In Orders from Nowhere, Vaughan Gething suggested writing to Andrew Dillon of NICE. The correspondence had an Xmas Eve denouement – hence the title.
It seems everybody figures NICE hold all the answers. The doctors who treated Stephen O’Neill claimed to be keeping to NICE Guidelines and only using drugs licensed by MHRA. They also conceded that Stephen was suffering from sertraline toxicity. See if you can spot any hint of NICE Guidance on managing SSRI induced suicidality in this correspondence. Get in touch if you know of any drug any regulator in any universe has approved for the management of SSRI induced suicidality.
Healy to Dillon December 11th
Re NICE Guidance
Dear Andrew Dillon
I recently sent versions of the attached letter to a number of Ministers/Departments of Health. Vaughan Gething suggested taking “my” concerns to you and to June Raine – attached. Mr Gething’s Department may not understand MHRA’s brief and so I am copying Dr Raine in here rather than writing separately.
The correspondence arose from a recent inquest at which I testified which returned a conclusion I regard as wrong and potentially compromising of the wider public health. This led to the attached letter to Michelle O’Neill MLA.
A key reason for this incorrect conclusion lay in the patient’s GP and later hospital doctors testifying they did not believe the literature on antidepressants was ghostwritten and they were fully adherent to NICE guidelines. They also stated they had used drugs approved by MHRA, internationally recognised for their excellence in these matters – the implication seemingly being that if approved by MHRA these drugs could do little harm.
Full adherence to NICE Guidelines and using a drug approved by MHRA seems to have persuaded the coroner to override a conclusion that I’d expect most people exposed to the facts of the case would have adopted.
A group of good doctors in this case have been misled by misunderstandings about NICE. In 2004, a NICE group working on Guidelines for the treatment of paediatric depression, drew attention to this hazard in an editorial in the Lancet – Depressing Research.
I testified about this hazard a few weeks later to a House of Commons Select Committee and I have been doing so at regular intervals since under oath in court, in peer-reviewed articles, in books and lectures – for instance in a public lecture in the Welsh Senedd building a year ago – and in an open letter to David Haslam earlier this year and in recent letters.
I can see constraints on any response you might offer and it’s not the job of NICE to police the academic literature. I will write to Mr Gething, telling him it’s my job not yours to alert him to this risk. The public though will likely want someone to tackle it.
Professor David Healy MD FRCPsych
From: Andrew Dillon, Sent: 20 December 2019 12:22
To: David Healy (BCUHB) ; ‘firstname.lastname@example.org’; et al
Subject: RE: Letter to M Hancock
Dear Professor Healy
Thank you for this and your earlier emails. I am sorry that you didn’t receive a response to your letter to David Haslam. He would have referred it to me for action and so the responsibility is mine.
The two main themes in your correspondence touch on the notion of ‘bureaucratic medicine’ and the provenance of the evidence that NICE takes into account when developing its guidance.
I know that you will be familiar with David Sackett’s definition of evidence-based medicine (BMJ 1996; 312:71-72). His description of a “conscientious, explicit, and judicious use of current best evidence in making decisions about the care of individual patients” and his view that “the practice of evidence-based medicine means integrating individual clinical expertise with the best available external clinical evidence from systematic research” continue to be an important reference for NICE. Our guidelines are just one element of evidence-based care, alongside clinical judgement and the values and preferences of individual patients, a position we capture with these words, which appear in each guideline:
The recommendations in this guideline represent the view of NICE, arrived at after careful consideration of the evidence available. When exercising their judgement, professionals and practitioners are expected to take this guideline fully into account, alongside the individual needs, preferences and values of their patients or the people using their service. It is not mandatory to apply the recommendations, and the guideline does not override the responsibility to make decisions appropriate to the circumstances of the individual, in consultation with them and their families and carers or guardian.
The issue of unpublished evidence is a challenge for anyone using evidence to inform treatment options. As you point out, the authors of our Depression in Children guideline drew attention to the problems caused by unpublished evidence some years ago. Since this guideline was published, there has been some progress in addressing the issue of unpublished evidence and positive bias in the published literature. Researchers are expected to follow the Good Publication Practice guidelines, which includes medical writers being named in the acknowledgements of papers.
Since 2016, the EMA has made all clinical data submitted to them for the marketing authorisation applications of new medicines available through a dedicated website (https://www.ema.europa.eu/en/human-regulatory/marketingauthorisation/clinical-data-publication).
This means that all clinical data that was used by regulators is accessible to everyone within 60 days of marketing authorisation (after registering for access to the database), whether or not is has been published.
Of course, these procedures were not in place when most SSRIs were approved, but post-marketing surveillance can trigger regulatory action and the EMA completed a referral procedure on SSRIs and suicide in 2005:
We agree with you that it is not NICE’s role to ‘police’ the academic literature, but we take the matter seriously and we are a AllTrials campaign which calls for all past and present clinical trials to be registered and their full methods and summary results reported.
Chief Executive National Institute for Health and Care Excellence
From: David Healy Sent: 23 December 2019 14:22
To: ‘Andrew Dillon’; ‘email@example.com’; et al
Subject: RE: Letter to M Hancock
Many thanks for your response, which I read as an invitation to push at a partially open door. I might agree with your characterization of my concern as being with “bureaucratic medicine”, if you are willing to extend bureaucratic to include management. Health seems to me to have become a service industry and it’s the management (bureaucratic) processes common to such industries that are now putting us at risk.
In August 2018, I wrote to Vaughan Gething detailing similar concerns. I attach this letter – omitting the local service issues it also covered.
Dr Frank Atherton, the CMO in Wales, responded (letter attached) saying:
I recognise some of the structural constraints you describe – these are not easily corrected, appearing as they do in all advanced healthcare systems. My suggestion would be for you to continue to engage with others who share similar views; collective analysis and advocacy are more likely to change systems than individual efforts.
With this advice in mind, and your response, this email seeks to engage a wider constituency.
We agree on what David Sackett said and I imagine you noted in my response to Leigh Smale that I referred to NICE Guidance rather than Guidelines. I am sure you are also aware that de facto NICE Guidelines have become something doctors either must, or feel they must, adhere to rigidly.
And even for doctors who view NICE as offering Guidance, the influence of the unpublished evidence can seriously skew that Guidance in a manner difficult to gainsay. By unpublished evidence I don’t mean trials that are never published, I am more concerned about trials published claiming efficacy and safety when the unpublished data suggests the opposite.
As regards unpublished evidence in general, I imagine the Dept of Health will be reassured by the points you make. I don’t wish to gainsay the impression that NICE are at least trying to do something, but I do need to spell out how from the point of view of someone concerned about safety, what has been done risks making the problems worse.
In the case of the articles on pharmaceuticals NICE works from, as you note Good Publication Practice boxes are more likely to be ticked than before – by the ghost writers. But the notional researchers (the apparent authors) are less likely to have seen or analysed the data from any of these trials than they were in 2004.
In agreeing with your suggestion that things have moved on since 2004, many readers of your email will likely forget that the pharmaceutical industry and health service companies have more incentive and resources to keep ahead of the game than anyone else – leading to “not easily corrected constraints that apply in most advanced healthcare systems” as Dr Atherton recognises.
In 2004, the BBC in particular were able to assist us all in tackling issues that appeared to be rotten apples in a barrel, but they seem unable now to get any purchase on what seem rotten barrel problems – tellingly illustrated for me in the failed efforts of two recent BBC programs to get answers from NICE on matters involving antidepressants and children.
Although NICE has access to regulatory material it didn’t have in 2004, this is not the raw data. Having been involved in the consultation process regarding access to data that EMA began in 2012 and liaising regularly with colleagues who have availed of EMA’s data offerings, I know that EMA now make available Company Study Reports (CSRs).
They do not make available Clinical Record Forms (CRFs) which are closer to the data. CSRs written in the late 1990s were likely more innocent than CSRs written since companies became aware that their CSRs might later be scrutinized. But, even in 2004, the account CSRs offered were only slightly less rosy than published articles that led to indictments for fraud.
Your suggestion that regulators do not operate on the basis of the raw data accords with my understanding. Their use of any extra data to which they might have access is for audit purposes rather than analysis. In this respect nothing has changed since 2004.
For someone concerned about the safety of drugs, the AllTrials initiative you mention compounds the problems. RCTs and AllTrials focus on efficacy. This focus makes RCTs the gold standard way to hide adverse events – as outlined in a recent BMJ article I co-authored, (as part of a collective effort to analyse the problem). In adding to the premium put on RCTs, the impression generated by AllTrials, that we are now forcing companies to make key data available, as I see it adds to the inability of prescribers to see the harms they are causing – as demonstrated in the O’Neill Inquest.
From my point of view, if AllTrials hadn’t existed the pharmaceutical industry might have had to invent it – and they were quick to sign up to it.
The post-marketing surveillance that might counter-balance the premium put on RCTs isn’t working. My reason for initiating this correspondence was that the recent O’Neill inquest is a telling example of how things are getting worse rather than better.
The SSRIs give the lie to the standard response given to soothe policymakers and others that of course there are some events too rare to be picked up by RCTs, or that happen only after months or years of exposure, for which we need post-marketing surveillance. With SSRIs a genital numbness happens in close to 100% of people who take these drugs, within 30 minutes of the first pill, but it was close to completely missed by RCTs and is still, 30 years later, not included in the labels of these drugs.
This is not a matter of improving our post-marketing surveillance methods – it’s a matter of re-empowering doctors and patients to notice problems happening commonly and in front of them – as Sackett would have wished – and developing methods to manage those problems as we did 50 years ago in the case of drugs like lithium but don’t do now. In the absence of any Guidance on how to handle treatment induced suicidality, the doctors in the O’Neill case flailed around and it seems to me contributed to a death that should not have happened.
I don’t hold industry or regulators to blame for this – I blame doctors and, in particular those who have held themselves out as experts on pharmacovigilance.
Most doctors now are too young to remember that 20 years ago we had a regular stream of publications like Drugs and Therapeutics Bulletins, reminding us that drugs are unavoidably unsafe. Very few of us read NICE Guidelines.
Now every medical student learns NICE Guidelines and parrots them back in order to pass their exams. These essentially only outline the benefits of treatments – and few younger doctors have ever seen a DTB.
Most doctors work in Family Medicine and depend on specialists like me – who are either the very people whose names are on the ghostwritten articles or whom they will feel too uncomfortable to ask back to speak if my message is at odds with NICE Guidance. I’m reluctant to speak to doctors in training as they’d fail their exams if they listened to me.
When it comes to adverse effects, regulators, in practice, strip away the names of patients from reports of problems, transforming these into hearsay and making it impossible for the regulator ever to link an effect to treatment. MHRA have had reports of Post-SSRI Sexual Dysfunction (PSSD) on their books for 30 years, and may now have over 1000 such reports, but seem incapable of regarding these reports of a problem that is leading young people to seek assisted dying as anything other than a signal.
We would all be a lot safer if RCTs, increasingly conducted in developing world settings, with invented patients, and notional investigators unable to attest to the reality of anything that has happened in the trial, were regarded as hearsay.
It would also help if regulators made the kind of efforts that companies make to determine whether their drug causes a problem, which involves consulting with patients and doctors as to what has happened and often deciding their drug has in fact caused a problem.
I have not gone into the details above in order to castigate NICE. We agree this is not primarily a problem for NICE –other than perhaps being seen as a fall-guy for doctors who don’t read or don’t understand the small print in the Guidelines. It’s rather difficult to pinpoint whose problem it is and I’d be grateful for any suggestions you might have in this respect.
As regards my efforts to engage a wider group of stakeholders, I will bundle up the recent correspondence about the O’Neill Inquest and ensure all Depts of Health in these islands get a copy. Unity on important issues seems lacking at the moment. Perhaps all 4 countries will be able to unite – in rejecting, or side-stepping, my concerns.
Other stakeholders that come to mind are some of the Royal Colleges, and the BMA.
From: Andrew Dillon Sent: 24 December 2019 10:32
To: David Healy (BCUHB); ‘firstname.lastname@example.org’ et al ;
I’m not really sure how to respond to this, to be honest. I’m not so naïve as to imagine that clinical trials are always either complete or completely accurate. I know that you’re not implying that, but it’s worth saying anyway. Our advisory committees are not uncritical either. It’s through the committee’s, applying our methods, that we can expose uncertainties and risks and try to reflect them in in our guidance. (by the way, when we refer to NICE ‘guidance’ we mean all the various forms of advice we publish, which includes our clinical, public health and social care ‘guidelines’).
I think you’re being a little hard on NICE. I imagine that you’re some way into your career, but people starting out tell us that they appreciate the steer that our guidance gives them. Interestingly, views are expressed on both sides of the argument about the extend to which our recommendations direct practice, as opposed to encouraging informed judgement. Both have their place, I think; the former particularly when there is a need to communicate a clear signal about safe practice. Through our Fellows and Scholars and prescribing advisors (clinical pharmacists) programmes, we reach significant numbers of health and social care professionals with a message about the extent and the limitations of NICE guidance. I haven’t formed the impression, in the contacts with the people involved, that they apply the guidance uncritically. If some young health professionals feel that they have to, better that they rely on our guidance than struggle without it, or worse, use less authoritative sources (if you’ll permit me to promote my own organisation).
I was interested in your point about ‘regulators (making) the kind of efforts that companies make to determine whether their drug causes a problem, which involves consulting with patients and doctors as to what has happened and often deciding their drug has in fact caused a problem.’ If by that you mean we should be tracking the effect of our recommendations, including the extent to which known as well as unanticipated risks have materialised, I would agree. Our ability to do that is limited partly by the availability of data and partly by resources. However, we are investing in an enhanced data analytics capacity, which we hope will put us in a better position to do this.
I don’t know who else you can approach to engage in this. The academic health science networks perhaps? Google, or one of the other new players in the data crunching business? Meanwhile, we’ll keep on going in our admittedly flawed way, to try and keep patients and health professionals safe.
NICE After Christmas
Unbeknownst to me, Andrew Dillon had indicated a few weeks earlier that after 21 years at the helm he was stepping down as the CEO.
He was widely praised by the Association of the British Pharmaceutical Industry (ABPI), among others, who mentioned his contributions to transparency.
David Haslam, the Chair of NICE, along with Andrew Dillon, an honourable man, is also stepping down.
Twas the NICE before Christmas
When all through the house
not a creature was stirring not even a mouse…
When it comes to the adverse effects of drugs, there is not a stir. The liberal media – the Guardian and New York Times, the BBC and PBS once stirred the shit. No longer. We don’t want to stop people believing in Xmas. They can’t be put off taking the sacraments – things that can only do good and cannot do harm.
And this is in great part down to NICE a creation of the Labour Party – who simply do not want to hear that the academic literature on which Guidelines are based is Fake through and through. A series of honourable men have presided over this, award winning journalists won’t go near the issue, Brenda Hale who tackled Boris Johnson doesn’t want to know and doctors… well doctors.
Meanwhile in terms of Guidance on managing SSRI induced toxicity, or any drug toxicity, you’d be better off asking St Nick before he takes off in his sleigh than anyone in NICE or MHRA, FDA or EMA.
Ask your doctor? Well doctors… BMA have been copied in to all this. Not a peep.