Bad Pharma

by Ben Goldacre

  If you have an ongoing medical problem, there will be a patient group that covers it, run by people with patients’ interests at heart. There are problems with some of these groups, as described above, which you could address, but I would strongly encourage a different path: you can join them, and then encourage these groups to lobby the companies with which they have relationships.

  For example, there is one important letter that every special-interest patient group should be sending to every drug company in the world, containing a simple query: ‘We are living with this disease: is there anything at all that you’re withholding? If so, tell us today.’ This letter serves two purposes. Thinking optimistically, it might prompt a declaration: someone might disclose trial data that they were previously withholding, and this will improve patient care. But if they don’t, and they have something they should be sharing, then you have still done something valuable: you have created anxiety; you have forced someone to attach their name to the responsibility of misleading you; and you have put a clear date-stamp on a company’s ongoing dishonesty. If a company denies withholding trial data on drugs for your disease today, in 2012, but is then caught out in 2014, and puts out a press release saying ‘Everything has changed now,’ you will know, for certain, that it was still willing to mislead and harm patients in 2012.

  Patient groups

  There is much more that patient groups can do, in their role as collective organisers, and I would strongly encourage these groups to meet, and consider what they can do to address the issues in this book, using the unique resources they have. At present, for example, it is nobody’s job to monitor missing trial results: so even though we have huge registries filled with details of ongoing studies, nobody is flagging up the trials which have completed, but failed to publish. We should remember that it was independent academics, investigating on a whim, who discovered that only one in five trials had met the reporting requirements of the FDA’s new law from 2007. The absence of proper, centralised compliance auditing for withheld trial results is a catastrophic failure in the information architecture of evidence-based medicine, but since it hasn’t been fixed, patient groups are in a very strong position to do the job.

  They can act as observers for their own area, monitor the registers, look at the completion dates of trials, and then hunt for the publications. If researchers fail to produce results within a year, patient groups should first name them – since this carries a valuable public sting, which can change behaviour in future – then contact them, and ask for the data that will improve their members’ treatment. Patient groups are also in a strong position, with their extensive grass roots membership, to find out about trials which are being conducted, but which have not been placed on a trials register. If there are patient groups who are willing to address the problems in this book, I would be happy to work with them to help develop further interventions, as would many other doctors and academics.

  Doctors

  Medics, in my view, need to think and talk about these issues much more, share what they know, and act. This might mean a number of things, as discussed earlier in the book: individuals could avoid industry marketing; declare what they’ve had to their patients; decline sandwiches and free flights; and so on. They could also engage with the senior figures in their professional societies and Royal Colleges, to try and encourage them to step back from the current dangerous positions that most hold.

  Medical schools

  Medical schools can teach medical students about how to spot bad evidence from the pharmaceutical industry, and in particular, how its marketing techniques work. There is some evidence from the US that students taught about these techniques are better able to spot distortions in promotional material, and this deserves much more concerted work: the current generation of trainee doctors will go on to practise medicine independently for at least three decades, without any further formal teaching. If we do not future-proof them, they will be taught by industry, with the encouragement of the government and – reading the latest collaborative document – with the encouragement of every eminent medical body in the UK. If there is to be any hope of defending the medical profession against the technical distortions used as marketing techniques by industry, young doctors must be trained to identify them.

  Ghostwriters

  Commercial medical writers – and the International Committee of Medical Journal Editors – need to fix their ridiculous guidelines, because everybody knows that they still permit ghostwriting to happen. Commercial medical writers could organise an amnesty where they expose every paper they’ve covertly written, and every ghost author they’ve ever paid, on ethical grounds, for the protection of patients. They won’t, but they could.

  Lawyers

  In the US, individuals and the state are better able to take action against those who have harmed them, often reframing the issue in terms of financial fraud. Drug companies are not the only target here, and many have argued recently that ghostwritten articles also present an opportunity.7 If a patient is injured, when their doctor is relying on the content of an article that was covertly manipulated, then the commercial medical writers, the ‘ghost authors’, could be held liable. More than that, the ‘guest authors’ – the academics who allowed their names to be put onto these papers, despite minimal contribution, often in exchange for money – might also bear responsibility. If an academic paper is used by Medicare or Medicaid, in the US, to justify off-label use of a drug, but that paper subsequently turns out to be ghostwritten and distorted, then, again, the authors may be liable for this act of fraud, perpetrated on the government. There are also anti-kickback laws to consider, and a clear precedent that the First Amendment right to freedom of speech does not shield fraud. This could rapidly become an interesting avenue.

  Journal editors

  Journal editors are the current gatekeepers for medical evidence, and they have dropped the ball. All journals should declare their industry income, in full, and no journal should permit any trial to deceitfully switch its primary outcomes: this practice misleads doctors and harms patients. All journal articles reporting unregistered trials should state this fact clearly, and the ICMJE should publicly declare that they have failed to police this practice, so that others know to fix it properly.

  The pharmaceutical industry

  There is a great deal to cover here, and much has already been said in the book, but there is a special call I would like to make, to the many good people who work in industry. It’s possible for companies and professions to be structured such that good people participate in projects which overall do great harm, without necessarily ever knowing. I strongly encourage you to familiarise yourself with the activities of your company, and the real details of any legal cases brought against them, and criticisms made in the academic literature.

  I also strongly encourage you to become a whistleblower, whenever you see things being done that are wrong, using three main vehicles, in ascending order of melodrama. At the simplest level, if you can handle the secrecy properly, write an anonymous blog, explaining what you see on a daily basis: the banal, mild distortions, the days when you’re asked to hunt in a data file and cherry-pick any pattern that makes your company’s drug look good, the unofficial sales advice you’re given as a drug rep, and so on. Next, I’d appreciate specific leaks, on ben@badscience.net (but please don’t send me confidential information from your work email address). Lastly, many of you reading this have access to large quantities of data or documents that would change the lives of patients, and help to prevent ongoing suffering and harm. I would appreciate a data dump, on the scale of the US Iraq and Afghanistan war records, and if I’m honest, I’m surprised and disappointed this hasn’t happened yet. If you need any help, just ask: I will do everything I can to assist you.

  Professional bodies

  The Royal Colleges, the faculties, and the societies, have failed us. There is still no professional body in the UK, other than the tiny Faculty of Pharmaceutical Medicine,
that has stood up and declared that withholding trial data is unethical, immoral and grounds for expulsion of its members. If these organisations have integrity, they will fix this; alternatively, if the senior members truly think that withholding trial data is fine and acceptable, then they should come out and say so, clearly, to their members and to patients. I’m not the first to raise this, so I have low expectations. If you are a member of one of these organisations, you could write and ask why it does not have a policy of chastising and expelling people who harm patients by withholding trial data. Please, send me their responses, or better still, simply post them online.

  Funders

  Money is short, outside of industry-funded research. Organisations like the National Institute for Health Research already do a great job, funding trials on important questions that drug companies won’t want to pay for: it assesses the benefits of older drugs, for example, or of treatments that don’t involve commercial products (I should say that I sit on one of these funding committees). But I believe that public funders should have two further priorities. Firstly, there are some small outstanding evidence gaps in our knowledge of how industry distorts medical prescribing practice, and this is not a field that the pharmaceutical industry is likely to fund. More than that, as I argued at the beginning of the marketing section: the great horizon for evidence-based medicine is finding new methods and tools for disseminating the evidence we already have, to doctors and other decision-makers. This will require innovative collaboration between coders, pharmacists, librarians, doctors and academics. At the simplest level, I’m always frustrated that I can’t press a button, when reading a systematic review, to say ‘notify me when this summary is updated with new trial results’. In more elaborate terms, it’s clear that we should be moving medical knowledge and trial results into structured databases; and perhaps embedding high-quality contextual advice into a doctor’s routine workflow, as they sit at the computer in their office.

  Academics and nerds

  This book is filled with exciting untapped areas. There are the big jobs, on organising medical knowledge, from the section above, but also an avalanche of smaller studies, many of which could be done as undergraduate dissertations. Do a fact-checking audit of claims from drug reps; gather quantitative evidence on industry sponsorship in your medical school; find out about your university’s policies on hiding trial data (or anything else); and then collaborate with other medical schools, to gather comparative national data. Share your ideas, and publish your results: we are all cheering you on.

  GLOSSARY

  ABPI – The Association of the British Pharmaceutical Industry.

  ACCME – The Accreditation Council for Continuing Medical Education.

  BNF – British National Formulary.

  CME – Continuing medical education.

  CSR – Clinical study report.

  EMA – The European Medicines Agency, the European drugs regulator.

  FDA – The Food and Drug Administration, the US drugs regulator.

  GCP – Good Clinical Practice guidelines.

  Generic drug – When a drug is first invented, it is owned by the company, and no one else is allowed to manufacture and sell it. A drug usually comes off patent eighteen years after it was first registered, and about ten years after it comes onto the market. After that, any company can make a copy of it, as with very old drugs such as paracetamol or aspirin. When this happens, you can’t make much profit out of selling it.

  GMC – General Medical Council.

  ICH – International Conference on Harmonisation.

  ICMJE – International Committee of Medical Journal Editors.

  JAMA – Journal of the American Medical Association.

  KOL – ‘Key opinion leader’.

  MHRA – Medicines and Healthcare products Regulatory Agency.

  NEJM – New England Journal of Medicine.

  Off-label – A drug is allowed onto the market by a regulator on the condition that it can only be marketed for use in a specific medical condition. Doctors can use the drug for other medical problems if they wish, but this use is ‘off-label’.

  Off patent – A drug that is out of patent (see generic drug).

  PMCPA – Prescriptions Medicines Code of Practice Authority.

  Publication bias – The phenomenon whereby trials with results regarded as unflattering or uninteresting are left unpublished.

  SSRI – Selective serotonin reuptake inhibitor.

  Surrogate outcome – Real-world problems are what we really want to change with medicines: things like death, heart attack or stroke. A surrogate outcome is something that is easier to measure, like blood pressure or cholesterol levels, that we hope is a proxy for those outcomes. Often they are not as good at predicting real-world outcomes as we think.

  ACKNOWLEDGEMENTS, FURTHER READING AND A NOTE ON ERRORS

  I have been taught, corrected, calibrated, cajoled, entertained, encouraged and informed by a very large number of people, including John King, Liz Parratt, Steve Rolles, Mark Pilkington, Shalinee Singh, Alex Lomas, Liam Smeeth, Josie Long, Ian Roberts, Tim Minchin, Ian Sample, Carl Heneghan, Richard Lehman, Dara Ó Briain, Paul Glasziou, Hilda Bastian, Simon Wessely, Cicely Marston, Archie Cochrane, William Lee, Brian Cox, Sreeram Ramagopalan, Hind Khalifeh, Martin McKee, Cory Doctorow, Evan Harris, Muir Gray, Amanda Burls, Rob Manuel, Tobias Sargent, Anna Powell-Smith, Tjeerd van Staa, Robin Ince, Roddy Mansfield, Rami Tzabar, Phil Baker, George Davey-Smith, David Pescovitz, Charlotte Wattebot-O’Brien, Patrick Matthews, Giles Wakely, Claire Gerada, Andy Lewis, Suzie Whitwell, Harry Metcalfe, Gimpy, David Colquhoun, Louise Burton, Simon Singh, Vaughan Bell, Richard Peto, Louise Crow, Julian Peto, Nick Mailer, Rob Aldridge, Milly Marston, Tom Steinberg, Mike Jay, Amber Marks, Reg, Mum, Dad, Josh, Raph, Allie, and Lou. I’m hugely indebted to the late Pat Kavanagh, Rosemary Scoular, Lara Hughes-Young, and especially Sarah Ballard, who is amazing. Robert Lacey has copy-edited my last two books, he is gentle. Louise Haines has been mighty.

  There are tools that made this book better, especially Zotero, Scrivener, Evernote, ReaditLater, IntervalTimer and Repligo. AntiSocial is a piece of software that irreversibly disables Twitter and Gmail on your computer when you’re working: I highly recommend it. In recent years I’ve had day jobs supported by the National Institute for Health Research, the Scott Trust, the Wellcome Trust, Nuffield College, Oxford, and the NHS, and a bursary from the Oxford University Business Economics Programme.

  It’s a common joke in evidence-based medicine that whenever you think you’ve had an idea, Iain Chalmers has probably written it up in an essay fifteen years earlier. He helped to formulate many of the key ideas in evidence-based medicine, and to spot the problems, and I hope I’ve attributed enough to him. There are also many other academics whose work comes up repeatedly: some of them I’ve met, most of them I’ve not, but where you see recurring names in the references, it’s fair to say that we all stand on their shoulders. There are huge rewards in medicine, but they’re not reliably in the right places. Many of the people whose work is cited in this book have taken a personal hit, on income and eminence, to work on the serious systemic problems in medicine. They are quiet heroes. It’s an honour to spread their work more widely.

  There are many excellent review papers on the issues raised in this book, and I have highlighted them in the references wherever possible. I have specifically tried to seek out papers that are freely accessible (look for the references to a journal called PLoS, in particular), though some sadly are still behind academic journal paywalls.

  There are also many excellent books that have covered some of the issues around bad behaviour by the pharmaceutical industry, though all are US-focused, many are now almost a decade old, and none have focused missing data. If you are keen to read more on any specific area, there are several books that have influenced my thinking over the years.

  Jerome Kassirer was editor of the NEJM, and his On the Take (2004) is great on the marketing issues, and t
he way that continuing medical education in the US has been captured by industry. Marcia Angell was also NEJM editor, and her book The Truth About Drug Companies (2005) first brought the story of marketing, institutional corruption and bad evidence to a larger audience. Richard Smith is the previous editor of the BMJ, and his book The Trouble With Medical Journals (2006) explains itself. Ray Moynihan’s various books on medicalisation are all excellent. Donald Light recently edited The Risks of Prescription Drugs (2010), which is a good collection on contemporary problems, especially the lack of innovation. Melody Petersen, previously of the New York Times, wrote Our Daily Meds (2008), which is excellent on marketing in the US. Daniel Carlat is a bioethicist, and his White Coat Black Hat (2007) is great on the ethical issues around drug testing. Tom Nesi’s book on Vioxx is magnificent.

  While criticising bad behaviour from industry is important, I’m also struck that the public have little opportunity to read about the basic techniques used to appraise new treatments, find out what works, and spot what harms. Testing Treatments (2006, second edition 2011) by Imogen Evans, Hazel Thornton, Iain Chalmers and Paul Glasziou remains the single go-to book on this topic in my view, published in several languages and also available free online at testingtreatments.org (I should mention that I wrote the foreword). Powerful Medicines (2005) by Jerry Avorn is the first attempt I’ve seen by a pharmacoepidemiologist to explain the science of side effects monitoring to the public. How to Read a Paper by Trisha Greenhalgh remains the medical-student bible on critically appraising academic papers, and it can be understood by anyone.