As a doctor, Ben Goldacre likes to have all the available facts about a prescription drug before he even thinks about prescribing it. However, when it comes to medicines, it’s nearly impossible to find all the existing data. As Goldacre described in this impassioned talk given at TEDMed 2012, there is a bias in medicine towards publishing studies that produce positive results. Meanwhile, studies with negative results often go unpublished, the information culled in the research all but disappearing.
“People will do lots and lots of studies and on the occasions that it works, they’ll publish. On the ones it doesn’t, they won’t,” says Goldacre in his talk. “This is a problem because it sends us all down blind alleys.”
You are probably asking yourself: how big is the problem?
The answer is, unfortunately, very big. As Goldacre notes, in this study published in Nature in March 2012, researchers tried to replicate the results of 53 basic preclinical cancer studies. Of those 53 studies, only six were replicable—one of the benchmarks of good science. And the situation is worse when it comes to trials of specific pharmaceuticals. In his new book, Bad Pharma, Goldacre sounds a warning bell on the fact that drug manufacturers are the ones who fund trials of their own products.
“Drugs are tested by the people who manufacture them, in poorly designed trials, on hopelessly small numbers of weird, unrepresentative patients, and analysed using techniques that are flawed by design, in such a way that they exaggerate the benefits of treatments,” writes Goldacre in his book. “When trials throw up results that companies don’t like, they are perfectly entitled to hide them from doctors and patients, so we only ever see a distorted picture of any drug’s true effects.”
In his talk, Goldacre turns his attention to a survey by Erick Turner of all the anti-depressant trials filed with the United States Food and Drug Administration. Of the batch, there were 38 studies that produced positive results and 36 that produced negative results. Of the positive-result group, 37 of the studies were published. Of the negative results group, only three were published.
In his book, Goldacre also notes a survey of trials of statins, a class of drugs used to lower cholesterol. In 2007, researchers looked at all the published trials on these drugs, finding 192 of them. They found that industry-funded trials were 20 times more likely to give results favoring the drug.
“This is a systematic flaw in the core of medicine,” says Goldacre.
To hear more about why this problem exists, watch Goldacre’s talk. Below, five specific drugs for which unpublished data gives a very different picture of both efficacy and side effects.
- Reboxetine. After reading positive things about this antidepressant in journals, Goldacre decided to prescribe it for a patient who wasn’t responding to other drugs. But upon further examination, Goldacre felt duped. While seven trials had been conducted comparing the results of reboxetine against a placebo sugar pill, only one was published — the one that suggested it worked better. However, the six other trials were done on almost 10 times as many patients. Meanwhile, three small studies — with 507 patients in total — showed that reboxetine was just as good as other antidepressants. However, data on 1,657 patients went unpublished. “As a doctor, I did something that, on the balance of all the evidence, harmed my patient, simply because unflattering data was left unpublished,” writes Goldacre in his book. “Reboxetine is still on the market and the system that allowed all this to happen is still in play, for all drugs, in all countries in the world.”
- Lorcainide. As Goldacre explains in his talk, this drug — which suppresses abnormal heart rhythms — was developed for use in those who had recently had heart attacks to increase chances of survival. The drug was tested in 1980 in a small study of fewer than 100 patients — half who took lorcainide and half who took a placebo pill. Of those who took lorcainide, nine died. Of those who received the placebo, only one died. The commercial development of the drug was quickly halted, and the results of the trial were not published. But the story didn’t end there. Over the next decade, other pharmaceutical companies had the same idea for a prescription drug. Several were brought to market and prescribed widely. Sadly, more than 100,000 people using these drugs died unnecessarily before the connection was recognized. In 1993, the researchers who did the early study published an apology to the scientific community.
- Rosiglitazone. This antidiabetic drug was released by GlaxoSmithKline (GSK) in 1999. However, as Goldacre writes in his book, “In that first year, Dr. John Buse from the University of North Carolina discussed an increased risk of heart problems at a pair of academic meetings. The drug’s manufacturer, GSK, made direct contact in an attempt to silence him, then moved on to his head of department.” While Buse stayed quiet, four years later, the World Health Organization contacted GSK about a potential relationship between rosiglitazone and heart problems. They spent the next two years analyzing the data. Writes Goldacre, The analysis “showed that the risk was real, but although both GSK and the FDA had these results, neither made any public statement about them.” It wasn’t until 2007, when an unrelated cardiologist published a study showing a 43% increase in the risk of heart problems in patients taking rosiglitazone that the medical community took notice. The drug was restricted in 2010.
- Oseltamvir, aka Tamiflu. While hospitals and pharmacies have been stockpiling Tamiflu to prevent complications from the flu, new research suggests that the drug might not be as good as we think. While two studies on the drug were published, researchers from the University of Georgia have uncovered an additional eight not published by the manufacturer, Roche. This new data suggests that the drug might not be effective for 30 hours, as currently believed. But the most interesting part of the article is what the researchers had to go through to get this additional data. They issued Freedom of Information Act requests and dug through documents that were largely redacted.
- Paroxetine. Sometimes, drugs are created for adults but prescribed for children. That was the case for this antidepressant, which GlaxoSmithKline wanted to market specifically for minors. Writes Goldacre in his book, “Between 1994 and 2002, GSK conducted nine trials of paroxetine in children. The first two failed to show any benefit, but the company made no attempt to inform anyone of this by changing the ‘drug label’ that is sent to all doctors and patients. In fact, after these trials were completed, an internal company management document stated: ‘It would be commercially unacceptable to include a statement that efficacy had not been demonstrated, as this would undermine the profile of paroxetine.’ In the year after this secret internal memo, 32,000 prescriptions were issued to children for paroxetine in the UK alone.” But that’s not nearly all. In their trials, GSK noticed an increased risk of suicide in children taking paroxetine, but because the use of the drug for children was “off label,” they were not required to disclose this information. They didn’t until 2003. Only then were doctors warned not to prescribe the drug for those under 18.
These examples are disconcerting. But Goldacre assures that this is solvable problem. He explains, “In any sensible world, when researchers are conducting trials on a new tablet for a drug company, for example, we’d expect universal contracts, making it clear that all researchers are obliged to publish their results, and that industry sponsors – which have a huge interest in positive results – must have no control over the data.”
For much more on this topic, read an excerpt of Bad Pharma, published in The Guardian.