This morning at 11:30, Dr Sarah Wollaston MP will ask questions in parliament about the ongoing scandal of missing trial data. This is widely recognised as a problem by academics and doctors, but governments, regulators, and journalists have neglected the problem, while industry simply denies it. Watch the questions live here or watch it later here.
As an example, we spent £500m stockpiling Tamiflu in the UK, but the company Roche are still withholding vitally important information about the trials on whether it works from Cochrane, the international academic collaboration who make gold standard reviews of evidence for doctors, patients, and governments.
Apparently this morning the Department of Health issued a statement to the Today Programme stating that all data from all trials on all drugs is already available by law. This is not true: but there may have been a communication problem, I am hoping to obtain a copy of this statement, and if you have one, please send it to firstname.lastname@example.org
Here are two articles from the Times today, on this issue. The first is a comment piece (by me). The second is a news piece about the growing number of doctors and academics speaking out on this issue, including editors of major journals, founders and directors of Cochrane, senior medics and academics, etc.
The question was raised, and answered by Norman Lamb. He said that he recognises there is more to be done on increasing access to information about trials, and says that he or Lord Howe will meet with us to hear our concerns.
He also, worryingly, says that the European Clinical Trials Register will fix these problems. This is not true, for two reasons. Firstly, it would only contain the results of new trials, while we need the results of all trials. Secondly, as I show in Bad Pharma, the EMA has failed even to produce a complete list of all the trials conducted in the EU, on their Clinical Trials Register, despite years of promises. In fact, for many years, this trials register was held entirely in secret, while now it is merely hopelessly incomplete.
An organisation that creates a transparency tool but keeps its contents secret cannot expect their promises to be taken seriously. We must hold their feet to the fire. Or in the words of Dr Sarah Wollaston MP:
I am also seriously concerned to hear that on the Today Programme the Department of Health issued a statement claiming that all trial information is already available. This would be an untrue claim. If their mind changed between last night and this morning, that may be because of the coverage in the Times, so for that I am grateful.
Here is a briefing note.
Missing trial data – Briefing notes.
Doctors need the results of clinical trials to make decisions. Unfortunately, drug companies and researchers are allowed to withhold information about clinical trials from doctors and patients if they wish to. This means that we are misled about the benefits and risks of treatments. We can be misled into prescribing an expensive new drug, for example, when in reality an older cheaper one is more effective. Patients are harmed and money is wasted.
This problem is extremely well documented in medicine, and widely discussed within the profession. Legislators have failed to engage on the issue. Our current best estimate is that half of all the clinical trials that are conducted and completed are never published in academic journals, and trials with positive results are twice as likely to be published.
Is there a specific example?
Tamiflu is a drug the UK government have spent £500m stockpiling. But Roche is currently withholding vitally important information about trials on this drug from the Cochrane Collaboration, the large international non-profit academic collaboration that produces gold standard systematic reviews of all the evidence on a given drug for doctors and patients. However, this is not an isolated case, and the evidence above shows that this problem is widespread throughout the whole of medicine.
What needs to change?
We need to ensure that all results of all clinical trials – past and future – on all drugs in current use are available to doctors and patients, so that they can make informed decisions about treatments.
Why does it matter if some trials are withheld?
There is always some variation in the results of trials done on any medicine, simply through the play of chance. If only the more flattering results are published, doctors and patients are misled about benefits, and harms can be hidden.
What has been done about missing trial results?
Not much, and what has been tried, has failed, due to lack of political will. The US government passed a law (the FDA Amendment Act 2007) which said that all results must be posted on the website clinicaltrials.gov within a year of completion of the trial. Industry claims that this solved the problem. In reality it has not. An audit was published in the British Medical Journal in 2012, which showed that only one in five trials had met this reporting requirement. Despite this fact, no fine has ever been levied against any company or researcher for failing to post results. Even if a fine had been levied, it is only $10,000 a day, or $3.65m a year: this is low for a drug company with annual revenue in the tens of billions.
Even if this law had worked, it would not have fixed the problem. We need access to all the results of all the trials on all the drugs in current use, not just the trials completed after 2008.
Are industry the only ones at fault?
Universities have failed to ensure that contracts with drug companies sponsoring trials run by academics allow the academics to publish the results of trials regardless of whether the company is happy with them. Ethics committees who give approval for research projects have failed to protect patients, because they do not insist that researchers publish results, and they do not check to see if researchers are withholding the results of previous trials. Medical membership bodies have failed to act on this issue, and have failed to state publicly that withholding the results of clinical trials is research misconduct. University academics also sometimes fail to publish results, though they represent a minority of all the trials research that is conducted.
Industry sometimes claim that medical journals will not publish negative trial results. This was a modest problem in another era of medicine, but was fixed a decade ago. There are several academic journals that will publish trials regardless of whether the results are positive. Academic journals are no barrier to publishing trial results.
Don’t the industry say that all trial data is given to regulators?
They do, but this is a false claim. For example, different regulators in different parts of the world were given different information about different trials on Tamiflu. However it is true that regulators and NICE get more information than is shared with doctors and patients.
It is also not enough, because regulators and NICE receive this information under commercial confidentiality clauses. The European Medicines Agency now regulate medicines for the UK. They fought for three and a half years to withhold information about trials on two weight loss drugs called Orlistat and Rimonabant from Cochrane, the international academic group who summarise trial data for doctors. The European Ombudsman made a finding of maladministration as the excuses given by the EMA were inconsistent and invalid. The agency has promised to revise its practices. However they have made and broken promises about transparency before.
Furthermore, the new draft of the European Union Clinical Trials Directive does not improve transparency, if anything, it makes things worse. This will become an important story over the next few months, as the bill passes through the European Parliament, and especially when the Rapporteur makes her first response.
Why do doctors matter, surely regulators decide if a drug works?
A medicine does not simply “work” or “not work”. Some drugs work very well, some work less well than other drugs, but are still better than nothing. A medicines regulator decides if a drug should go on the market at all, and they have a low bar for approval. This is a good thing: we need some less effective drugs to come on the market. For example, a patient may have idiosyncratic side effects from the best available drug for their condition, in which case it is useful to have a less effective drug to try next.
Doctors and patients need all the information about all the clinical trials that have been conducted on a drug in order to make informed decisions. A clinical decision (“shall we use this drug on this patient?”) is very different to a regulator’s decision (“overall, is it in the interests of society that this drug should be on the market for use at all?”).
Furthermore, regulators can sometimes miss important problems with medicines. For example, the harms from the drugs Vioxx and Avandia, both now taken off the market, were spotted by academics and doctors rather than regulators. This is not because regulators are incompetent: these are difficult problems, so it is good to have many eyes working on them. It is also good if the evidence behind regulators decisions can be independently assessed, to ensure regulators have made good decisions.
Is this about clinical trials data on individual patients?
No. That is a separate issue but also a very important one. If we had better systems for sharing clinical trial data about individual patients, this would create more transparency, and permit more accurate assessments of drugs by doctors and patients. This is an important area, which presents some confidentiality problems, though they can be overcome. It is, however, a different issue to the bigger problem of information about the results of trials and their conduct being withheld.
Withholding information about clinical trials from doctors and patients is against the public interest, harms patients, and wastes money.