Making medicines safer for all of us

Adverse drug events are now the fourth leading cause of death in hospitals.

It’s a reasonable bet they are an even greater cause of death in non-hospital settings where there is no one to monitor things going wrong and no one to intervene to save a life. In mental health, for instance, drug-induced problems are the leading cause of death — and these deaths happen in community rather than hospital settings.

There is also another drug crisis — we are failing to discover new drugs. [Read more...]

Archive for Study 329

Study 329 Trick, Treat or Treximet

The plan this morning was to continue coverage of the Opioid Epidemic but perhaps because it is Halloween a link to a Wall Street Journal story on how Drug Makers Turn Cheap Generics into Expensive Pills arrived by email. It was difficult to resist. Here’s why.

The article features Treximet, a combination of sumatriptan and naproxen, used for migraine. These two drugs are ancient. They come from the last millenium and should cost almost nothing. But Treximet is retailing at $750 for 9 pills. The two drugs as separate ingredients can be bought separately for next to nothing.

This hike in price is heading toward the 5000% hike in price Martin Shkrelli hit the headlines for when he jacked up the price of Daraprim a year ago. This was noted at the start of a series of posts on Pharmaceutical Rape by Laurie Oakley.


Behind Treximet lies trials comparing the combination to sumatriptan alone and to naproxyn alone and to placebo showing some benefit it seems – if these trials can be believed. The trials were run by GlaxoSmithKline, for whom sumatriptan had been a flagship drug for decades, and Pozen Inc. Once concluded, in 2014 GSK transferred the rights to Pernix.

This smelt fishy to RxISK’s super-sleuth Johanna Ryan. What’s going on here? In related posts she also fingered the weight loss med Qsymia as egregious and both as indicators of something new this way coming and Qsymia features in the WSJ article also along with Duexis for pain, as price gougers extraordinaire. The new development seems to find new painkillers and new weight loss meds fertile ground. Regulators seem all too happy to license ever more dangerous variations or combinations of older drugs in these groups already on the market and causing serious problems.

This willingness has offered an opening for financial speculators to make a killing speculating on the likely profits these drugs could yield and getting out of the game before much happens. Pernix may be primarily an umbrella for some smart speculation of this sort. It certainly doesn’t look like a pharmaceutical company or anything to do with the actual delivery of healthcare.

Martin Shkrelli has since made it clear he’s been taking Efexor, and it made him feel so good, “invincible”, that he figured people will be more than willing to pay pretty well any price for these wonders of modern science.

Begs the question as to what the regulators in FDA are taking.

It’s more complex with Andrew Witty. You’d have to think GSK didn’t run the trials of Treximet just to hand it over to Pernix, no strings attached. What is the business model here. Is it Rape or is it Pernication? What drug do I have to be taking to make it work?

Mi Grain and Your Grain

When trying to get Study 329 published, we ran up against Liz Loder in the BMJ. She seemed completely resistant, stating on many occasions that we were biased – had conflicting interests. The article only got published after a conflict of interest duel. We wrote to the editor, Fiona Godlee, laying out Dr Loder’s possible conflicts. BMJ have never offered a satisfactory response to this letter.

The main conflict lay in Dr Loder’s husband working for a law firm, Ropes and Gray, who have worked on most of GSK’s recent difficulties including in relation to Study 329.

But in addition we pointed out that:

Dr Loder has made public statements favourable to GSK products, including: ‘For all of these reasons my mantra is that “You haven’t failed sumatriptan until you have failed to respond to a full dose of injectable sumatriptan given early in an attack!” There is also evidence that combining a triptan with an anti-inflammatory drug might improve the likelihood it will be effective’.


This article was published just before GSK’s Treximet (combination triptan and antiinflammatory) came on to the market – years after Dr Loder was supposed to have renounced all links to companies and their marketing.

Trick or Treat or Treximet?

Go Figure: Murder or Accident?

Harold Shipman

Harold Shipman was a doctor in Britain, who was arrested for murder in 1998. He turned out to be a true Angel of Death, the most prolific known serial killer, who killed it is thought between two and three hundred of his patients by prescribing opioids in large doses.

After his trial and conviction and jailing, he committed suicide in jail with no-one any the wiser as to why he had behaved the way he did.

His killing spree led to a change in medicine’s regulatory apparatus, ostensibly to ensure this couldn’t happen again. Boxes were put in place, and mandatory courses on a range of issues from consent to continuing education. All of this takes time away from seeing people. But if Shipman’s case held any lesson it is that he was assiduous at ticking the boxes that registration bodies like the General Medical Council in Britain put in place. The system probably makes a future Shipman more rather than less likely.

Short of doctors who have been struck off for proven cases of negligent care or abuse of patients, we might all on average be safer with doctors the regulator is having problems with, who for the most part are more keen on seeing patients than spending time ticking boxes, rather than with the doctors who are in good standing with the regulator. But what can the system do? We, the public, won’t let it do nothing.


While Shipman’s killing spree with opiods was unfolding, North America was sinking into a prescribed opioid epidemic that now accounts for 100 deaths per day, over 30,000 per year, over half a million since the epidemic began, perhaps the single greatest cause of death in America today.

The way in to the epidemic was laid during the 1980s with the marketing of Oxycontin by Purdue Pharma, supported by Abbott, and later Janssen pushing Fentanyl. There was an astute marketing of an idea – that people with real pain do not become addicted to opioids – a myth equivalent to the myth of lowered serotonin in depression. This was allied to new standards of care for pain management which hinged on RCTs, all of which demonstrated that opioids were effective – again exactly the same dynamic exploited by companies marketing SSRIs. The result was mass prescription of opioids by doctors, many whom felt trapped between clinical wisdom and the risk of being sued. But clinical wisdom never makes it into guidelines or standards of care, and since standards of care emerged in the 1980s managers sack doctors who don’t stick to the guidelines – or refer them to their registration body.

Prescription-only status is another part of the regulatory apparatus. It was introduced in America in 1914 in response to escalating concerns about one of the first opioids – heroin.

Purdue’s marketing of Oxycontin exploited this deftly. Companies and doctors were in the clear provided all prescriptions were for physical conditions causing pain. Patients and doctors on cue, to a background tune being piped by Purdue and Janssen, engaged in a dance about the management of pain. There were no problems for doctors prescribing for pain but if they had been prescribing the same drugs for addiction they would have been struck off.

Faced with the horrific consequences of this dance, FDA intervened in extraordinary fashion two years ago. In 2014, a century after the introduction of prescription-only status, FDA made a potent opioid, naloxone, available over the counter in US States who were willing to endorse this option. Given intravenously or nasally, naloxone can save lives by reversing the effects of an overdose by other opioids.

To murder is human. To mass murder needs a regulatory apparatus. We have no problem viewing Shipman as a murderer. But what about Purdue or Janssen? And what about the role of regulators from FDA to the GMC?

Go Figure: Study 329


Editorial Note: This post merges the Go Figure sequence of posts from several weeks ago with the 329 series.

In the wake of the French Revolution of 1968, the government was still tottering when on February 4 1970, sixteen miners died with twelve others maimed in an explosion at a mine in Lens in France. The mine owners pitched the event as an accident. It didn’t help their case that Northern France had a history mine disasters, including the worst ever in Europe in 1906 at almost exactly the same spot, when over a thousand men lost their life. This is the reason we have insurance, the owners said. This is what the law of 1898 was all about.

The miners called on the Left-Wing philosopher Jean-Paul Sartre to prosecute their case. Sartre claimed there were no accidents, the owners knew there would be a certain number of deaths each year and factored this into their costs. There had been five the year before in a nearby mine. While injured workers thought of their job as risky and that when injured they had just been unlucky, they were wrong. It was murder, Sartre said.


Calling these deaths murder made the disaster an event. If it was murder, the owners should go on trial. Capitalists in general should go on trial. Perhaps even the government. This was a pivotal moment for the Left. Despite the lack of trust of many in France in their government, society didn’t follow Sartre down this route. Perhaps because the French State made a display of solidarity. The Left lost its grip to a New Left and others who were looking for a Third Way.


A century earlier faced with a predictable number of deaths from anesthesia each year, American surgeons in the first place, because anesthesia was developed there, faced questions about the ethics of doing evil in order to do good. If the patient died from the anesthetic, were they murderers?

We somehow collectively decided that no they weren’t. No-one sees us as duped for doing so, even though roughly one per thousand anesthetic events today lead to death. There are likely more deaths from anesthesia than from mining disasters each year. There are certainly more deaths per person year exposure.


The death of the comedian, Joan Rivers, some years ago under light anesthesia for an endoscopy, rather than plastic surgery, threw a light on this issue. Deaths under anesthesia have been happening for one hundred and seventy years but every so often a death or other happening becomes an Event. In this case, nothing happened.

“With plastic surgery, the general anesthetic is like a black-velvety sleep, and that’s what death is – without waking up to someone clapping and going, ‘Joan, wake up, it’s all over and you’re looking pretty’.

What is less forgivable perhaps is that anesthetists have a blind spot for the difficulties that may emerge days or weeks later linked to their procedures and agents.

String of Pills

Where do the adolescents or young adults or any of us who commit suicide or homicide triggered by an SSRI taken, or given, on the back of studies like 329 fit? What do we call this?

British papers over the weekend were full of a trial involving a footballer and an inebriated woman where he was convicted of rape in the first instance, with the sentence later quashed in controversial fashion. American and world media were filled with stories of Donald Trump and the question of consent. In the light of Study 329, is the consent people or their families have given to take a medication like paroxetine any more valid than a consent that after the event an inebriated woman is claimed to have given?

Like the mining accident in Lens, the Evans rape trials, the Trump candidacy, the republication of Study 329 is an event, possibly a pivotal moment. Details outlining trends in the rate at which articles are ghostwritten, or patients enter hospital because of treatment induced injuries are features of the contemporary landscape but every so often these features come together to make an Event and Events make history. They create a before and after – even if not necessarily right there and then.

Study 329 Taper Phase


Editorial Note: It was tempting not to run a post today for fear it might get lost in the wash of the Clinton-Trump debate. But today is the fourteenth anniversary of the day FDA issued an approvable letter for Paxil for children, as well as the fifty-fourth anniversary of the 1962 FDA Act that created the playing field on which Study 329 happened. It’s also World Mental Health Day. Freud might have been amused.

Study 329 recruited its last patient in 1997. The data was analyzed in 1998. Faced with the results, the plan was to pick out the good bits and publish them. This meant never publishing the Continuation Phase – the 24 week extension to the Acute Phase of Study 329. The good bits of the Acute Phase became the Keller et al 2001 paper – possibly now the most famous clinical trial publication of all time.

The Continuation Phase was published for the first time by Le Noury and colleagues in September 2016 – eighteen years later. It is available HERE along with reviews from JAACAP – not the journal in which it was published.

Among the most interesting findings are those from the Taper Phase. Publishing the Acute Phase only meant that the Taper Phase vanished. Whatever way the data are cut, whether in terms of all severe adverse events, or behavioral adverse events or suicidal adverse events the picture remains the same – the taper phase of this trial was the riskiest period for those on active treatment. This will come as no surprise to many who have learnt the hard way after the event. Others like Bruce Springsteen do not yet seem to have put two and two together – see Born to Withdraw.




The years 1997 and 1998 may be as important as the data. The Moneybags image above, an internal SmithKline Beecham image from 1997, brings out how the issue was viewed within the company. SmithKline were at the time under intense pressure from Lilly and Prozac, with Lilly running adverts for Prozac like the one below.

Prozac withdrawal

The material accompanying adverts like this specifically targeted Paxil – Seroxat, the drug that was Prozac’s most direct competition. This is an unusual thing for a pharmaceutical company to do. They have to be on solid ground. It was also deeply cynical in that Prozac also causes dependence and withdrawal but with the right study its long half life can appear to make this problem disappear.

SmithKline’s response was maybe even more cynical. For women worried about getting pregnant and possible risks to their unborn baby, the short half life of paroxetine and being able to get off it quickly was a blessing. If you were on Prozac, the implicit message was you were “fucked”.

Extraordinarily in the face of this – twenty years ago – most doctors and regulators affected not to notice anything. Were they being adults trying to ignore children squabbling in the back of the car, or were they like terrified kids in Jurassic Park trying not to draw the attention of the Tyrannosaurs sniffing the air around them?

Most doctors still affect surprise at the idea SSRIs might come with withdrawal problems. Regulators since 2002 know very clearly about the problems but have decided to leave any communication of these issues in company hands.

It needs an anthem from someone like Bruce Springsteen to find the millions of people whom “Progress” has left washed up on a shore they never wanted to be on. An anthem to put the island on which they have been Shipwrecked on the map.

Study 329 Continuation Phase

Continuation efficacy

Editorial: We interrupt the Go Figure series of posts to return for two posts to Study 329. We will then return to Go Figure.

All the fuss about Study 329 centers on its 8 week acute phase. But this study had a 24 week Continuation Phase that has never been published. Until Now.

We might have Marty Keller to thank for this Continuation Phase. His big deal was the long term treatment of depression. He made his name on the back of claims that depression was more chronic than people thought, and that treatment might have to carry on much longer than had been thought. Or perhaps industry made his name for him as this was such an appealing idea for them. He rapidly became one of the go-to key opinion leaders for antidepressants, along with Charlie Nemeroff in the US and Stuart Montgomery in Europe.

Rhode Island, and Brown University, must have looked Providential to Pharma in the early 1990s. Peter Kramer of Listening to Prozac fame was there too, also advocating pretty well permanent treatment for people who might be somewhat less than entirely chipper 100% of the time. Well would you take off a pair of spectacles if your sight was crisper wearing them? Why treat a drug any differently? We need to get over these Calvinist hang-ups about feeling better on medication, maybe even feeling better than well.

The continuation phase

The Acute Phase of Study 329 ran for 8 weeks – slightly longer than usual because the worry even before it began was that it was going to be difficult to show that treatment worked. But then after that the children could enter a further 24 week continuation phase. The idea was to see how ongoing treatment shaped up. The continuation phase was never published. Never even talked about.

To coincide with the Autumn Equinox, the International Journal of Risk and Safety in Medicine has published Study 329: The Continuation Phase.

The full text is available on Study, along with the reviews from the Journal of the American Association for Child and Adolescent Psychiatry (JAACAP). Click HERE.

JAACAP is not IJRSM – what’s up?

The original Keller paper was published in JAACAP. The former editor Mina Dulcan and the current editor Andres Martin and the American Association for Child and Adolescent Psychiatry were lobbied heavily over the years by Leemon McHenry and Jon Jureidini, trying to get them to do the decent thing and retract the paper. In vain.

So it seemed like a good idea to send Study 329 The Continuation Phase to JAACAP first. We were pretty certain they would refuse it. But they needed to be given the opportunity to dig a deeper hole. And they did. The reviews are published on Study


The primary theme of the reviews is there is too much data here, tell us what it means. Interpret it for us, and then we can tell you whether we agree with you or not. The paper may in fact have too many tables and figures but the key message behind Study 329 is that there is no authoritative interpretation. A lot hinges on almost arbitrary categorizations of the data. You the reader need to be able to play around with the data; you may well spot things others have missed. When it comes to adverse event data in particular there is no expertise. Things need to be picked over.

Off a cliff

There are some things about the Continuation Phase that are almost too shady to pick over too much, and we did not spend time on them. At the end of the Acute phase, for instance, there was an option for patients to continue into the next phase. It’s easy to understand why people doing poorly in the trial might have opted to drop out at that point. Some people doing well can also have been expected to drop out. What is harder to explain is why most of those doing well who dropped out were on placebo. A mixture of reasons were offered such as non-availability of trial medication but two thirds of those dropping out while doing well were taking placebo. This does not seem random. From GSK’s point of view having a large number of people doing well on placebo enter the continuation phase would not have been ideal.

The findings

Despite the placebo dropouts, there are more data to pick over in this study that for any other study. The data are not flattering for paroxetine – Paxil / Seroxat.


The relapse rate on paroxetine was higher than for imipramine or placebo. This relapse rate makes paroxetine a Gateway drug. People taking it end up worse off and their difficulties are interpreted as further evidence of illness, and in the real world this is likely to lead to additional diagnoses and additional treatments.

As the Efficacy data above shows, there is no evidence that patients given paroxetine show any benefits compared to placebo over the longer run – even when the deck is stacked against placebo by eliminating many of those who are taking it and doing well.

The suicide data also continued to be a problem into the Continuation phase as the graph below shows.


The word AG1TA here stands for agitation.

But there was another even more interesting finding in this paper which will feature in next weeks post. If you download the paper you will likely spot it. Surprisingly in the light of the finding, or unsurprisingly given the finding, the reviewers paid no heed it.

To be continued…

Go Figure: Sally’s Problem with Whinging about Medicines

99 percent

Two weeks ago in response to the last post in the Study 329 series, Sally MacGregor added the comment – that features as a post below. It’s spot on. The problem is how to avoid being marginalized, becoming part of a 1%. How to capture the attention of the 99% for whom the meds work just fine thanks. There will be more on this theme over the next few posts.

The whole point about the revisited Study 329 for me was that it was so meticulously and scrupulously carried out by a team of researchers, in a way that left no wriggle-room for it to be dismissed as ‘bad science’. (Even though some have tried). So it stands as a solid, irrefutable, excellent piece of science, which will be cited in the literature, and, crucially, is likely to pop up on Google. It’s kind of embedded now, which seems to me to be its lasting legacy. That it concerned GSK and paroxetine is secondary – Eli Lilly were just as murderously and indifferently mendacious – and in the UK get much, much less publicity for their callously commercial behavior.

The difficulties in getting through the hoops to the BMJ must have been incredibly frustrating, but – is anyone outside the ‘Study 329’ inner circle, or people who suffered from paroxetine, truly interested in the fact that someone’s husband was related to someone else who might have hindered the path to publication. It GOT published – which is the really important thing.

No one I know, service users included, had even heard of Study 329 – although several friends were very interested in the RIAT paper – mainly as a concrete example of meretricious Big Pharma mendacity. Hell – I’d never heard of Study 329 until 2013.

I’m playing devil’s advocate to a degree but sometimes I think we have to get real about life outside, where the chemical imbalance theory still reigns supreme, and if the public perceive any problem at all with antidepressants it is far more likely to be along the lines of ‘well he/she wasn’t taking their medication, that’s why they went berserk’.

No one will like me saying this – but I am sometimes reluctant to point people in the direction of David’s blog and Rxisk because, coming at it from a newcomer’s perspective, people don’t understand why AllTrials is suspect, (no one ever slows down and explains, for one reason) what on earth the BMJ did wrong, why Fiona Godlee is Nurse Ratchett, or what on earth is the point of laboriously transcribing an exchange between Goldacre just to show he’s an unreliable twat… And those new visitors will probably never repeat the experience.

Drawing new people in, from places other than those (like me) who’ve been dreadfully harmed is really important: doctors, researchers, scientists, our children, relatives, neighbours, philosophers, writers, journalists. We honestly HAVE to broaden the debate otherwise it simply becomes an incestuous gathering of victims and activists, who already know all there is to know about the damage. If visitors from outside don’t feel welcome, or just don’t understand what everyone’s going on about, then what on earth is the point?

Believe me, I’ve tried getting people to engage with Rxisk and David’s blog but it hasn’t worked – for all the reasons I’ve just stated. But they are surely the truly important people to get on board? Otherwise both forums just end up with ‘I am a victim’ hand-wringing on a big scale which does nothing to spread the message…no one apart from Johanna offered me any help with my ‘Take a Rxisk report to your GP’ request – and that was a challenge thrown out by David, which I didn’t especially want to do, but was willing to give it a go.

Similarly, with the Complex Withdrawal site (which I’m deeply interested in, as it might just offer some hope in the future for people like me) – I asked around, as requested, got some small bits of information from my hairdresser and a friend who’s a beautician– but there is no where sensible to put it, and the comments section has (yet again) been mainly co-opted by people offering well meaning but probably useless dietary advice or repeating, yet again, their stories. Surely the challenge was to GET MORE AND DIFFERENT people involved in collecting information?

It seems to me that every single attempt to move forward just dies – because no one slows down long enough to consolidate the practical ideas which might just make a difference to future victims. It’s too late for me, but I’d like to do my best to see that help is there for all those still to come. We can carry on preaching to the converted till the cows come home, but seems to me far more important to educate, persuade, chip away at societal views about antidepressants (non addictive, no such thing as withdrawal etc etc) without making people who take them feel alienated and ashamed.

I’m fed up with bashing away and for anyone who wants to reply ‘I did not like this comment’ – go figure.

Club 329: Part 4


Editorial Note: This post perhaps should be called: There’s Something about Leonie. The image above is of a Rapid Response she submitted to a BMJ editorial by Richard Smith and Fiona Godlee that BMJ published and unpublished and republished and re-unpublished. The full story is here. It again hinges around Study 329.

The full transcript of her exchanges with Ben G is below. It’s important. It suggests AllTrials see the side effects of drugs as irrelevant.


Hi, this is to do with, specifically to do with Study 329. The BMJ took a year to publish the Restoring Study 329, the reanalysis of Study 329. I was just wondering what you thought of the fact that, would it be a factor that the BMJ clinical editor is married to a partner in Ropes and Gray, the same law-firm that GSK paid to defend them in the US department of justice action?

Ben Goldacre

I don’t know – also I don’t care. Honestly, conflict of interest, it happens – and it is clearly problematic. I don’t find it interesting, in the same way as I don’t find fraud interesting. So, I’m not saying that it’s not important.  I’m just saying, it’s not my thing. Conflict of interest is also what people who don’t understand trial design talk about, and it’s kind-of all that we talk about. So, all of the, almost all of the popular discourse around problems in research in medicine that you see mentioned it’s almost all about financial conflict of interests and I just feel like that’s really, that’s all we talk about, it’s well covered by the .. but it’s not, it’s doesn’t set me on fire.

I don’t think you can honestly say the BMJ are the bad guys in the world of suppressing medical research, and if anything, they’ve got a reputation of being, a kind of sanctimonious obsessives about research integrity.

And when you say it’s been a year to publish the reanalysis of Study 329, But firstly, I don’t think we needed a reanalysis of Study 329, honestly – it was a rhetorical act. It was interesting that someone went out and did it. But we knew that within six months of the trial being published, that the trial was crap. We knew within six months of the trial being published that it had been misreported and all of the stuff that I had showed you on that slide that Study 329 misreported it’s from a paper by Jureidini that came out almost as soon as Study 329 did. So we already knew, we already knew Study 329 was crap and study 329 was published 20 years ago.

I’m glad that someone did a reanalysis of the underlying data, to show, yet again, that it was dodgy and I’m glad that they got some media coverage, for showing, yet again, that Study 329 was dodgy, and showing that people switch outcomes. But actually, I’m more interested in the fact that there has now been 29 cohort studies, showing that on average about a third to a half of all trials switch main outcomes. It’s not about Study 329, it’s about these endemic structural problems throughout the whole of healthcare that we can fix. And I don’t know what fix you’ve got in mind, cos the fix that I’ve got in mind in the misreporting of outcomes, is, we need to hold journals to account when they misreport outcomes, and we need to stop then doing that bad stuff.  But I don’t know what you do about, I don’t know who ?? person is.. but I don’t know what we should do about that?


But it’s still unretracted?

Ben Goldacre

Study 329? You think the BMJ should retract Study 329?


I think it should be done by somebody.

Ben Goldacre

But the BMJ didn’t publish Study 329.


No, but, GSK then.  Whoever, whoever wrote it.

Ben Goldacre

Yeah, I mean I think it should probably be retracted. BUT, again, what’s the purpose of a retraction, because anybody switched on knows that Study 329 was crap – and, also are we going to retract the tens, possibly hundreds of thousands of clinical trials that also switched their main outcomes?  I mean, maybe we should, but that’s a really big piece of work and why would we, why would we be more interested in Study 329, than the tens or hundreds of thousands of trials which we know switched outcome?  I mean this is a systemic structural problem. I mean Study 329 has rhetorical value for getting the media interested and engaged. But why are we trying to retract that one trial, rather than a hundred thousand trials that also switched outcome?

Unknown Male

I just want to ask a similar question, about Study 329 as well. Is outcome switching not basically fraud then? Is it not fraud, switching?

Ben Goldacre



Would you call it fraud?

Ben Goldacre

So, it’s really interesting isn’t it.  I think, it’s very, very interesting how the lines have been drawn between fabricating your data, actually going into the spreadsheet in excel and deleting the number that’s there, for the, for the patient’s blood-pressure, and typing in a new one. That’s fraud.

But, all of these different design shortcomings, which we know, are associated with over exaggerating treatment benefits and downplaying side-effects. That’s not regarded as fraud. I think that’s really, really interesting and it’s not a, it’s not a position that I’m sure, that I think I can respect. Because what I think is when you get the wrong answer you hurt people, and it’s your job to get the right answer.


In the case of 329, wasn’t there a lot of teenagers damaged? Young people were prescribed the drug on the basis of it.

Ben Goldacre

Yeah, but, like, why are you talking about teenagers and Paroxetine? We’re talking about the whole of medicine. Like, so what’s special about 329? What’s special about ..


Is it not the canary in the coal mine, no?

Ben Goldacre

Is it what?


Like, it’s the most publicised example of a corrupt study?

Ben Goldacre

Emm, yeah. It is this year, yeah.


Sorry. Why is it not important?

Ben Goldacre

Why is study.. No,  Study 329 is important but I don’t understand why you’re so interested in Study 329, when we’re talking about structural problems throughout the whole of health.


Well it’s just one of your examples, that you had in your slides.

Ben Goldacre

Yeah, I’m happy – I’m totally fine with that but I think, I think it’s a real strategic error and a backward step, to be preoccupied with one study, when you’ve identified structural problems throughout the whole information architecture of evidence based medicine. When you’ve identified a problem that hits, like a third of all trials, then I’m not sure that I care about ..


Can I just ask another question, because I find it this stuff interesting. You’re talking about antipsychotics and antidepressants, and you know, the dodginess of some of the trials, but, do you not think the medicalizing human distress is also the bigger problem?

Ben Goldacre

That’s about informed choice. So I get a sense of, that maybe you’re coming from a particular standpoint in mental health and that’s a partial view but I think with, with any treatment the right thing to do is to say, look, for the problems that you have, there is very good evidence, that this is a medicine which exists, which is been invented, which is available on prescription. From the best evidence we have, overall it looks like it has the following benefits and the following harms and it’s for you to choose whether you’d like to take it.

Now that works really well for most treatments, except it’s not done properly. So first of all you need to get better at disseminating information, not just to doctors but to patients, as I said… That would be my funding priority for a whole year to get better shared decision making between doctors and patients and then you can say, well look, with this statin, you get the following objective benefit with the risk of the following side-effects, you make your own choice.

And the thing that we know, from all the research on decision making, is that different people make different choices. So some people when they’re offered Statins say “ ..I don’t want to be medicalized… Even if the does overall reduce my chances of heart-attack, stroke and death, by two percent over the next ten years, I’m not interested”. But, some people will say “Yeah, I definitely want that drug, are you insane? Why wouldn’t I take that drug – I want to live?” So different people make different choices on the basis of the same information and I don’t know if it’s for me, to bring my own personal views and prejudices, which are actually, between you and I, and the room, more aligned to yours, than to most of the psychiatric profession, but I don’t think it’s for me to bring my own prejudices to bear on that. I think it’s for me to help my patients make an informed choice on the basis of the best currently available evidence and if they want to take an antidepressant, where the best currently available evidence shows that it will reduce your Beck Depression Inventory score by 2 out of 30 points, which is a modest, but nonetheless true benefit – then, that’s a choice for them to make, it’s not for me to bring my prejudices about medicalization of society to bear on that. That, that’s a choice for them.

Lady asks a question (edited out)

Ben Goldacre

But I really want to know what you think about?


I just think when you’re saying about informed choice, a lot of people were prescribed antidepressants 15 years ago, maybe 20 years ago in the 90’s or whatever – they didn’t have informed choice, the side effects only came out when the general population of millions of people were prescribed them. So you’re talking informed choice now, but there wasn’t then… so.. so


Yeah, I agree, it’s really shit isn’t it

That’s what we want to fix with proper trials and proper side effect monitoring and all of that but that’s very different than the broader, sort of cultural view you seem to have of medicalizing human distress.


I think the pharmaceutical industry have medicalized human distress, yeah.


There’s no doubt that people have marketed pharmaceutical products for medical problems for medical treatment. But nonetheless if this treatment will knock two points off your..


But you’re talking about scales and stuff that’s not applicable to people’s subjective experiences of these emotional problems.


Well it’s not perfect but it’s the best that there is and an informed patient will make an informed choice, they will look at their depression metric and they’ll know what those questions are and they’ll look at at and see well we can knock two points off that- and they can look at that and go well what would that actually mean? Yeah you’re right actually that could just mean just not waking up in the morning two days a week..


It’s all subjective…


So people can make an informed choice, and if they don’t make an informed choice, well that’s their decision too, and people throughout society on all treatments vary hugely on whether they want to be involved in shared decision making, and that’s their right…


It doesn’t happen in reality though. Well certainly not in Ireland- maybe in the UK, but in Ireland it’s not happening.


Well that’s what we want to happen. But also you have to be really careful about imposing your prejudices on other people.


I’m not imposing my prejudices on other people.

Club 329: Part 3

Shane Clancy

Editorial Note: This post by Leonie Fennel carries on from parts 1 and 2 in this series. There will be one more post.

I dreamt I met my son Shane last night – in a jewelry shop, of all places. I was admiring the beautiful costume jewelry, when I overturned the dainty display and went clambering to pick up all the pieces. It seems I don’t escape my klutziness in the land of my dreams (or my love of all things bling).

The shop doorbell tinkled and in walked Shane who had been gone for so, so long, as handsome and animated as ever. All thoughts of scattered jewels were instantly forgotten while I launched myself at him. Shane’s younger brothers and sister appeared behind me (it is a dream after all) and he laughed happily while telling them stories of what he’d been up to. They all looked on, transfixed, fascinated as always by their big brother – listening attentively for once. All the while my arms were wrapped around him, with my head buried in his chest, clinging on for dear life, crying happy tears and feeling a joy in my heart that somehow seemed so alien.

Then I woke up and cried all over again when I realized it just a silly, silly dream and my lovely Shane was still dead – it’s ‘Marbh’ in Irish but means the same, deceased, dead, unadulterated and irrevocably dead.

Seven years and I’m still haunted. Haunted by the ‘what ifs’. What if I hadn’t insisted that Shane see a doctor? What if the doctor had not believed in the biological model of treating heartache? What if we had insisted on knowing all the facts BEFORE he took an SSRI, not afterwards? What if we had known Study 329 was a crap trial before 2009, not afterwards? What if Shane’s arms around me were actually real, not just a dream? ENOUGH. Back to a modicum of normality. I don’t know why I shared that with you, apart from the fact that the feeling of Shane is burned into my mind today.


As this is my second post, I should point out that my previous one effectively opened a virtual Pandora’s box, by annoying Ben Goldacre. He said I misrepresented his talk in Dublin – which of course I didn’t. He also said that David Healy was ‘fully responsible’, presumably for publishing the post on his website – have to be careful of misrepresentation here but I’m a little offended at the suggestion that I’m not entirely responsible for myself. To be perfectly honest, I’m not quite sure what all the fuss was about – but it seemed to have had a Streisand effect, which I suppose is a good thing. I think I’ll leave Drs Healy and Goldacre to debate the finer details, although I’d love to see them debating in the same room. Wonder if I’d get an invite if I promised to take a responsible adult as a chaperone, although Dr Dishy has now officially disowned me!

Moving on

Tragedies similar to Shane’s are increasing every year, largely due to the increase in the prescribing of psychotropic drugs. Last week, very near to where I live, we heard police cars racing past with their sirens blaring and helicopters flying around overhead. It turned out that a man who lived nearby had tried to strangle his four children. Thankfully they all survived, but it was a close call, with two of the children being airlifted to hospital. Reports suggest that this man is a nice guy who loves his children dearly, so how do we marry these two opposing images? Every year we see the same tragedies, filicides, siblicides and infanticide, all with the same SSRI-induced hallmarks – yet very few ‘get it’.

When I hear of awful incidents like this latest one in Wicklow, the first thought that goes through my head is “Please, let these kids survive”. The second is “I wonder if this poor man was recently prescribed an SSRI or was he in withdrawal?” That most medics wrongly believe that psychiatric drugs are not addictive only exacerbates the problem. Hence, this latest incident is all too similar to many previous ones in Ireland, where the outcomes were not always as fortunate.

Despite the FDA and EMA warnings, on the rare occasions when medication is implicated, medical professionals will spout ‘autonomy’ and ‘informed consent’ – as if this is an actual possibility. Tell me, what medic tells a patient presenting with distress, that the proposed treatment may cause suicide, violence, emotional blunting, akathisia and among many other awful effects? Oh yeah, let us not forget the sexual dysfunction. Who will tell a vulnerable patient that the most dangerous time with psychiatric drugs is upon starting, changing dose (up or down) or discontinuation? Tragically, in the case of withdrawal, many healthcare professionals will justify these incidents by saying it’s the ‘mental illness’ returning – stating that If these people took their meds as prescribed, there wouldn’t be a problem. Thus, was this Wicklow native and his family given even the tiniest degree of informed consent? If this ‘out of the blue’ case turns out to be yet another instance of prescripticide, I can only hope that this man’s family, who know him better than anybody, will understand that these drugs can be the catalyst.

Lastly, the opinion of a lone member of the Irish Police force “There is only one thing I know that enables a parent to want to kill their children, SSRI antidepressants”. Whether this latest case is SSRI-induced, we will just have to wait and see. I’ll keep you posted.

Club 329: Part 2

Following last week’s post, Club 329: Part 1, Ben Goldacre went into orbit claiming his views on medicalization and Study 329 had been misrepresented. He offered a SoundCloud as evidence. The link can be found in the comments after the last post. Seems to me Leonie got the content of the Q & A right.

In the course of listening to the BG SoundCloud though something else came into view. Leonie began questioning him about the links between one of the BMJ editors and the law firm Ropes and Gray who had represented GSK in the Department of Justice case that led to a $3 Billion fine for GSK. This didn’t seem to worry Ben. He’s not worried about Conflict of Interest – nor am I. His view on Fraud is unclear.

There was something else in the recording that caught my interest. BG didn’t figure the question of BMJ taking over a year to publish Study 329 was an issue. You couldn’t possibly make out that BMJ are the bad guys in the business of suppressing good science, he said. If anything they have a reputation for being too sanctimoniously obsessive about research integrity.

As BG says people react differently to things. If you and he were both facing a course of treatment, some piece of trial data about a drug might speak to him and not you and vice-versa. Maybe the Study 329 story doesn’t speak to him and others of you out there – doesn’t give you a feel for GSK or BMJ.

Perhaps a completely different story will work for those of you to whom Restoring Study 329 doesn’t speak.

A curious development

Six months into a tortured year long review process for Restoring Study 329, something odd happened. On February 9 2015, I was approached by the BMJ:

I’ve long been interested in the controversy over the role of serotonin in mental illness and wonder if you would like to write an editorial for us on: “What is the evidence that serotonin plays a role in depression?”

I replied:

More than happy to think about doing this for you. But I probably need to make something of a pre-conflict of interest conflict of interest statement, which you will need to consider.

The idea of serotonin in depression is inextricably linked to the marketing of the SSRIs. There is vanishingly little evidence that serotonin is involved in depression – but I can probably put more evidence on the table for its role than almost anyone can.

The idea of a role for serotonin in depression has been an extraordinary marketing trope – one that is critical to perceptions when it comes to the way the role of SSRIs in suicidality and birth defects is viewed. Because of this latter aspect and my role in some of these debates many people viewing an editorial by me would probably have a blood boiling moment. That’s just viewing the existence of an editorial – might not be too bad if they read it.

BMJ decided to run with it, but then seemed to have a crisis. They got back to me with an extensive conflict of interest declaration and there was a considerable delay after I got this back to them before anything else happened. Given all the fuss BMJ were making about Conflicts of Interest on Restoring Study 329, it was difficult not to think there was some link.

Finally they ran with the editorial but had difficulties with the title. So, So Long and Thanks for all the Serotonin became Serotonin and Depression. The Marketing of a Myth. It did well in terms of impact factor despite the fact that Sense about Science, with which Ben Goldacre is closely linked and which kicked off AllTrials with him, mobilized to get dissing comments about it from the Royal College of Psychiatrists and others.

Sense about Science spends a good deal of time mobilizing responses to material that might seem not supportive of corporate interests – see the Sense about Science series of posts to which BG also took exception.

Curiouser and curiouser

But there is a more interesting story behind this one that I only know about by accident.

A short while before the BMJ approached me, I had been sent an article – on the Myth and Marketing of Serotonin. It was a very good read. At least as good as mine if not better. One of the authors asked what I thought about the possibility of BMJ being interested. I said I thought it was unlikely.

Despite my advice, the authors went ahead and submitted anyway – on February 7 2015. They got a quick reply.

We are in the process of commissioning an editorial for The BMJ, looking specifically at the role of serotonin in depression, and so you will be able to send in a rapid response to that directly.

When dealing with BMJ you can get the feeling they (the editors) are using authors to run their own agenda. You might get in touch with a draft article on Access to Clinical Trial Data and they get back saying “sorry we’ve just had something else from someone else on this topic”. The something else when it appears a good deal later turns out to be a neutered version of what needs to be said.

Many people figure journals shouldn’t have an agenda. In fact the first medical journal, the Lancet, began with an agenda – against food adulteration. The later appearing BMJ ran its own campaign against nostrums – around 1900. These campaigns put these journals at considerable risk of reprisals. BMJ sees itself under its current editor as on a mission. It’s more recent campaigns, far from putting the journal at risk, have been for Evidence Based Medicine and Against Overdiagnosis, along with AllTrials to which pharma has signed up.

Whatever the BMJ mission its not a mission to tackle adverse events – the modern equivalent of adulteration or nostrums. It was close to paralyzed in the case of Study 329 by having to conceded a drug might have a side effect.

The educational articles BMJ runs on most drugs contain little or no mention of adverse events. While it published Restoring Study 329 after a gun was put to its head, its educational articles on antidepressants play down any risk of suicide, are comfortable with giving antidepressants to children, never mention withdrawal, and deny links to birth defects.

But here’s the rub. The article that was better written than mine was written by Leonie Fennel and Maria Bradshaw – neither of whom have healthcare or neuroscience backgrounds. Both write like angels. Leonie was a hairdresser before a family tragedy mobilized her to find out more about the drugs that led to the death of her son and others. The latest version can be accessed on ResearchGate and an earlier version is attached here.

You’d have to think BMJ couldn’t cope with the idea of a hairdresser joining the club. Only credentialed nerds need apply.

Leonie’s input along with that of people like Anne Marie Kelly who have done so much to establish the role of SSRIs in promoting alcoholism proves again and again – and is the inspiration behind – that motivation counts for more than expertise.

When my article came out, Sense about Sense mobilized against it – having been sent a copy by BMJ, with whom they, BG and GSK are part of an AllTrials coalition.

Parts 3 and 4 to come.

Club 329: Part 1

Ben Goldacre

Editorial Note: This post is by Leonie Fennel. It’s one of two involving Leonie.

Last week, Dr Ben Goldacre gave a public lecture in the Royal College of Surgeons in Dublin (organised by the 3U Partnership and the very lovely Dr Ruth Davis). Dr Goldacre is a doctor, academic, campaigner and writer; he is also a psychiatrist and self-professed nerd. I was eager to hear what he had to say, not least as the subject-matter was ‘Bad Trials’ – so off I toddled to Stephen’s Green with a friend in tow, a psychotherapist. He, like me, has a personal interest – he has witnessed first-hand the devastation that can be caused by nothing more than a GP’s farraginous prescribing. He is also a very kind, funny, charming companion, so I was delighted to have any excuse to meet up with him. I’d also say he’s a handsome chap but the husband I abandoned for the day wouldn’t be too impressed.

Having read Goldacre’s ‘Bad Pharma’ book, we were both curious to hear what he had to say. Incidentally, he once called my English friend Fid a ‘Smeary Conspiracy Theorist’ – so apart from guaranteed entertainment, I wasn’t too sure what else to expect. In fact, his talk was fast-paced and as excitable as he is – he hops around like a Duracell bunny on speed and lets out intermittent roars, which effectively kick-starts the heartbeats of anyone not paying attention. Nevertheless, he attempts to make data and statistics fun, a nigh-on impossible task.

Needless to say, as the subject concerned ‘bad-trials’, he specifically mentioned GlaxoSmithKline’s notorious Study 329, although bizarrely managed to do so without mention of GSK (usually both are referenced synonymously). He seemed like an amicable chap and was quite happy to answer questions afterwards in the Q & A session.

I was interested to explore his views on Study 329 and asked his opinion on why the BMJ took a year (of much wrangling) to publish Le Noury et al’s reanalysis of it – and did he think it had anything to do with the BMJ’s clinical editor being married to a partner in Ropes and Gray, the same law-firm hired by GSK to defend the action brought by the US Dept. of Justice, where Study 329 was a central element.

Goldacre said that he didn’t know and didn’t care – that fraud and Conflicts of Interest were not of interest to him. He asked what the fixation with Study 329 was, as it was just one of the many trials where data was misrepresented?

He expressed the opinion that everyone knew from early on that the original study 329 was flawed and nobody really relied on it. I sincerely doubt that the authors would agree with him on that, but he is entitled to his opinion. The oddity as the photo shows he was talking about outcome switching which is what happened in this originally well-designed trial.

He went on to say that the reanalysis (Restoring Study 329: efficacy and harms of paroxetine and imipramine in treatment of major depression in adolescence) was nothing new. It only did what had largely been done before and confirmed what everyone actually already knew.

Expressing my concerns that Study 329 was still not retracted, he asked what I wanted to happen, that all papers that are re-analysed and found wanting, be retracted? Erm, yes!

My friend, the dishy therapist, then said that wasn’t the crucial point being that 329 harmed so many children? Ben said once people have informed consent, they can make their own choices – like this is a common practice. He pushed Ben on the now-common practice of problematizing distress, with Ben suggesting that people need to be careful not to push their own biases onto others.

It struck me as odd that BG seemed disturbed by the discussion turning to study 329 – yet he had specifically brought it up himself. I thought it even odder that he didn’t give the re-analysis by Le Noury et al any credit at all. I got the distinct impression that while his forte may be in data and stats, the enormous numbers harmed by these fraudulent trials were given little consideration.

How can anyone say that people can make an autonomous choice to take a drug, when the (usually ghost-written) studies are manipulated to give positive results, while hiding serious harms?

The mammoth undertaking by Le Noury et al deserves huge recognition for exposing just this – that truly informed consent is impossible unless the full facts are provided.

Editor’s Note: Ben Goldacre has taken exception to this characterization. He comments below. His tweets on the issue can be found BG tweets