Bad Pharma

The practices of large pharmaceutical companies and scientific journals could “cause hundreds of thousands of unnecessary deaths every year”. This was the stark message Dr […]

Dr. Goldacre at the Oxford Union

The practices of large pharmaceutical companies and scientific journals could “cause hundreds of thousands of unnecessary deaths every year”. This was the stark message Dr Ben Goldacre delivered when he spoke at the Oxford Union in Michaelmas term. Also the main focus of his book Bad Pharma, he outlines the way in which big corporations are allowed to freely exercise dangerously bad scientific practice. Hiding undesirable results and rushing studies on the effectiveness of potential treatments are shockingly abundant in the medical industry. More significantly, this has far reaching consequences for both doctors and patients alike.

The pressure on pharmaceutical companies to create a steady flow of sell-able products means that drugs can often be rushed through the different stages of clinical trials without proper analysis of the data collected. This increases the chances of fluke results and false data making it through to publication. The competition between businesses also provides a disincentive for drug companies to test against any similar treatments produced by their competitors, since they don’t want to prove that their treatment is any less effective than an alternative. As a result, key information about the side effects and the relative effectiveness of new medication is either never collected or it is hidden from public view. Goldacre argues that this makes it impossible for doctors to decide on the most appropriate treatments for their patients.

the results of around half of all trials conducted are never published

Goldacre also claims that the willingness of scientific journals to publish unexpected and interesting results over results which merely replicate expected data creates a bias towards fluke studies. If the chance of the expected outcome occurring is 0.99 and 100 studies are performed, scientific journals are far more likely to publish the 1 fluke study which goes against expectation than any of the other 99. This, like any other scientific hypothesis, is testable, and in March 2012 a study on the bias publication of studies was published in Nature. The test looked at 53 ‘new and revolutionary’ findings in the field of cancer research, all of which had been published in peer reviewed scientific journals. Shockingly, the study discovered that only 6 of these findings were readily replicable in the lab, and it concluded that the most likely cause of this was over publication of anomalous results and outlying data. In highly respected scientific journals, this publication bias could mislead doctors and send them on a wild goose chase for ineffective medicine.

Lorcainide is a drug used to treat tachyarrhythmia

More worryingly, bias publication of unexpected results only seems to happen for studies which produce positive conclusions, results which show a potential leap forward in science. Studies that show unexpected, negative results can often go completely unpublished. A prime example of this is lorcainide. Lorcainide is a drug used to suppress arrhythmic heart patterns, and in 1980, a study in the US on the effects of lorcainide on recovering heart attack patients was conducted. The logic was that since post-heart attack patients often have abnormal heart rhythms, lorcainide would help suppress this and would improve the chance of survival. Of the 100 people in the test group, 50 were given lorcainide and 50 were given a placebo. Of the 50 that were given lorcainide, 10 died. Of the 50 that were given a placebo, there was only 1 death. As a result of this test, funding for lorcainide was rapidly cut, it was regarded as a failure and the study was never published. The bias non-publication of these results combined with the rushed analysis of drugs under development within pharmaceutical firms meant that over the next 5 to 10 years, drug companies all began rolling out their own anti-arrhythmic treatments to prescribe to heart-attack patients, and because the logic behind them made sense, they weren’t thoroughly tested and were readily used within hospitals. It in fact took so long to recognise this error that from their introduction to the realisation of the dangers posed by these drugs, over 100,000 people died unnecessarily.

Industry-Testing is nowhere near as reliable as it needs to be

While Goldacre has made it clear that the situation has improved since 1980, there is still a problem. In his Michaelmas term talk, Goldacre listed examples of when publication bias has had damning consequences for patients across the world. Publication bias is still a huge problem within medicine, and studies on it estimate that not only are the results of around half of all trials conducted never published, but positive trials are twice as likely to be published as negative trials. Goldacre likens this to tossing a coin 100 times, withholding half of the results and claiming the coin always lands heads up. Letting someone get away with this approach in a single study would be an example of gross research misconduct, but it happens all the time within large scale medical research, and Dr. Goldacre claims it is “a cancer at the core of evidence-based medicine” that we continue to “blindly tolerate”.

Why does this problem exist? Drugs companies are legally allowed to pick and choose which studies they publish. If a study produces results they don’t like, they try to keep it hidden. If the results don’t show anything new, they are ignored by either the researchers or the scientific journals, and if the results are positive, researchers actively try harder to get the results published.

This is not a difficult problem to overcome, but we need to be active about it. Drug companies have too much to lose by admitting the drugs they’ve spent millions of dollars on are not the best available, are ineffective, or are even potentially dangerous, so we need to force them to be truthful. We need to force them to publish all the studies they have ever conducted on humans, past and present, and we need to force them to publish every study they conduct in the future. Only then can doctors have the transparency of evidence they need to provide proper care to their patients.

“A cancer at the core of evidence-based medicine”

This is the aim of the AllTrials campaign, an on-line petition to call on governments, regulators and research bodies the world over to change the law on publication of scientific research. It has already attracted big names such as GSK, the Royal Society of Medicine, the British Pharmacological Society and Cancer Research UK, but it still hasn’t been implemented, and for every day these trials remain unpublished, people may die unnecessarily. This is not something we should apathetically tolerate. It is not right that drug companies value their profit enough to reduce the standard of care for the lives they should be saving, but this is what is happening right now in large scale medicine. I urge everyone to sign the petition and show your support, so that countless lives may be saved.

About Andrew Smith