Abstract
Background: Conflicts of interest
Conflict of interest statements are now de rigueur in lectures and articles. The presumption is that links to pharmaceutical companies might conflict with doctors’ duties to patients. But unless these duties to patients are specified, conflict of interest statements risk tokenism.
Ironically we want doctors to be biased – by treatments that truly help and by good evidence. The only people with no conflicts of interest when it comes to clinical care are those with nothing to offer. But these biases must be open to scrutiny for which access to data is critical. I am conflicted by my involvement in an adverse event reporting website: Rxisk.org. But unlike pharmaceutical company presentations at meetings, or articles in journals, I am hiding no data from you – whereas they invariably are.
Common interests?
Symposia on links between clinicians and industry regularly have titles that include win–win (Brown, 2012; Powrie-Smith, 2012). Win–win is also the basis for adverts to recruit clinicians to industry today which speak to a vision of developing needed treatments and producing the best evidence; adverts that many would find hard to distinguish from the medical mission. As a result perhaps, an increasing number of clinicians have spells in and links to industry. Finally, governments clearly believe in win–win and actively encourage clinicians to partner industry. In the UK, the government has created networks of clinicians to facilitate trials of new drugs (see www.inpharm.com/news/172632/abpi-launches-new-nhs-partnership-guide).
Despite token declarations of conflicts of interest, in practice medical journals do not permit a conflict. Thus, the BMJ ran an issue on academic fraud on 12 November 2011. There are few medical researchers who haven’t fudged a figure or put on unwarranted spin on data, and so this issue can make us all feel uncomfortable. The BMJ names fraudulent doctors who fall short of the standards of good clinical and laboratory practices to which the pharmaceutical industry operates.
The lead article on the topic of scientific misconduct was by Elizabeth Wager, Chair of the Committee on Publication Ethics (COPE) that overviews the practices of journals (Wager, 2011). She is a medical writer. Neither her article nor any other pieces refer to industry malpractice, despite a series of fraud cases and billions of dollars in fines, and even though the same BMJ issue had a small News item about GlaxoSmithKline’s payment of a $3 billion fine to the US Government for a series of dubious practices (Hawkes, 2011).
Some years previously, I had written an article for the BMJ showing a stark contrast between how suicidal acts were reported in the trials that brought fluoxetine, paroxetine and sertraline to the market and what the real data actually looked like. Everything in the article was in the public domain. Having answered all points raised by the reviewers, while reviewing the galleys for the article, I got an email from the editor saying that ‘we’ve run into a legal wall with our libel lawyer reluctant for us to publish your piece’ (Healy, 2012a).
I have had approximately 10 articles in different journals rejected on this basis. The most interesting was from Index on Censorship, who trumpet themselves as bravely taking on the Government. Given an article on antidepressants and the risks to children, after all legal questions had been answered and documents provided for every single point, they decided to self-censor (Healy, 2012a).
In 1999, John Cornwell wrote Hitler’s Pope, a book on the role that Pius XII failed to play in drawing attention to the plight of the Jews in the Second World War (Cornwell, 1999). In 1996, Cornwell had also written The Power to Harm, a book on the risk of suicide and homicide on Prozac (Cornwell, 1996). Angels & Demons aficionados know that tangling with the Vatican is not a good thing. In Cornwell’s estimation, while the Vatican were unhappy with his book, they were no problem compared with Eli Lilly, who apparently threatened to sue him in 50 different countries (Cornwell, 2003 personal communication).
In Ireland through to the 1980s, the influence of the Catholic Church was such that every major Irish author of fiction had works that were banned by the State (Carlsson, 1990). The censorship was as comprehensive as anything in Nazi Germany or Stalin’s Russia. This seems unbelievable now, but when it comes to outlining treatment-related adverse events, everyone reading this article is living through something quite comparable.
When it comes to conflicting interests, doctors face both carrots as well as sticks. The pharmaceutical industry bills itself as the greatest investor in research in key Western economies – up to 40% in the UK. And the pensions of most doctors likely depend on the performance of pharmaceutical company shares, so causing trouble for pharmaceutical companies is not in our interest.
From the United States, there are daily reminders that this is the most polarized Congress in history with Democrats and Republicans unable to agree on the day of the week. The one thing both sides agree on is speedier access to medical drugs and devices, and there is currently a Bill before Congress to speed this up and to get the FDA to take into account not just the efficacy and safety of drugs but also the fact that drug manufacturing leads to jobs in the United States (Scott, 2012).
In the UK, a recent edition of New Internationalist, a Marxist publication, comes with a standard diagram of the type Pharma would be proud, suggesting that 15–20% of us are mentally ill and in need of treatment (Godrej, 2012).
Against this background, it is difficult to see how medical and industrial interests can conflict. Even having a provocative piece like this does not trouble marketers, who have learnt how to brand Che Guevara to sell goods (Heath and Potter, 2005). Some years ago after a talk I gave on antidepressants and suicide, a public relations person responsible for Prozac greeted me afterwards with: ‘You’re David Healy. I am so pleased to meet you. You are doing more for the sales for Prozac in the UK than anyone else’.
Real conflicts of interest
If declarations of links to industry amount to tokenism, what might a statement covering real conflicts look like?
In 2005, I received documents from Eli Lilly under a freedom of information request. Document 103 stated [put] ‘third parties in the audience to monitor what he says and see whether he can be sued’. While this is in a company document, ‘third party’ means clinical colleagues. In 2006, I had a letter from the General Medical Council to investigate my behaviour on the basis of complaints made by colleagues – not by industry. All of the correspondence above is available online (Healy, 2012b).
In 2011, at a lecture at the Royal College of Psychiatry annual meeting, Professor David Nutt gave a talk with a slide ‘No Psychiatry without Psychopharmacology’, which said that psychiatry is under threat from three sources: first, treatment deniers like Irving Kirsch or Joanna Moncrieff; second, from those who regard addictions as lifestyle options rather than diseases; third, from scaremongers – where he listed me. There is no evidence I have ever denied that any psychotropic drugs work and I would agree with the notion that for medical colleagues wanting to get pregnant but hooked to antidepressants being unable to stop is not a lifestyle option.
Terming me a scaremonger seems to be an oxymoron, in that anyone who makes claims that cannot be supported about problems caused by drugs is likely to be sued but warning about problems that do exist is in fact why drugs are available on prescription only. In contrast, scaremongering is often how companies market medicines – if you don’t treat your children with antidepressants they are going to grow up to be drug abusing, alcoholic, career failures, divorced and suicidal (Newsweek, 2002).
On the basis of my experience then, if the issue is keeping patients safe, a conflict of interest statement might include blocked promotions, a lack of invitations to participate in events, academic stalking and libel. While negative experiences like this clearly can cloud judgements, when the issue of conflicts of interest is discussed it is more often framed in terms of gains. On this basis, my conflict of interest statement should also include the fact that I have had more help from colleagues within industry than from clinical colleagues when it comes to bringing adverse events to light.
Whence clinical conflicts arise
In 1962, following a disaster with thalidomide, we put a system in place to ensure the safety of drugs. It is this system, I have argued elsewhere (Healy, 2012a), that is leading us collectively to tear ourselves apart and individually to serious conflicts of interest.
Thalidomide was an over the counter drug in Germany, where its problems were first noted. After thalidomide, one safety step was to make all new drugs available on prescription only, thereby putting them in the hands of a profession sceptical about over the counter drugs. This was done because there was every reason to think that, like thalidomide, new drugs would be riskier than over the counter drugs or alcohol and nicotine that people manage for themselves.
We complain about academics getting a few thousand pounds for lectures even though there is little evidence that money leads academics to change their views. Academics get asked to talk because of the views they have and not the other way around. In contrast to the small fees academics get, clinicians in the UK earn about £100,000 a year primarily because they are entrusted to manage risky drugs on prescription only. But far from managing these risks, we have become a risk laundering system, so that the major hazards of a range of drugs from dopamine agonists to hypoglycaemics are accepted by us on average 10–15 years after patients have established their occurrence. Had thalidomide been prescription-only, it would likely have remained on the market for a good deal longer than happened.
Risks get laundered through another safeguard put in place in 1962 – randomized controlled trials (RCT). As of 1962, one drug had been through a placebo controlled RCT before coming to market. This had demonstrated thalidomide was safe and effective. RCTs can make a contribution to demonstrating efficacy or rather a lack of efficacy, but RCTs undertaken in medical conditions are close to the most perfect way imaginable to hide adverse events. If a selective serotonin re-uptake inhibitor (SSRI) causes anxiety, for instance, this can be hidden against a background of anxiety. Doctors can even be persuaded that schizophrenia rather than olanzapine causes diabetes (Le Noury et al., 2008).
Fifty years after thalidomide, most pregnant women avoid over the counter drugs from coffee, tea, and alcohol to nicotine, and many foods such as soft cheeses or uncooked meats. But antidepressants have become the most commonly taken drugs in pregnancy because doctors tell women that if they don’t have their nerves treated their babies will have birth defects (Healy et al., 2010). We tell them this even as the evidence mounts that antidepressants double the rate of birth defects and miscarriages and appear to cause learning disabilities in the children exposed to them in utero (Healy et al., 2010).
Fifty years ago, cars didn’t come with seatbelts but they now have airbags and some won’t start unless you have your seatbelt on. Fifty years ago, in contrast, drugs were often labelled as poisons, but now if I cite Paracelsus’ dictum ‘every drug is a poison, and the art of medicine lies in finding the right dose’ in a legal report, lawyers for the pharmaceutical company will get it struck out as prejudicial against their client. We cannot say a drug is a poison anymore. Instead, we have proposals to stamp images of pregnant women on antidepressants to overcome the scruples women may have about taking them (Koren, 2007).
This extraordinary situation has come about because prescription-only arrangements make doctors the ultimate consumers. But doctors consume by putting drugs into patients’ mouths and so consume without side effects. Companies meanwhile focus the biggest marketing budgets on earth on this small group of consumers. The system puts a premium on efficacy rather than safety to the point that doctors now blithely put patients on 10–15 supposedly efficacious drugs apparently blind to the exponential increase in risk involved.
Aside from the risk to patients from risk laundering, the worry for doctors is the risk of professional suicide. When the FDA proposed a Black Box warning for antidepressants, the American Psychiatric Association issued a news release stating ‘The American Psychiatric Association believes that antidepressants save lives’. They should have written that ‘The American Psychiatric Association believes that psychiatrists save lives’ (Healy, 2012a).
If the drugs work well and have few significant side effects, it doesn’t take an expert to manage them. Nurses, pharmacists and others are increasingly prescribing, are cheaper than doctors, and can more readily be constrained to keep to guidelines. Any sensible provider of health care is pretty soon going to consider employing them rather than doctors. Against this background, the American Psychiatric Association’s support for antidepressants rather than psychiatrists is a suicide note.
Consider the dilemmas posed by the NICE guidelines for schizophrenia (2002) and later bipolar disorder (2006), which recommended Zyprexa first line. NICE’s 2002 recommendations appear to be based on 234 publications on this drug. These 234 publications come from the four clinical trials that brought this drug to the market, but none of these publications contain the data on the capacity of this drug to raise glucose and lipid levels, cause weight gain or trigger suicide, making it impossible I believe to prescribe this drug with informed consent (Healy, 2012a). I have never prescribed it. Why do NICE take a different view? I have been told that Lilly threatened to pull out of the United Kingdom unless their drug featured prominently in the guidelines (Healy, 2012a).
Hence, when drawing up a Professional Development Plan in 2010, I listed one of my goals as avoiding loss of a job for practicing good medical care. Doctors increasingly work in a world where our interest to provide good medical care is likely to conflict with healthcare managers’ interests to have us conform to guidelines. Now there’s a conflict of interest.
Guidelines and protocols are part of an industrialization of heath care. One expression of these developments came to me in a recent job planning invitation to consider how ‘measure compliant’ my practice is. We are losing our capacities for professional discretion and discernment here.
P-artisans?
Is there a survival route? Pinel’s famous dictum had it that ‘It is an art of no little importance to administer medicines properly, but it is an art of much greater and more difficult acquisition to know when to suspend or altogether to omit them’. These days it is algorithms rather than art that get patients on medicines. But it still takes a doctor to know when not to prescribe or when to stop medicines. No algorithm can do this. The market cannot understand ‘No’.
As opposed to the mass-production that gets people on medicines, we need to specialize in tailoring treatments to patients. In many areas of consumption, goods or services of this sort are branded as artisanal. Given that doctors should be offering professional discernment as a service and should be advocates for their patients and for possible non- consumption, the complex of issues involved might be caught in the term ‘P-artisans’. Alternately, we may need to develop ‘scaremonger pride’. Perhaps to adapt Dave Nutt, our slogan should be: ‘No psychiatry without scaremongering’.
Fifty years ago, doctors and pilots were both viewed as being in the business of keeping us safe. Pilots are critically concerned with our safety because if we go down, they do too. In contrast, if a patient goes down, doctors can blame it on the patient’s illness rather than the drug. Perhaps linked to this, when pilots report adverse events, their reports are taken seriously and lead to change. Doctors don’t report adverse events but if they do their reports are filed as anecdotal and lead to no change. If the public suspect we aren’t as concerned about their safety as pilots are, we may be in big trouble. Patients don’t care about getting from Sydney to Perth 30 minutes earlier or having their antidepressant work 2 days earlier, they are concerned about survival (Fried et al., 2011).
If we became experts in treatment rather than just a conduit for drugs, Pharma might even welcome us out of the closet. The best way to discover new drugs remains through noticing adverse events. It may be no coincidence that as adverse event reporting has atrophied in recent years, company pipelines have dried up.
