Psychobabble: Placebo Predicament
By Keir Liddle
This post may come across as something of a departure from my usual 21st Floor output, as it doesn’t rally against alternative medicine, the lack of evidence-based policy or the like. This post will likely be more personal and self reflective than others I have written for the site, but it does delve into a sometimes controversial area of medicine and psychiatry: the efficacy of anti-depressant medication.
Here comes the personal bit. There will be some stats and science later, I promise; feel free to skip a few paragraphs (three of them, to be precise) if you don’t want to know the context for my own placebo predicament. It’s not really necessary, and it’s probably not particularly interesting or edifying to read…
I was diagnosed near the end of October with Bipolar-2, a disorder classified by intense bouts of depression and hypomania (at least one bout thereof) which is bipolar’s (the DSM category formerly known as Manic Depression) less sexy relative, in that the mania is less pronounced and the depression more pronounced. In personal terms, it’s a bit of a shit deal. Without the mania– who knows?– we might never have had Skeptics on the Fringe and various other things, but that has to be tempered with missing things like TAM because I have in effect been suicidal, and likely missing QED because much of my money has been spent on various manic whims (none of which I might add are particularly exciting). Now, it’s easy to say you feel suicidal or depressed, and for people who haven’t experienced this, to shrug it off to an extent or otherwise fail to understand the severity of those statements. At least, that was true for me before I experienced it myself. So to put it in a bit of context, imagine you are assailed by a persistent desire not to be here, not to be anywhere, to simply no longer exist because it seems easier. You are relentlessly bombarded with “black thoughts”; you imagine your family and friends dying; you feel abandoned, as if everyone you know hates you or doesn’t care about you. You find yourself “testing” friends; putting them through challenges to prove that they like you– challenges that they can’t ever pass as they are mainly designed to confirm your own sense of self loathing and lack of self worth.
One experience that perhaps encapsulates all that is that every time I had to cross the road I had to fight some internal battle not to just close my eyes and walk out into traffic. If I had to sum up my depressive phases in one phrase the following would come close: depression – it’s like vanity for masochists.
I was prescribed Citalopram (starting on 10mg, soon upped to 20mg, currently on 30mg– this dosage may need to increase a little more, but we shall see) and it did nothing initially — although it did mobilize me enough to start self-harming in the initial few weeks. However, once the dosage was upped, my mood appeared to lift. The depression started to lift, I had a few manic periods, and in general things started to get better. This is pretty much where I am now, my moods are still swinging from high to low, but the darkness is a little lighter.
I don’t share this with you for your pity, platitudes or anything of that nature: it’s presented purely for context. I would hope no one would judge me too harshly for sharing the above, and I apologise to anyone who felt that it is outwith the purview of this blog. I would hope to regain your favour by delving into some statistics and science from this point on.
A paper on the efficacy of Citalopram came to my attention via Twitter: Are Regular Doses of Citalopram for Depression only Placebos? Meta-analysis And Meta-regression Analysis Of Pre-registration Clinical Trial Data
The paper aims to explore published and unpublished trials of the drug, and combine them using meta-statistical techniques to determine whether it is an effective treatment for depression. Meta-statistical techniques provide a means of combining data from different studies (so long as they fit rigorous selection and inclusion criteria); they are not without their issues, but are generally considered in medicine and other fields as a kind of gold standard of evidence. To my dismay as a skeptic, the above paper reaches the following conclusion:
There is insufficient evidence that 20 mg daily of citalopram is clinically useful in patients with mental depression, probably not at all effective in major depression. Surprisingly, this dose was not tested before it was recommended. Two fixed-dose studies with 20 mg daily have been performed after the registration of the drug in Denmark. Neither study showed a statistically significant difference between the drug and the placebo in absolute depression scale scores. Increasing the recommended doses of citalopram is warranted.
The paper’s analysis shows that at doses of 45mg and above, Citalopram is as effective as other anti-depressants but at doses under 45mg Citalopram performs worse than comparator drugs. The study implies that Citalopram may be effective but it raises questions personally for me as to whether or not it has had an actual effect on my condition. As a skeptic with a particular interest in alternative and complementary medicine, it would be ironic to discover that a drug I begrudgingly have to thank, to put it bluntly, for still being here might possibly be a placebo.
Can I in good conscience criticise the treatment of chronic conditions with ‘alt med’ woo? After all, if it turns out that the drug that has helped me so is nothing more than a placebo with side effects, how can I countenance against the notion that placebos can be useful in the treatment of some disorders? How do I make peace with the ethical dilemma associated with prescribing such placebos — holding such as I do that deceiving patients should be anathema?
There is some research that suggests placebos may work even if you explicitly tell patients that they are placebos. That may simply be down to the value of the medical ritual or some other psychological aspect of treatment that warrants further research. This might overcome my ethical objections, but really, to be consistent, wouldn’t that mean I’d have to grudgingly accept homeopathy and other such nonsense as effective placebos? I suppose that I could temper that with the notion that at least Citalopram does something at certain doses, whereas alternative medicine does nothing at any dose beyond placebo.
What is more worrying is the process which I went through when first faced with this paper: I searched the relevant databases for papers, and paid attention specifically to those that supported the idea that my meds were efficacious. I ran through all the problems and issues with the techniques of meta-analysis and meta-regression. For example, meta-analysis is not as good as mega-trials, and the inclusion criteria and search strategy have to be spot on. One issue with this study could be that they look at papers which compare Citalopram to placebo, and to comparator drugs, which perhaps indicates that the inclusion criteria isn’t that great.It is important to get this aspect right as one of the first meta-analyses, conducted on psychotherapy in psychology, led renowned psychologist Eysneck calling it “mega-silliness” for showing that it worked. There is an argument that Eysneck was simply dismissing the research because it contradicted his beliefs, but looking again at the analysis, it was clear that the inclusion critera were far, far too broad. Meta-analysis can frequently throw up different results from so called mega-trials. Meta-regression can be a fairly blunt statistical tool and the assumption of independence could impact on the results, predictions and inferences drawn from this method; a multi-level modelling approach might have been more forthcoming in that regard but perhaps impossible given the source data.
The reason why I find the above worrying is because it’s quite similar, as near as I can tell, to the kind of things supporters and proponents of alternative medicine and the like do: blatant cherry picking. One study does not a scientific consensus make, and I really shouldn’t pick the ones I like and reject the ones I don’t based on personal bias with a liberal sprinkling of cognitive dissonance: presuming that if my experience doesn’t fit with the science, the science must therefore be wrong.
My inability to fully resolve this is a worry: not as huge a worry as the fact I may be on drugs that do nothing– but a worry nonetheless…