As regular readers might well recall one of my recurring ‘things’ here is newspaper health correspondents (many of them are crap) and in particular health correspondents at the Independent (who are even more crap than most of the rest).
Now I have my ‘thing'(s) and the Indy has it’s ‘thing'(s), one of the most prominent of which being its preference for trying to distinguish itself from its main competitors by having a ‘campaigning’ front page rather than just straight news. So it was no great surprise to see the Indy leading the charge this week on the PLoS Medicine published study on past trials of a range of anti-depressant drugs.
They are among the biggest-selling drugs of all time, the “happiness pills” that supposedly lift the moods of those who suffer depression and are taken by millions of people in the UK every year.
But one of the largest studies of modern antidepressant drugs has found that they have no clinically significant effect. In other words, they don’t work.
Now this story nicely illustrates one of the problem I have with health correspondents – given half a chance of a decent headline they’re all to prone to editorialise an interesting story to death to the point where it ceases to contain anything that remote resembles either accurate or meaningful content. Fortunately on such occasion one can generally turn to the always reliable Ben Goldacre for an accurate reading of what the real story actually is and this is happily one such occasion:
It has been misreported as a definitive nail in the coffin: this is not true. It was a restricted analysis [see below] but, more importantly, on the question of antidepressants, it added very little. We already knew that SSRIs give only a modest benefit in mild and moderate depression and, indeed, for some time now, the NICE guidelines themselves have actively advised against using them in milder cases since 2004.
Two days, and two more heavily editorialised articles later, the Indy managed to provide a first class example of the other problem I have with health journalism. When they not diving over the barricades and editorialising stories to death, they’re just uncritically publishing whatever the latest bit of PR screed that been lobbed in front of them by Big Pharma or, failing that, cobble something together out of old material to try and reheat and old story, the latter being how we arrive at this story from Thursday of this week:
Anti-smoking drug linked to suicides given NHS approval
The anti-smoking drug Champix should be offered to smokers who want to kick the habit despite fears that it is linked to suicides, according to the Government’s health watchdog.
The National Institute for Health and Clinical Excellence (Nice) has recommended that the drug, which tests indicate can double the chances of a smoker quitting, be prescribed to those wanting to give up the habit.
Now this is old news. NICE actually approved the use of Champix within the NHS, which has the generic name varenicline, when it issued its draft guidance/appraisal for its use back in May 2007, at which point its press release stated that final guidance was expected to be issued during the course of the following July. In practice the way NICE operates that the notional period between the issuing of draft and final appraisals is there to allow for any appeals against the content of the draft guidance and from the look of thing there were no appeals lodges regarding this particular drug which meant that the guidance issued in May 2007 can be taken as final.
What appears to brought this old story to the attention is both that its apparently been ‘linked’ to a number of suicides and that it get a mention in some newly issued general guidance, by NICE, on the delivery of NHS smoking cessation services, and what NICE actually has to say on this occasion is simply that:
NRT, varenicline or bupropion are recommended as treatment options for people who are planning to stop smoking on or before a particular date (target stop date) and alongside proven behavioural counselling and group therapy
Varenicline or bupropion should not be offered to young people under 18 or to pregnant or breastfeeding women.
All of which would rather mundane and routine were it not for this suicide ‘thing’ that’s suddenly cropped up on the Indy’s radar, so what’s that all about…?
The European Medicines Agency has ordered that the drugs come with a safety warning after Champix was linked to 37 suicides in America and those of two men in the UK. Omer Jama, 39, was found dead at his home in Bolton last October. He slashed his wrists weeks after starting a course of Champix. And Wayne Marshall, 36, a father-of-two, was found hanged in January, shortly after completing a 13-week course of the drug. His widow said she believed the drug had played a part in his death and reported the case to the Medicines and Healthcare products Regulatory Agency (MHRA).
As well as these cases there have been two other attempted suicides and 60 suicidal-type adverse reactions since the drug’s introduction. The MHRA, which licenses and monitors drugs in this country, says it will be closely monitoring the use of Champix.
Got that? Right… so now let’s get down to the science bit, which I’ll try to keep a straightforward as possible even if it means resorting to a bit Dr Who-ish blather about ‘wibbley-wobbly, timey-wimey stuff”.
Champix is supposed to both reduce the cravings and symptoms of nicotine withdrawal that smokers go through while trying to give up and make smoking, itself, much less satisfying and it does this by ‘binding to nicotine receptors in the brain’. In other words it mucks about with a bit of the biochemistry of the brain to try and stop you feeling like you’re gasping for a fag.
Now, the thing about the brain is that our current understanding of how it works is a bit… well, rough and ready and then there’s the mind-body problem to take into account, the upshot of which being that its nigh on impossible to predict exactly what effect all this mucking about with the chemistry of the brain will have, if any at al, on the thought-processes of the people who’s brain chemistry is being tinkered with. In short there’s always the possibility of unpredictable side-effects and its bloody difficult to figure out in advance just exactly, if anyone, might end up with them.
On top of that, should someone who’s taking or has recently taken a drug that messes around with the chemistry of the brain, like Champix, happen to go and top themselves then its only natural that the grieving relatives of the now terminally ex-smoker are going to go looking for something or someone to blame, and the drug they taken is a pretty obvious place to start, especially if its only come on the market fairly recently. So all in all, its not that uncommon to see this lind of ‘new drug linked to suicide’ story cropping up in the press…
…although coming only a couple of days after the Indy has tried to monster Big Pharma over withholding a few of their less impressive trials of anitdepressants what is a little unexpected is the rather conciliatory and unquestioning tone of the rest of the Indy’s coverage of Champix.
For example, whenever you’ve got a halfway decent looking side-effects story on the go then one of the obvious moves is to give the company who produce and market the drug a ring to see what they might have to say for themselves, which in this case means Pfizer, the producers of viagra and the world’s number one friend to men who struggle with getting a stiffy.
A spokesman for Pfizer, the drug’s manufacturer, welcomed Nice’s decision and said that although symptoms of depression had been reported in some patients taking Champix, no causal link had so far been established.
And its highly unlikely that any ‘causal link’ will ever be established because even if the drug is causing some people to shuffle off this mortal coil under their own steam actually proving that its because this or any other drug is messing with their head can be the next best thing to impossible.
In situations like this direct proof of causality is not going to happen – it’s the mind-body problem again, we just don’t know for sure exactly how biochemical and bioelectrical activity in the brain maps onto actual conscious thoughts. There are times when we can make a damn good guess and be sure we’re logically spot-on – if you take a smoker whose not had a fag for a few hours then its perfectly logical to assume that the biochemical effects of nicotine withdrawal will produce thoughts like ‘I’m gasping for a fag’ and ‘where the fuck did I put that pack of Benson and Hedges?’, not because we can actually relate those specific thoughts to what’s happening at a biochemical level in the brain but because experience and logic tell us that such connections are too stupidly obvious to be anything other than correct. Getting from A to B here is not only a straight road, its a one-way street with dirty great walls on either to make sure you don’t drift of the obvious path.
Getting from A – this drug does something to my brain – to B – oooh, look there’s an express train coming, I feel like jumping in front of it – isn’t anything like such an obvious connection to make, which is why we can’t make a direct causal link.
In these situations the best we generally do is look for a statistical correlation that looks way too suspicious to be anything other than causal or, failing that, one big enough for us to figure that its not worth taking the risk of dismissing the possibilty that drug X is actually fucking with people’s heads so badly that some of them start reaching for the multi-packs of paracetamol. So what you do is pull together a few numbers for the suicide rate in the general population, compare them to the rate in patients taking drug X, do a bit of statistical jiggery-pokery to match the numbers up and hopefully cancel out any statistical kinks you might get from differences in the two population groups, and then see if the number that get spat out the other end look a bit too suspicious for comfort or not. All of which, when you’re dealing with a new drug on the market can be a bit on the tricky side, because of the effects of population differences you need to try and cancel out are the distortions that can easily crop up when you’re trying to compare the incidence of something that generally fair rare in a pretty big population to the incidence of the same thing in a very small population.
Forget causality here, the best you’ll get is a correlation that close enough to look too damn suspicious for it to be written off to coincidence or bad luck, as I’m sure the talking head from Pfizer knows all too well – given the newness of Champix the very least that Pfizer can bet on is that it way too early to generate the kind of numbers that will make the regulators edgy enough to pull the drug off the market, giving them a window of opportunity in which they can push the drug for their worth in the hope recouping as much its development costs as possible before we find out, one way or another, whether it has been fucking people up, just in case they do get unlucky with this one.
The Indy’s too casual acceptance of a stock line from one of Big Pharma’s PRs is far from unusual, in fact its not something I’d been inclined to spend too much time commenting on were it not for a couple of other comments in the article, comments that are attributed to members of the NICE panel who have approved this drug for use:
“This is a relatively new drug and we don’t have the advantage of evidence as to its effects,” said Sir Alexander Macara, the chairman of the National Heart Forum, who chaired the Nice committee. “We are being properly cautious about recommending it to people who might be vulnerable to the drug.”
Now, Sir Alexander does bring to the table a pretty hefty clinical background, he’s a former Chairman of the British Medical Association for starters and, as the article notes he’s current the chairman of the National Heart Forum, which campaigns heavily on smoking cessation and tobacco control.
What seems a little less convincing, however, is his reference to NICE being ‘properly cautious about recommending it to people who might be vulnerable to the drug.’ given that ‘vulnerable’ in this case, as with so many others, seems to consist in the guidance only of ‘young people under 18 or… pregnant or breastfeeding women’ – all of which looks very much like the kind of standard boiler-plate warning you’ll easily find on just about any drug on the market that hasn’t designed specifically for use on kids.
Now I know that the Indy doesn’t give much in the way of detail about the suicides that have been allegedly linked to Champix (or Chantix as its branded in the US) but neither to the two suicides that have linked to the drug in the UK fall into the ‘no kids, pregnant women and new moms’ criteria – and nor would you expect them to be as the certification of drugs for use by under 18s is almost always handled separately from their certification for use by adult and is usually given only after the drug has been on the market long enough for there to be a body of data on adult use sufficient to indicate the the drug is pretty safe.
By way of contrast, following a number of cases in the US that have linked the drug not only to suicide but in one case to a high profile shooting:
A notable example, however, that drew national attention to Chantix was the case of Carter Albrecht. While in a drunken rage, the Texas musician banged on the door of a homeowner, who feared an intruder, and shot and killed the young man. Although Albrecht was later found to have tested for a high level of alcohol, he’d also begun taking Chantix, the smoking-cessation drug, a week earlier and complained of vivid dreams. Hallucinations are noted as only a rare psychiatric disorder, but “Chantix dreams” were cited by Albrecht’s family and girlfriend as possibly contributing to his outburst.
…packets of Chantix sold/prescribed in the US now carry the following warning, which has been added to its label by Pfizer.
“…patients who are attempting to quit smoking with Chantix should be observed for serious neuropsychiatric symptoms, including changes in behavior, agitation, depressed mood, suicidal ideation and suicidal behavior.”
Fair play to Pfizer for adding that warning to its US product, although one might have hoped that the Indy might have asked Pfizer whether the same warning is or will be included on the drug’s labelling in this country, just as they might have asked the members of the NICE committee who approved the drug for us NHS smoking cessation services, whether this same kind of warning might not have been usefully incorporated into its own guidance.
As for Carter Albrecht, whether or not the drug may, or may not, have played any part in his actions will almost certainly never be known…
The Dallas Medical Examiner, Jeff Barnard, wrote in his report that he sought to test the body for the drug, but was unable to do so. Why? “No test exists in our laboratory,” he said, according to a local TV station. He also stated that he was “unable to find any lab which does this testing beyond the manufacturer.” For its part, Pfizer declined to perform the test, and sent Barnard a sample and instructions for developing its own test. But that could take weeks or months, he claimed, and the sample expired after a few days.
Equally interesting are the comments of another member of the same NICE panel, Christine Owens – who not uncoincidentally happens to be the Head of Tobacco Control at the Roy Castle Lung Cancer Foundation – and, yes, the phrase ‘conflict of interest’ is occurring to me in regards to the reported composition of this committee.
Christine Owens, from the Roy Castle Lung Cancer Foundation, also sat on the committee. She said it was aware of the possible link to suicide but added: “It is a clinical judgement and [this drug] should not be given to anyone where there are any concerns or people who have suicidal thoughts. However, there is an argument over whether people have suicidal thoughts because they have given up smoking or because of drugs.”
Sorry?
Remember what I was saying earlier about the fact that even though we don’t understand how the biochemistry of the brain maps on to thought there some are some connections that are so blatantly obvious that we can take them as read?
Well having suicidal thoughts just because you’ve given up smoking just ain’t that kind of logical or obvious connection, not on its own. Sure, nicotine withdrawal as the icing on a pretty shitty cakeload of other problems might not be too helpful in the game of suicide avoidance but you’ve still got to be badly fucked up from the get go to come to the conclusion that topping yourself is a better option that nipping down the all-night garage for a pack of Silk Cut, and whether she realises it or not the impression that Owens comments create is one which hints that there would have to been an almost stone cold certainty that Champix causes people to commit suicide for this committee ot even think of insisting that Pfizer’s carries its US warnings on its British packaging let alone put of guidance recommending that careful assessments be made of candidates before prescribing the drug.
After all, what’s a few suicides compared to the long term social, medical and financial costs of smoking – as long it ain’t you and family that end up burying a loved one long before their time.
In its own way, the Indy’s supine acceptance of the ‘official line’ here is as bad a piece of journalism as its overplaying of the antidepressants story of a couple days earlier.
Between the information on Champix and its conjectured link to suicide that’s already out there in the public domain, especially from the US and all easily found using nothing more complicated that Google, and the rather glib dismissals they’ve received from both Pfizer and from members of the NICE committee responsible for drawing the guidance on the us of the drug by the NHS, there are plenty of searching questions that the Indy could have asked, even without going for the over-the-top approach they adopted in reporting the PLoS study, none of which they appear to have been bothered to ask.
Next time you see someone in the MSM bitching about bloggers and questioning what we do, just remember that one of the answers is increasing becoming ‘your fucking job!’
Top notch writing.
John Gibson
I should also comment that it’s prescribed as an anti-depressant in addition to a stop-smoking aid. It was originally designed as an AD, and it was only in the initial trials that they noticed a propensity for participants to give up smoking.
In light of the above article, I feel fairly flippant saying this, but obviously increased risk of suicide is also a ‘side effect’ (ha!) of other ADs.
Sure, nicotine withdrawal as the icing on a pretty shitty cakeload of other problems might not be too helpful in the game of suicide avoidance …
People can say many things, Unity but they can’t say you’re not well researched and argue a fine case. This was interesting reading in a topic I didn’t give a sh=t for before.