The case of the missing evidence

Elementary, my dear Watson. It didn’t need the return of Sherlock Holmes to work out that you can’t take a fully-informed decision without knowing all the evidence.

I wrote a year ago about the scandalous situation that exists in the field of medicines research, where pharmaceutical companies can quietly hide away inconvenient data from clinical trials that don’t have the outcomes they had hoped. The All Trials campaign, backed by BMJ, brought the issue into the open. We’ve heard a lot from the industry since about its good intentions to behave better in future, to register all clinical trials as they are started, and to disclose all information to the regulators. The industry seems to hope this will make the problem go away.

So doff the deerstalker to the sleuths of the UK’s Public Accounts Committee, which today published its own report on the subject. Their outrage is patent, if phrased in Parliamentary language. ‘We were surprised and concerned to discover that information is routinely withheld from doctors and researchers,’ they say. And, they note, despite the promises of the pharmaceutical companies, ‘None of the latest proposals from regulators or industry adequately addresses the issue of access to the results of trials from previous years on the medicines in use today.’

They have a rather important example. Back in 2009, the world faced an outbreak of pandemic flu. The outbreak was real and cost lives. Fortunately, its effects in the UK were less lethal than originally anticipated. This means many now dismiss it as hype. But as the PAC observe, pandemic flu remains a very real threat. It’s higher on the list of civil emergencies than coastal flooding  or terrorist attack. Being well-prepared is vital.

If we had an outbreak of lethal flu, we’d want access to the drugs most likely to save lives, and very quickly. The drugs would need to slow the transmission from person to person, and prevent life-threatening complications. So the UK has spent millions of pounds on stockpiling the drugs that seemed to have the best evidence in support of them. Tamiflu, an anti-viral drug in tablet form, is one of them.

The UK’s drug regulator told the investigation that they were confident they’d seen all the data relating to the effectiveness of Tamiflu. Yet when the internationally-renowned research experts Cochrane Collaboration looked at the data, they came to a different conclusion. ‘We found a high risk of publication and reporting biases,’ they wrote in their review of the effectiveness of the drug. ‘We are unable to draw conclusions about its effects on transmission or complications… Full clinical study reports are unavailable to us.’ In other words, they couldn’t say whether Tamiflu would do what we will most need it to do in the event of a pandemic, because not all the evidence was made available.

The drug’s manufacturer has now agreed to supply the full study reports, and Cochrane will review the evidence and its conclusion. The PAC noted that it was ‘extremely concerning’ that this situation had taken five years to resolve.

The committee’s report makes some very simple recommendations. All study results should be published, including previous trials of drugs in use today. All trials should be registered, retrospectively if necessary to capture trials started before registration came into force. The drug regulators should demand all the information to which they are entitled (and share it with each other), and the evidence regarding the UK’s stockpile of Tamiflu should be reviewed, once Cochrane has finished its work.

Elementary, perhaps. Let’s hope Mycroft is listening.

%d bloggers like this: