astonishing passage in Bad Pharma relates a near four-year period of stonewalling by the European Medicines Agency (EMA) when researchers from the respected Nordic Cochrane Centre tried to get hold of data on two diet drugs they were reviewing.
The EMA, responsible for authorising medicines in the EU, refused, saying it was protecting the commercial interests of the manufacturers. The researchers complained to the EU Ombudsman, who told the EMA to hand over the data. Still the regulator refused – during which time one of the drugs, rimonabant, made by Sanofi-Aventis, was even taken off the market due to the risk of serious psychiatric problems and even suicide. When the EMA eventually relented it told the researchers the data was in fact held by the UK regulator, the Medicines and Healthcare products Regulatory Agency. The MHRA said it had shredded it all.
“The scandal is a million times worse than MPs’ expenses or phone hacking.”
Then there are the medical bodies that refuse to condemn research misconduct – “the silverbacks of the profession have failed us”, says Goldacre – and whose members prescribe drugs produced by companies that have offered them hospitality or sponsored training. Oh, and patient associations that purport to be independent but are largely funded by, guess who, big pharma.
“The scandal is a million times worse than MPs’ expenses or phone hacking but structurally similar. Some people should be in prison. We need the public to peer in and all hell break loose.”
Goldacre stresses that the industry has long produced valuable drugs and saved countless lives “on an epic scale”, and that good people exist among companies and regulatory agencies. There is corruption, but he almost bends over backwards to emphasise that it’s as much about people having no interest in identifying misconduct, or even just finding it too difficult to understand.
The European Parliament heavily criticised the EMA for allowing its executives to take up highly paid posts in industry almost immediately after leaving the agency. It said the EMA failed to take seriously the potential conflicts of interest resulting from the revolving door phenomenon. But cosy relationships between industry and regulators can sometimes develop simply because people on both sides of the fence like to talk work stuff with each other.
“It can be people being insightless,” says Goldacre. “What I’ve found from 15 years of being a grown-up is that competence looks better from afar. When I’ve talked to even senior regulators, some of their answers have not made sense intellectually. But there’s a different blend of drivers for each individual.”
Goldacre’s book has forced the industry – one of the UK’s leading exporters and investors in research and development – on to the back foot. In a statement, the Association of the British Pharmaceutical Industry said “the examples he refers to are long documented and historical, and the companies concerned have long addressed these issues”.
Goldacre insists that takes no account of Roche continuing to withhold information on Tamiflu. “The ABPI saying it’s in the past is very simply and demonstrably a false statement. It’s an extraordinary state of affairs – there are lots of very ethical people in the industry and they are badly served by the old-fashioned PR of the ABPI. You can see everything about this problem in that statement from the ABPI.”
An ABPI spokesperson told The Big Issue in the North that it is already best practice for its members to publish all data, positive and negative, that it recognises “there is still work to be done by all stakeholders” in ensuring greater transparency, and that it is “working collaboratively with the wider healthcare sector to achieve this”.
The spokesperson added that the industry had supported trial registries in the EU and the US, and was part of the Ethical Standards in Health and Life Sciences Group, which is consulting on medical education for doctors sponsored by the industry and on the disclosure of financial relationships between healthcare professionals and commercial organisations. But the ABPI did not directly answer when asked if it would support the mandatory disclosure of all trial data.
One ABPI member, GlaxoSmithKline – the world’s second biggest pharma company – responded to the pressure last month by promising to publish on its website detailed patient data as well as the trial results it currently puts up.
But this is the company that earlier this year was fined £1.9 billion by US authorities after admitting three misdemeanour criminal charges, including marketing the anti-depressant Paxil to young people when it was only approved for adults, and failing to stump up safety data about its diabetes drug Avandia.
Goldacre is suspicious of broken promises and what he calls “fake fixes”
And GSK has made similar promises before on releasing data, setting up its own clinical trials registry in the late 1990s before it was shortly shut down again following a change in chief executive.
Goldacre is suspicious of broken promises and what he calls “fake fixes” but nevertheless takes heart from recent moves to more openness. He urges bigger, better and more transparent trials, says doctors should declare links with the industry and patients should demand more information on the way the industry works. Health ministers have agreed to meet the signatories of the letter to the Times to hear more about their concerns.
“I now feel very hopeful that the medical and academic professions will do something to address these serious threats to patient safety,” says Goldacre. “I hope that some of the many ethical professionals who work in the pharmaceutical industry will take a stand too. There are too many people in their community who pretend that these problems don’t exist. Patients are harmed by their silence.”
Perhaps it was no surprise that governments around the world rushed to stockpile Tamiflu when faced with dire warnings of the possibility of a flu pandemic.
The UK government started to place bulk orders with the manufacturer Roche in 2003 and is estimated to have spent £500 million.
Tamiflu isn’t a vaccine but was claimed to be able to ward off pneumonia and other complications associated with flu, and to slow down its transmission.
Even the independently minded medical researchers at the Cochrane Collaboration once believed that to be the case, as they said in their 2006 review of the drug.
But publication of that review brought a comment from a Japanese researcher, who pointed out that the Cochrane researchers were drawing their positive conclusions from industry-funded analysis that summarised 10 trials. But only two of those trials had been published.
Since that “serendipitous” intervention, Cochrane researcher Dr Tom Jefferson tells The Big Issue in the North, “I have been atoning”.
Atonement has been pursued through a lengthy battle with Roche to get hold of the data. It’s a battle that has been joined by the British Medical Journal, which has published reams of correspondence between the Swiss company and the researchers because of what it says are the drug maker’s broken promises.
“Sixty per cent of the treatment trials are unpublished,” says Jefferson. “We have part of that but don’t have it all.”
He says so far there is “not a shred of credible evidence” that Tamiflu reduces complications and hospitalisations, and that “no one in government, the World Health Organisation, wherever, properly looked at the evidence” that it reduced the transmission of the virus.
In a statement, Roche said: “Roche complies with all legal requirements regarding the publication of data and does not generally make patient level data available due to legal and confidentiality constraints. Roche provided the Cochrane group with access to 3,200 pages of very detailed information, enabling their questions to be answered. The Cochrane group asked more questions and requested further data, but declined to sign a confidentiality agreement.”
Jefferson says it is “impertinent” for Roche to tell researchers what data they need
It says more detailed clinical trial reports are available for use by investigators on a password-protected site.
But Jefferson says it is “impertinent” for Roche to tell researchers what data they do or do not need, and finds the idea that independent researchers should sign a confidentiality agreement ridiculous. After all, it was open publication of data that brought the Japanese researcher’s critical intervention. Jefferson says Roche has not moved its position since 2009.
“The reality is governments have got a problem in backtracking,” he adds, because they might have to admit they have been hoodwinked.
“Who is responsible for this? British taxpayers in the midst of this economic crisis are being asked to fund this intervention.”
Publication of research in peer-reviewed journals is a cornerstone of science. Before papers are published they are checked by other scientists, and once out in the open the results can be further scrutinised.
In 2004 the International Committee of Medical Journal Editors agreed that none of them would publish clinical trials that hadn’t been pre-registered. The aim was to ensure all trials were registered and stop the pharmaceutical industry and researchers from selectively publishing only the best results.
But in 2008, researchers found only half of the 323 trials published in the top ten journals were adequately registered and a quarter weren’t registered at all. “The ICMJE editors had simply failed to keep their word,” concludes Goldacre.
Fiona Godlee, editor of the British Medical Journal, puts that down to a flaw in implementation rather than a lack of will.
“The BMJ’s commitment to pre-registration of trials is absolute.”
“ICMJE’s policy was a turning point in trial registration and should be credited with changing the culture, leading to a sizeable and lasting uplift in the number of trials registered each year,” she says.
“I don’t think anyone has looked at this since then but, speaking for the BMJ, the commitment to pre-registration of trials is absolute and we no longer allow authors to plead ignorance. If the trial was not registered from the start, we will not publish it.”
Last month Godlee said the BMJ would only publish trials if researchers committed to make the detailed, anonymised patient data on which they were based available to other researchers.