Clinical trials—where doctors test new medicines and treatments on humans— keeps medicine state-of-the-art. But too many trials don't publish their results, says Dr. Ben Goldacre, so he co-founded All Trials, a campaign to get all clinical trial results on the record. Bob talks with Goldacre about this venture.
******************** THIS IS A RUSH, UNCORRECTED TRANSCRIPT ********************
BOB: Over the years, we’ve returned repeatedly to what gets into medical journals and how clinical research is represented -- or misrepresented -- there.
But at least as important, is what does not get into medical journals: the results of thousands of clinical trials that never see the light of day, mainly because pharma doesn’t like to advertise its failures.
If science falls in the forest and nobody is there to hear it, does it permit other scientists and physicians to make informed judgments? Rhetorical question. Yet, at least one study suggests that half of American clinical trials are never made public. This is the backdrop renewed plea from the World Health Organization this month, urging doctors worldwide to fulfill their ethical obligations to publish all their findings. Ben Goldacre is a physician and journalist, and a co-founder of All Trials, a group dedicated to that very goal.
We asked him for a recent example of a transparency fail.
GOLDACRE: Well I suppose the poster child for trials transparency recently has been Tamiflu. So this is a drug made by Rosch, the UK government spent about half a billion pounds stockpiling it in one year alone. The American government has spent hundreds of millions of dollars on it. Many of the clinical trials which were conducted weren't made available to doctors, researchers, and patients. We now know that we were mislead about its effectiveness. Now it's important to be absolutely clear here that Rosch broke no laws at all, and that is the problem. What we have here is a strange cultural blind spot in medicine. We don't see withholding whole studies as research misconduct in the same way that we would view, for example, deleting individual data points from within one individual study.
BOB: A huge part of the information ecosystem is the medical journal which medicine and public health experts have relied on for, I guess close to a couple of centuries. Why have these publications ceased to be sufficient arbiters of success and failures and newsworthy and marginal.
GOLDACRE: It turns out that actually- and this may sound strange - academic journals are pretty bad places to report the results of scientific research. They miss out a lot of important information on adverse events, the nuts and bolts about how the trial was conducted - if it really was a fair test - or whether it was flawed by design. So, when we say we need to results of clinical trials to be made available, i think we probably don't really mean in an academic journal article anymore. I think there are much better places to report those trials.
BOB: And turns out we have the technology for that - the internet.
GOLDACRE: [Laughs] Yeah, that's exactly correct. So, clinicaltrials.gov was initially set up as a register of trials in progress, and the idea is that you force people to register the trials they're doing, and then you've got something to go and copmare what's missing - what people have done but not reported. Now, clinical trials dot gov has developed a feature where people can report the results of their clinical trials straight onto that database with absolutely no delay. And crucially, it's very difficult to misreport your trial on clinical trials dot gov, because there's not really a lot of room for narrativizing or waffling, or reanalyzing your data in a more flattering way. there's just a box where you have to put the result, and you either put it there or you don't. We now know from published research taht the completeness of reporting on clinicaltrials.gov is actually much better than for when trials are reported in academic jouanls. Now that's really important because it knocks the legs out of one of the I think spurious objections that researchers and drug companies make against campaigners like me, which is they say, well it's not our fault because when we have an unflattering results or a negative result, academic journals aren't interested in publishing it. Well I'm not interested in you publishing it in an academic journal anymore. i think it needs to be reported first and foremost on clinical trials dot gov.
BOB: As we observed earlier, you are not new to this agitation game. We have come a long way, and clinicaltrials.gov is exhibit A. Where are we now, who is able to avoid this basic mandate of the Food and Derug Administration.
GOLDACRE: well, pretty much everybody right now - the FDA has a law, called the FDA amendment act of 2007 - and the FDA is allowed to impose a ten thousand dollar a day fine on people who fail to report the clinical trials that they've conducted. No such fine has ever been imposed. And actually to be honest, although 3.65 million dollars a year would obviously be quite a lot for you or I, that is a parking ticket to a major drug company or a major institution even.
BOB: Here I guess we have to explain that the publishing of clinical data is a 2 part process; first you register your trial, and then publish the result. And there's a huge gap between those trials that are registered and those that are followed up by the publication of the full result.
That's absolutely right. So when in the 1980s we realized that the results of trials were being routinely withheld. The first thing to really campaign for is the registry: this is the clinical trial i'm going to run, this is how many people, these are the drugs i'm testing these are the outcomes i'm measuring. The idea was that then people would never again withhold the results their clinical trial because it would be too embarrassing. But it's nobody's job to go through and say "hey, this one's published its results but this one is still hiding its results." Now, you could say that that should be the job of regulators. But since they've failed to do that, the case that I made this week in the academic journal Plos medicine, was that doctors and payers, health services around the world, should take this matter into their own hands. We need to set up routine ongoing audits, we can then very quickly and easily start to say the following drug companies, the following funding bodies, the following institutions, are the best and the worst at withholding the results of clinical trials. We can also say the following drugs are the best and the worst for missing data. And that's really important because if there were two apparently effective treatments, the clinically cautious thing for a doctor like me to do is to prescribe the drug where you can trust the evidence more. But also finally we can say which researchers are the worst for failing to share the results of clinical trials. When a researcher goes back to a committee and says, I'd like to conduct a clinical trial, please give me access to 700 patients in this hospital district, the administrators reviewing that application should be able to say, no, you're not allowed to touch any more patients until you publish the results of the clinical trials that you've already finished.
BOB: You have lately gotten a pretty hefty ally in all of this: the World Health Organization weighed in on the subject of incomplete data --
GOLDACRE: Yeah! What's been really great about the WHO announcement is that they've said we also need the results of clinical trials from the past. The trials on the drugs that we're using right now today. When you talk to industry today they say, well maybe we can be more transparent for now, but you've got to understand that it's very difficult for us to go back and get trials from 5, 10, or 15 years ago, even after the time at which they promised they were fixing this problem, they were continuing to lose, shred, mislay, often very important information that would really inform patient care. They've been lucky for 30 years that because this issue is a little bit technical, a little bit boring, it's protected from popular scrutiny by a kind of force field of its own tediousness. That's what I think we're going to have to fight through with things like the All Trials campaign.
BOB: Ben, thank you so much.
GOLDACRE: Thanks, chief.
BOB: Ben Goldacre is a physician and journalists, and co-founder of All Trials.net, a group dedicated to transparency in clinical research.