Goldacre is right - greater honesty is needed
One of the most trenchant critics of pseudo-science in healthcare, the neuro-psychologist Dr Ben Goldacre, has turned his guns on the pharmaceutical industry.
I am only part way through ‘Bad Pharma’ (Fourth Estate, London), but I do know Ben and have heard him speak several times. The current theme emerged a couple of years ago, when he supported a debate at The Oxford Union on this topic. I have to say that, after 38 years working with big pharma, the whole issue causes me considerable pain. Let me outline why.
On the opening page there is the following statement: “Industry-funded trials are more likely to produce a positive, flattering result than independently-funded trials.”
This establishes the theme for the book; drug companies distort and even hide research data, and thus mislead regulators, doctors and patients. But I am not writing a book review here, and in any case I am quite familiar with some of the abuses that Ben sets out. If there are any surprises, they would include the vast scale of book-cooking that seems to go on.
A casual reader could get the impression that there is really nothing that drug companies say about their products that we can believe. That would jar with those of us who work in drug development, and who know the huge grinding effort that goes into getting the data right.
But the fact is inescapable that the process does go wrong, sometimes deliberately so, because Bad Pharma is well enough referenced to back up the overall message.
A painful necessity
That such a book could be written is a tragedy. I spend most of my spare time in voluntary work supporting evidence-based medicine, for example with the charity HealthWatch. The many opponents of real medicine are having a field day, claiming that drugs do far more harm than good, and citing Goldacre in support.
Some of you might have seen a new magazine on the newsagent’s shelves, called ‘What Doctors Don’t Tell You’. It says nothing good about real medicines, and is full of false claims for ‘alternatives’. The editor dismisses previous books by Goldacre on the fantasies of homeopathy, chiropractic etc., and delights in his broadside against big pharma (and ignores his praise of pharma’s major contributions to public health).
So what do we as an industry think about all this? Presumably the ABPI speaks for the UK research-based companies, and in a press statement devotes 63 words to addressing what the book actually says, finishing with: “...the examples he refers to are long documented and historical, and the companies concerned have long addressed these issues.”
It continues with 557 words on how the industry is heavily regulated, how difficult and expensive drug development is, how it is meeting unmet medical need etc etc. I and many of my colleagues do not see that as a balanced and constructive response.
Indeed, the clip reproduced above is demonstrably untrue. Publication bias continues, studies are still disappearing from public view, and there are still big gaps in safety data. I feel badly let down by such complacency.
Sell, sell, sell....
It would be rather facile to say that everything is fine in R&D, and the problems only start when the data get to the marketers. I am reminded of something I heard from one of my competitors, in the early days of my first business.
The gist of it was that the content of one’s offering was much less important than how hard it was sold. That seems to be a message that has struck home, with global sales and marketing spend about twice that of R&D.
Nevertheless, I am realistic enough to know that no matter how good the product, it simply won’t get used unless someone is paid to sell it. Companies will do whatever they can to sell their products, so I don’t think we can ever legislate to stop them from spending their own money on that.
What we can do though is to ensure that what the sales reps say is decent, legal, honest and truthful. Perhaps the messages will become much simpler for having to rely only on solid data, and will cost less to deliver.
Meanwhile, everything clearly is not fine in R&D. Clinical trials absorb the major part of resources for that, and for decades have been governed by various flavours of Good Clinical Practice. But it has dawned on me that it’s perfectly possible to have 100% compliance with GCP, yet still do bad science and mislead the consumer.
For example, does it add to our knowledge to carry out a placebo-controlled trial when there is an effective treatment already available? Yet that is what I have found myself doing, because that’s the study that the regulator wanted. Surely what we want to know about a new drug is that it offers an advance in patient care, not that it is better than doing nothing.
Even with active-controlled trials, there are ways of doing bad science. It is unlikely that the study drug and the active comparator will have identical response profiles. So there is scope to make your drug look better, by carefully selecting the time point which you can designate the primary response parameter. Don’t laugh, I have seen this done.
The regulator (a different one) bought it hook, line and sinker. For such trials as these, millions were spent on gruelling GCP compliance, necessitating many thousands of air miles, and rooms full of paper files. Surely I am not alone in questioning the sense of all this?
A broken model
It is becoming more widely appreciated that the current GCP model is not sustainable. The concept of the data trail, which emerged from the original FDA GCP guideline of 1977, is firmly rooted in paper-based systems. Any clinical research associate knows that there is no guarantee that all data in the study database will have originated in the patient’s medical records.
When errors and omissions occur, we all know that gaps in the medical records are commonly filled in from data in the case report form, just to keep the auditor happy. People repeatedly fly into remote countries to monitor sites that only recruit one patient, or none at all. Hence we can easily be looking at £50,000 to £100,000 per patient to get the study done.
It is very widely accepted that because of this lack of realism, and many other problems that I have highlighted previously, GCP and the current drug development process are not meeting the expectations of a more sophisticated and challenging consumer base.
Ben Goldacre and his ilk have got wise to what is going on, and with Bad Pharma already among the best sellers his arguments, even if flawed in places, are falling on receptive ears.
Some commentators are asking whether the focus of R&D should not be on new products, as driven by pharma, but on quantified health gains. Thus, the placebo-controlled trial that I mentioned above would be worthless, as it could not demonstrate a health gain over the status quo. R&D would then have to proceed as more of a partnership between industry and healthcare stakeholders.
A new paradigm
How could this work? I can think of two main branches to follow. Firstly, clinical trials need to be far more relevant to clinical practice. Many medical journals are now demanding that authors provide a structured synopsis of their paper, including a clear statement of what we now know from their findings, that we didn’t know before, and how it will change clinical practice. This needs to be brought forward into the original proposal for doing the trial, including a firm commitment to publication.
Yes, a protocol has objectives, but they are very commonly more esoteric. Secondly, we need Good Scientific Practice as an integral part of GCP. This falls squarely into the laps of the ethics committees and the regulators, both of whom have been strangely silent so far about Bad Pharma.
To enable these things to happen, the crushing burden of GCP drudgery needs to be relieved, so that resources become available to do better science that matters to more people. That doesn’t mean lowering standards, it means raising them by refocusing.
We need to put the effort into where it will have the most benefit, and that means a risk based approach to trial management. Should we keep monitoring those unproductive investigators? I don’t think so, we should put them on hold and go where we have the most data.
It may well be that the regulators are alert to some of this. Last year the FDA published draft guidance on a risk-based approach to trial monitoring. There is increased emphasis on using intelligence about the study to govern what is to be monitored.
Such intelligence would be managed centrally, implying the need for advanced systems to analyse the study progress data feed in real time, and to inform managers of future potential problems.
We are in this together
Goldacre has shown us where our science is not good. Ethics committees and regulators now have no excuse for not spotting protocols that pulled the wool over their eyes before. One ethics committee member I know suggests rejecting submissions from investigators with a track record of failure to publish (as does Goldacre).
The legislators have the opportunity to include essential checks and balances to ensure that good science is done, and that patients are better protected by increased emphasis on where the biggest risks are, and much less on wasteful backwaters.
I am doubtful that either industry or regulators will drive this process alone, and it will need far more creative collaboration than I can recall ever happening before.
Les Rose is a freelance clinical research consultant and medical writer.