by Ben Goldacre
Let’s take a look at this mysterious world. When adverts were first made legal again in the US, they could only appear in print, because of a requirement to include all the side-effect information from the drug label. Since 1997 the rules have been relaxed, and now the side effects can be abbreviated (they are read out at jabbering speed over the end of the TV adverts). After this change the pharmaceutical industry’s annual advertising budget rose from $200 million to $3 billion in the space of just a few years. Notable single spends include Vioxx, at $161 million, which was taken off the market because of serious concerns over hidden data; and Celebrex, at $78 million, also later taken off the market because it was harming patients.
Various approaches have been used to try to evaluate the impact of these adverts in the real world.4 One study observed patients visiting their doctors in Canada, where direct-to-consumer drug advertising is still banned, and the US. It found that those in the US were more likely to believe they needed medication, more likely to request specific drugs that were advertised on television, and more likely to receive a prescription for that drug. In other words: the adverts worked. The doctors in the US, meanwhile, were more likely to report being worried about whether the drugs requested by their patients were appropriate.
Another study took a more proactive, experimental approach. Trained actors, posing as depressed patients, were sent to visit doctors in three American cities (three hundred visits in all).5 They all gave the same background story, about the problems they were having with low mood, and then were randomly assigned to act in one of three ways at the end of the consultation: to ask for a specific named drug; to ask for ‘medicine that might help’; or to make no specific request. Those who did what adverts drive patients to do – ask for a specific drug, or ‘medicine’ – were twice as likely to receive a prescription for an antidepressant pill. Whether you think that’s a good thing might depend partly on whether you think these drugs are actually worth using (the evidence overall shows that they’re pretty ineffective for mild and moderate depression). But whatever your view on antidepressants, the evidence very clearly shows that what patients say to their doctor, and what they request, has a significant impact on what is prescribed. Engaged and informed patients are something most doctors would want to see, but here the question is whether the information the patients have received is truly helpful, and whether doctors can resist inappropriate requests for pills.
So the same study sent more actor-patients to see doctors, but this time they gave a clear history of ‘adjustment disorder’, a term used by some people to describe the simple human phenomenon of feeling bad in the immediate aftermath of a very bad thing happening in your life. This is a normal and appropriate thing for your feelings to do, though it’s unpleasant, as any normal feeling person will know, and pills to treat it aren’t a good idea. But patients presenting with ‘adjustment disorder’ who demanded a specific named drug still got it, in 50 per cent of cases, compared with 10 per cent of those who didn’t request any medicine. This is the dark side of these adverts, and as a doctor, I’ve always been surprised by people who say that it’s doctors who force pills on patients. Doctors are generally nice people, and eager to please. They will get bounced into giving people what they want, and a lot of patients have been persuaded, through whatever social processes are at play in their world, that pills fix things. I’ll rephrase that for something that’s coming later in this chapter: a lot of people have been convinced that they’re patients.
So the evidence shows that adverts change behaviour, and they change it for the worse. This becomes much more worrying when you look at which drugs get advertised. One study gathered data on 169 drugs on the market, and looked for patterns in their promotion.6 Firstly, drugs are advertised more when the number of potential patients, rather than current patients, is large. This is an interesting finding, because it means that people are turned into patients, which is good news if they’re sick, but bad news if they’re not. Secondly, drugs get advertised more when they are new. That may seem inevitable, but it can be very problematic. As we’ve already seen, new drugs are often not a good idea: they’re the drugs we know least about, because they haven’t been around long; they’ve often been shown only to be better than nothing, rather than the best treatments we already have; and lastly, even if they’re equally effective when compared with older drugs, they will be more expensive.
We’ve already seen how AstraZeneca managed the transition from omeprazole to esomeprazole, the ‘me again’ drug, and we can also see it in their advertising strategy. The company spent $100 million on omeprazole in 2000, the second-highest advert spend that year. Then in 2001, as it was shortly to come off patent, AstraZeneca dumped the drug, and put $500 million into advertising esomeprazole instead, the ‘me again’ drug. But as we’ve already seen, these two drugs are almost identical, and esomeprazole is basically no better than omeprazole, just much more expensive.7 The advertising campaign was highly effective, so we waste money on drugs that are no better than those that already exist.
As I’ve previously argued, when you take a step back from pharmaceutical industry marketing, it is simply a process whereby patients pay money to drug companies, in order for them to produce biased information, which then distorts treatment decisions, making them less effective. This is not simply my view, or an interpretation from the basic principles of economics: we can also watch the phenomenon in real time, by tracking the cost of drugs, and their advertising budget. One study looked at clopidogrel, an ‘anti-platelet’ drug that can help prevent blood clotting, and is given to people at high risk of various heart problems.8 It’s popular, and expensive – in 2005 it was the world’s second-biggest-selling drug, at $6 billion. Clopidogrel came to market in 1999 with no advertising, and was used widely, with no advertising, until 2001. Then the drug company introduced television advertising, spending $350 million in total. Oddly, this had no impact on the number of people taking the drug, which continued to increase at exactly the same rate. So nothing changed, except for one thing: the price of clopidrogel went up by 40c per tablet. As a result, Medicaid alone paid out an extra $207 million. For me this episode is strong evidence – if it wasn’t already obvious – that it’s patients and the public who pay for the industry’s expensive marketing campaigns.
That would be fine if we were paying for reliable information, cautiously explained, but the reality is that even if the adverts are adequately policed (I will review the endless to-andfro on this later) they are still focusing only on pills and commercial products, which in turn distorts our whole outlook on medical interventions. Any sensible public-health campaign to inform people about reducing the risk or the consequences of a disease would look at prescription drugs, of course. But it might also inform patients with equal vigour about things like exercise, alcohol, smoking, diet, recreational drug use, social engagement, and perhaps even social inequality. A public education and engagement programme costing $350 million – the amount we have just seen spent on clopidrogel alone – would achieve a great deal for all of those goals, but instead, using money from patients and public, it is spent on TV adverts for one pill.
This is a theme that we will see recurring: one of priorities being distorted, alongside individual treatments being sold. But first, we must note that adverts are not the only way in which drugs are advertised.
Celebrity endorsement
In the 1952 Hollywood movie Singin’ in the Rain, Debbie Reynolds plays Kathy Selden, a talented singer who hides behind a curtain, and covertly provides the sweet singing voice for an on-stage starlet who merely mimes the words. In a recent interview, Debbie Reynolds suddenly starts explaining that ‘Overactive bladder affects you because it defects you…effective treatment is available.’9 The interview didn’t mention that she was working for Pharmacia, a company pushing a new treatment for overactive bladder. In another recent interview, Lauren Bacall encourages readers to get tested for macular degeneration, which she says can be h
elped with Visudyne. Neither she nor the interviewer mentions that she is being paid by Novartis to promote the drug.10 The mom from Serial Mom (seriously) drops references in interviews to arthritis drugs, and is paid to do so by Wyeth and Amgen.11
This is a new phenomenon, but it has become so widespread that American celebrity shows are now having to screen for endorsements before interview segments go ahead. CBS, for example, recently reported that an interview with Rob Lowe – of all people – was dropped by NBC because of concerns about him promoting a drug for people on chemotherapy.120 So when PR News gushingly reports that a character on ER with Alzheimer’s disease was treated with the new drug Aricept thanks to the work of a PR firm working for Pfizer, it comes as no surprise.13
The TV spots where sponsorship is openly declared are often much more bizarre: although it’s only directed at the US market, for a flavour of how this world works, I strongly encourage UK readers to find Barry Manilow’s ‘Get Back in Rhythm’ video online, promoting a drug that treats abnormal heart rhythms (‘Hi, this is Barry Manilow, and that’s the rhythm to my song “Copacabana”…’). Jon Bon Jovi’s painkiller advert is even slicker. And Antonio Banderas is the voice of a bee for Merck’s Nasonex.
But sometimes the hidden hand can be more subtle. You will remember, I’m sure, the media stories around Herceptin, a drug which has a very modest effect on survival for some kinds of breast cancer, at a cost of serious cardiac side effects, and tens of thousands of pounds for each treatment. From 2005, access to this drug became a spontaneous cause célèbre for the British press, and the extent of the distortion is best expressed in the words of one GP who was suffering from breast cancer herself, and later wrote of being caught up in the bombardment of media coverage: ‘I began to feel that if I did not receive this drug then I would have very little chance of surviving my cancer.’ When she stepped back from the maelstrom and looked at the data as it stood at the time, she was surprised: ‘More careful analysis of the “50 per cent benefit” which had been widely quoted in the medical and non-medical press, and fixed in my mind, actually translated into a 4–5 per cent benefit to me, which equally balanced the cardiac risk…This story illustrates how even a medically trained and usually rational woman becomes vulnerable when diagnosed as having a potentially life-threatening illness.’14
The dominant theme of this media coverage was that NICE should approve Herceptin for use on the NHS. Yet, bizarrely, the campaign was orchestrated before any evidence had even been given to NICE. The health minister, similarly, said the drug should be approved, but again before the data on its effectiveness was available.
What can explain all this? One group of academics tracked down every single newspaper story on Herceptin, to try to understand what had happened.15 They found 361 articles in total: the overwhelming majority (four out of five) were positive about the drug’s efficacy, and the remainder were neutral, with none negative. Side effects were mentioned in less than one in ten articles, and were often played down as minimal. Some articles came right out, in the face of all common sense, and declared that this miracle cancer drug had no side effects at all.
Half of the stories were about the problems in getting a licence for Herceptin’s use in early-stage breast cancer, but they almost never mentioned that the manufacturer, Roche, needed to apply for that licence itself, and hadn’t even done so yet. Many of them attacked NICE, but they hardly ever mentioned that it couldn’t even look at the drug for this use until it was licensed, and until the government asked it to.
Perhaps most remarkable was the use of individual patients, in two thirds of all the articles. Although journalists chose not to mention how they had found these women, in reality they were being provided to the media by lawyers and PR firms. Elaine Barber and Anne Marie Rogers each appeared in dozens of articles: they were handled by Irwin & Mitchell, which was shortlisted for a Chartered Institute of Public Relations award for its work on this project. Lisa Jardine, Professor of Renaissance Studies at Queen Mary, University of London, who was suffering from breast cancer, told the Guardian she was contacted by a PR firm working directly for Roche.16 The charity CancerBackup also appeared in these stories repeatedly, often hawking a survey finding which had been delivered to it by Roche, which also funded the charity’s work.17
Why would the involvement of a PR firm working for the drug company not be declared up front? Here’s an unusually open explanation. In 2010 the British government proposed a scheme allowing pharmacists to substitute all prescriptions for branded medicines with the generic alternative. Generic drugs, as you now know, are identical copies of a molecule, but manufactured more cheaply by another company when the original inventor’s exclusive patent has run out. Doctors who have been influenced by advertising from drug companies often write the brand name of the drug on the prescription, rather than the true scientific name. This proposed new law would allow pharmacists to ignore the brand name and substitute an identical generic copy of the drug, made by whichever company sold it most cheaply, potentially saving the NHS huge amounts of money, at no risk to patients. A letter of protest immediately appeared in The Times, signed by various patient groups and experts, and received positive coverage elsewhere in the broad-sheets. ‘Plan to Switch to Cheaper Medicines will Harm Patients, Say Experts’, reported The Times. It even had a case study: ‘Patient Given Seroxat Substitute Felt Unwell Within Two Days’. But Margaret McCartney, a GP who writes for the British Medical Journal, found out that the letter was coordinated and written by PR company Burson-Marsteller, paid for by the drug company Norgine. Peter Martin, chief operating officer of Norgine, was the major influence behind the campaign, but he did not sign the letter himself. Asked why not, he cheerfully replied, ‘There was no conspiracy. The frank truth, the honest truth, is that I thought that having a pharmaceutical company in there would sully the message somewhat.’
Despite that story, we shouldn’t allow ourselves to become too obsessed with conspiracies. Cancer and health are – as I’ve written endlessly outside this book – areas where many newspaper journalists distort facts as a matter of casual routine, without any commercial assistance, partly through a lack of knowledge, but also, I suspect, through a desire to be crusaders. We have already seen how figures can be reported misleadingly, in the coverage of the Jupiter trial on statins. But the simple tradition of using human stories, even when they do not represent the reality of the evidence, provides an open goal to companies, hoping that their drug will get positive coverage. In 2010, for example, the British media was filled with journalists angry about NICE’s recommendation not to fund Avastin, a bowel-cancer drug that costs £21,000 per patient. Overall, on average, when added to all the other treatments, the drug had been shown to increase survival by just six weeks, from 19.9 months to 21.3 months. But the newspaper stories featured Barbara Moss, who paid out of her own pocket to have Avastin in 2006, and was still alive four years later.18 I am pleased for her, but someone who has survived four years is no illustration whatsoever of what happens if you take Avastin for bowel cancer. You could find patients who survived for four years without Avastin too, and neither they nor Barbara Moss tell you anything informative whatsoever about the drug’s effectiveness.
Individual stories such as these are the bread and butter of health journalism. But beyond the desire to report on individual ‘miracle cures’, there is also a murkier problem. Journalists, like all of us, like to explain the world around them. Sometimes, though, a deceptively simple explanation for a complex phenomenon can be very powerful: it can prime the reader to accept a specific treatment, but it can also change our whole cultural understanding of a disease.
More than molecules
The idea that depression is caused by low serotonin levels in the brain is now deeply embedded in popular folklore, and people with no neuroscience background at all will routinely incorporate phrases about it into everyday discussion of their mood, just to keep their serotonin levels up. Many people also ‘know’ th
at this is how antidepressant drugs work: depression is caused by low serotonin, so you need drugs which raise the serotonin levels in your brain, like SSRI antidepressants, which are ‘selective serotonin reuptake inhibitors’. But this theory is wrong. The ‘serotonin hypothesis’ for depression, as it is known, was always shaky, and the evidence now is hugely contradictory.19 I’m not giving that lecture here, but as one brief illustration, there’s a drug called tianeptine – it is a selective serotonin reuptake enhancer, not an inhibitor, that should reduce serotonin levels – and yet research shows that it is also a pretty effective treatment for depression.
But in popular culture the depression–serotonin theory is proven and absolute, because it has been marketed so effectively. In drug adverts and educational material you can see it recycled, simply and plainly, because it makes absolute sense: depression is caused by too little serotonin, therefore our pill, which raises serotonin levels, will fix it. This uncomplicated notion is attractive, even though it has little support in academia, perhaps because it speaks to us of controllable, external, molecular pressures. As one US newspaper said recently about depression: ‘It’s not a personal deficit, but something that needs to be looked at as a chemical imbalance.’20
This is not a belief that arose spontaneously out of nowhere: it has been carefully fostered and maintained.21 A recent advert for paroxetine by GSK says: ‘If you’ve experienced some of these symptoms of depression nearly every day, for at least two weeks, a chemical imbalance could be to blame.’22 Or a patient’s guide to Pfizer’s SSRI: ‘Zoloft may help correct the chemical imbalance of serotonin in the brain.’ The same claims are found in adverts around the world, directed not only at adult patients, but also at children. One museum exhibition on the human brain sponsored by Pfizer started at the Smithsonian Institution in Washington and then toured the US. Half of us will experience ‘brain dysfunction’ at some stage in our lives, it happily explains: ‘Chemical imbalances in the brain – often involving the neurotransmitter serotonin – are almost certainly involved.’23