What with having a job, and some science of my own to be doing, it's taken me some time to get around to blogging on media claims that antidepressants are ineffective for two reasons: firstly, it's a complex subject worth doing justice - something the media have failed to do; secondly, I have a personal interest in this study myself, having been diagnosed with clinical depression during my doctorate. Here's my attempt to make sense of it all.
The Hype:
The study dominated science news in the U.K. last week. As a former user of Citalopram, the following headlines on the 25th and 26th of February interested me a great deal. The first thing I noticed was that most of them were actually bullshit. See if you can spot why...
"New generation anti-depressants have little clinical benefit for most patients, research suggests. A University of Hull team concluded the drugs actively help only a small group of the most severely depressed." (BBC) [2].
"Prozac, used by 40m people, does not work say scientists. Analysis of unseen trials and other data concludes it is no better than placebo." (Guardian) [3].
"Anti-depressants taken by thousands of Brits 'do NOT work', major new study reveals. Anti-depressant tablets taken by millions of Britons may be a waste of time and money, research shows. An analysis of dozens of studies involving thousands of patients revealed that some of the most widely-prescribed anti-depressants work little better than dummy pills." (Daily Mail) [4]
"Depression drugs don’t work, finds data review. Millions of people taking commonly prescribed antidepressants such as Prozac and Seroxat might as well be taking a placebo, according to the first study to include unpublished evidence." (The Times) [5].
The BBC get full marks for accurate reporting. The others fail because they fail to appreciate the incredibly significant point that placebos work very well for depression. The Daily Mail goes further, deliberately using inflammatory language like "dummy pills" presmably to try and push the point that the use of antidepressants is junk science (slightly ironic, given the Mail's track record). At any rate, for the papers to report that antidepressants don't work is - while tediously predictable - completely wrong.
So what is actually going on? For that we need to look at the paper.
The Paper:
The study the media have picked up on is called "Initial Severity and Antidepressant Benefits: A Meta-Analysis of Data Submitted to the Food and Drug Administration" [1]. For those not used to reading science literature, a meta-analysis is basically a systematic review of existing data - in this case, sourced from America's FDA under the Freedom of Information Act.
The abstract begins: "Meta-analyses of antidepressant medications have reported only modest benefits over placebo treatment, and when unpublished trial data are included, the benefit falls below accepted criteria for clinical significance. Yet, the efficacy of the antidepressants may also depend on the severity of initial depression scores. The purpose of this analysis is to establish the relation of baseline severity and antidepressant efficacy using a relevant dataset of published and unpublished clinical trials."
In other words we actually already knew that antidepressants performed only slightly better than placebo pills in treating depression. The purpose of this particular study is to see how much that gap changes with the severity of depression. The abstract tells us - in the subtle way that scientists reveal such things - a third thing too, that the published studies are systematically biased in favour of the drugs and their manufacturers.
Indeed the authors go on to state: "meta-analyses are often limited to published
data. In the case of antidepressant medication, this limitation has been found to result in considerable reporting bias characterized by multiple publication, selective publication, and selective reporting in studies sponsored by pharmaceutical companies". Don't let the tempered scientific language fool you - this is incendiary stuff. Drug companies stand accused of deliberately manipulating the scientific literature, duplicating positive results and hiding negative ones to produce a distorted view. No wonder GlaxoSmithKline were pissed off.
So once they had all this hidden data, what did they find? Ultimately, the graph below. Improvement is on the y-axis, initial severity of depression along the x-axis. The basic story you're seeing here is that the difference in performance between placebos and drugs becomes larger as the severity of depression increases, and only becomes clinically significant when depression is very severe.
A couple of things are quite striking at first glance. Firstly, there is a statistically significant (although not clinically significant) improvement with the drugs over placebos. (Edit: Basically - and a medical doctor might correct me slightly on this - statistical significance means that over the studies there is a noticable difference between the two sets of results, big enough to not be random noise; clinical significance means that there is a clear-cut difference between them, big enough to get excited about.)
Secondly, the increase in the size of the gap between placebos and drugs at the higher levels of depression is due to a reduction of the effectiveness of placebos, rather than an improvement in the score of the drugs.
This placebo effect is remarkable. The authors comment: "The response to placebo in these trials was exceptionally large, duplicating more than 80% of the improvement observed in the drug groups. In contrast, the effect of placebo on pain is estimated to be about 50% of the response to pain medication"
Looking at this data, I can't help but feel that the media response is hideously off-the-mark. This data isn't saying that antidepressants are ineffective or that they don't work, it's saying that the placebo effect is having an extraordinary effect when it comes to helping those with mild depression. Ben Goldacre has an excellent discussion about this effect on his blog, better than anything I can write, so I'll direct you there. But come back, I'm not done yet.
A Patient's View:
"The plural of 'anecdote' is not 'data'" said, erm... somebody. My own anecdote isn't evidence to weigh up against the study, but it's a story that I feel may have some bearing on the broader issues involved, so I'll relate it here.
It was the second year of my Ph.D. when I became chronically unhappy with my lot. It was for the usual reasons of a 20-something postgrad - a combination of an unhealthy relationship and the general stress of a Ph.D. At any rate, over a period of several months I ventured to my G.P. for help.
I visited twice. The first time I spoke to my doctor complaining of anxiety. He diagnosed anxiety, and prescribed me beta-blockers. For the first few days they really seemed to work, but gradually I lapsed. Some months later I returned, only to find that my doctor had left, and that I had to explain my problem to a replacement. To this second doctor I stated that I was depressed, upon which he diagnosed depression and prescribed Citalopram pills. Again I took them for a while, but lapsed. Later, once I had a job sorted out and the end of my Ph.D. was in sight, I became better of my own accord.
My point is that ultimately I was miserable because of things I needed to change in my life. Walking into a GP, I was able to socially-engineer my doctor into giving me antidepressants or anxiety meds by saying that I was depressed, but it wasn't what I really needed. What I needed was a friendly ear, some advice, and the motivation to put my affairs in order. What I got on each occasion was 10 minutes of being prescribed whatever I asked for.
I'm not alone. The number of people being handed antidepressants is staggering - 31,000,000 prescriptions were issued in 2006 alone [6]. Is this necessary? The real implication of this report is that millions of them could potentially be treated without drugs. Getting more exercise, changing routines, taking control of work or finances - all are valid cures for depression. Why then are we handing out tens of millions of pills?
Conclusions:
Any summary of this story has to begin with the fact that the media have spectacularly failed to grasp the implications of this report through ignorance or - in the case of the Daily Mail - hearing what they want to hear. This new meta-study doesn't prove that antidepressants don't work, but that the right psychology can give patients a massive boost.
Ben Goldacre is right when he says that we need to get a better understanding of the mysterious placebo effect, but in this case that's half the story. We seem to have settled into a convenient stable minima when it comes to the treatment of depression. Handing out pills - placebo or otherwise - is a great short-term solution for all involved - patients get the pills they turn up demanding, doctors can deal with them in a timely manner, and pharmaceutical firms make money. As a long-term solution for dealing with the nation's mental health though, I'm far from convinced.
There are two final points I'd like to make about the reputation of science and medicine. No doubt quacks will pick up on this story and show that it proves that doctors and scientists are all incompetent and in the pay of Big Pharm. In fact, that's just bollocks. Government-funded scientists have published this study that skewers the pharmaceuticals, leaving outfits like GlaxoSmithKline spluttering into their cereal. The next time someone talks to you about corrupt scientists, just point them to this case.
We do have a problem here with Big Pharm though. Pharmaceutical companies were able to systematically manipulate the science literature to promote their products and squash unfavourable reviews. Some sort of intervention is required here from on high, and Ben Goldacre makes some interesting suggestions here (also includes some great references looking at Big Pharm tactics).
I'll be interested to see how it all pans out. Meanwhile, healthy food, less beer and caffeine, and a few miles on the exercise bike every day are keeping me healthy in body and mind.
[1]Kirsch, I. (2008). Initial Severity and Antidepressant Benefits: A Meta-Analysis of Data Submitted to the Food and Drug Administration. PLoS Medicine, 5(2), 0260-0268.
[2] Anti-depressants' 'little effect' (BBC).
[3] Prozac, used by 40m people, does not work say scientists (Guardian).
[4] Anti-depressants taken by thousands of Brits 'do NOT work', major new study reveals (Daily Mail).
[5] Depression drugs don’t work, finds data review (The Times).
[6]Antidepressant prescribing soars (BBC).
Interesting. What does "the benefit falls below accepted criteria for clinical significance" mean?
Basically (and a medical doctor might correct me slightly on this): statistical significance means that over the studies there is a noticable difference between the two sets of results overall big enough to not be random noise; clinical significance means that there is a clear-cut difference between them big enough to base treatment policy on.
It's essentially just a question of scale - as Ben Goldacre puts it on the BadScience blog, a clinically significant difference is a statistically significant one big enough to get excited about.
I hope that helps a bit...
Martin is the editor of layscience.net.
Follow Me!
RSS | Twitter
I have been following the story on this study with great interest. I enjoyed reading about your experience with antidepressants. My experiences were a bit different, but relevant to understanding just how so many people end up with prescriptions for them. And I would like to add to the body of evidence...er anecdotes.
I am a nearly 50 year old woman. I am an ABD in biology. I had, still have, a hard time narrowing down my field(s) of interest. I ended up making my field of study human reproduction. (His name is Matthew and he is now 15 years old. Hence the ABD.) About a year after my son was born I began to have some sort of gastric pain. It would hit me sharply for a couple hours and then leave me aching all over for a much longer time. The GP I saw said it was fibromyalgia. The rheumatologist she sent me to confirmed. All I needed was antidepressants, the protocol for fibromyalgia; exercise; and to stop taking NSAID's for the ache as they were causing the digestive problems and they don't work for fibromyalgia anyway. Actually the NSAID's worked quite well for the ache, I did not take them during the sharp pain as they made me nauseous, but my opinions, apparently, were irrelevant, the doctors "knew" what they were talking about.
I took the prescription for awhile, but to no avail. I was becomming increasingly frustrated with the doctors who did not listen, and even scolded me that if I were following directions I would be better by now. Since I was not better, I obviously not following directions. The idea that the diagnosis might be wrong was not entertained nor would it be discussed. I was depressed and non-compliant and that was that. Changing doctors did not help, my medical file said it all.
Setting my bitterness aside... It turned out I had gallstones. I finally had emergency surgery after more than five years of listening to the fibromyalgia nonsense.
What is more relevant to the antidepressant issue is that, sometime after surgery I started to have digestive problems, again. And I ran into the same mind set in the doctors I saw. GP1: you have fibromyalgia, take Wellbutrin. GP2: you are depressed, take Effexor. Gastroenterologist: you have IBS due to fibromyalgia/depression, here is a prescription for an antidepressant. I actually did go to a reputable psychologist to get an opinion on my mental state. Her diagnosis was for very mild depression, maybe, but did not recommend drugs. Her advice was to keep knocking on doctors' doors as she could not find any indication that my symptoms were psychosomatic.
My symptoms became very debilitating. I lost thirty-five pounds, my joints turned swollen and red, I could not eat for nausea/diarrhea, I started to suffer malnutrition. I did find a doctor, a rather old-fashioned one who has been in practice for over thirty years, who did not brush me off. He properly diagnosed me with an auto-immune disease, the specifics of which he said he did not have the expertise to identify, and referred me to a rheumatologist. But the first rheumatologist I saw, can you believe it, he said I had fibromyalgia and that would treat me with exercise and ... antidepressants! It turns out I have celiac disease. A gluten free diet has kept me symptom free. But heaven help me if I make a mistake.
In my circle of friends, family and acquaintances, most of whom are mature adults, whenever conversation turns to the subject of medicine, I am continually shocked to find how many of them are on antidepressants. Anecdote is not evidence, but in my experience, antidepressants are prescribed for just about every ailment in the book. I used to think that this pertained mostly to the treatment of women, but I have learned more and more of the men I know are taking them too. They are just a little more reluctant to say so.
So what is happening here? The combination of marketing pressures from Pharma, their ability to lobby and get regulations relaxed so that any kind of doc can prescribe antidepressants, and the way health management bureaucracies pressure doctors to "treat 'em and streeet 'em" as quickly as possible? Yes, yes and yes. But we are also training the doctors themselves to think this way. I would be interested to hear other stories from other scientists. Are you taking antidepressants? What do you think?
We are now relying on people ON DRUGS for critical analysis of a study? GOD HELP US!
Thanks for you comment!
Firstly, I'd say that when you couple the powerful effect of placebos in cases of depression with the figure of around 30,000,000 prescriptions for antidepressants, you have strong evidence of over-prescription.
I can't back this up with evidence, but my opinion is that doctors themselves probably aren't ultimately to blame for over-prescription of anti-depressants - ultimately they're under a lot of pressure to provide quick solutions to increasingly demanding patients "I've google my symptons and I demand 60mg of Citalopram per day or I'm complaining". There are probably - sadly - a hell of a lot of patients who would rather have a quick-fix pill than make changes to their lifestyle/diet/fitness, and will demand accordingly. The drugs companies are offering a pill which (in fairness) was already known by existing studies to be only slightly more effective than placebos, and they're offering it because there is a demand.
What I'm saying is - in my opinion now - I don't think you can target doctors or even the bureaucracy, because regardless of how train the doctors are, people will still demand the magic pills, or worse get them from other sources (expect a rise in homeopathic remedies perhaps). I think the problem is systemic - everyone involved has settled on the easiest solution. It's a stable equilibrium convenient to all parties, and that's going to take a lot of shifting..
Martin is the editor of layscience.net.
Follow Me!
RSS | Twitter
What are you, Daily Mail boy?!
Martin is the editor of layscience.net.
Follow Me!
RSS | Twitter
Thank you. I think that does help clarify.
Does anyone know where this idea that patients demand prescriptions comes from? It sounds defensive to me. When you can't afford $80/hr for therapy and you've tried to kill yourself twice recently, and you tell a nurse and she hands you a free month's sample of Celexa, what do you do? I honestly want to know if anybody has done a survey to find out what percentage of patients actually bully doctors. The doctors are the ones who own the prescription pads and medical degrees. Are they actually claiming that it is the patients' fault for suffering and asking the doctor to help? Isn't it the doctor's responsibility to decide what to prescribe or not? I really would like to know if you have read any such survey.
Thanks, I updated the post itself to improve it based on your feedback :)
Martin is the editor of layscience.net.
Follow Me!
RSS | Twitter
Some interesting questions, I'm tempted to do a full blog on this soon with a more thorough look through the literature.
Firstly, nobody is trying to blame patients at all - that would be a far too simplistic view to take. There have been studies done, and here is an example:
"In [mild depression cases], when patients requested antidepressant medications, prescribing
was almost completely driven by patients’ requests;
the clinical presentation had virtually no effect on the decision to prescribe. In these visits, few patients who did not request medications received them and more than one half of the patients who did request them received a prescription regardless of the symptoms they reported" (http://www.annfammed.org/cgi/reprint/5/1/21).
Even if a patient isn't necessarily pushy (although anecdotally at least a number apparently are, particularly parents), diagnosis and treatment is a two-person process, it depends as much on the patient as on the doctor, especially with depression where the doctor is essentially asking the patient questions. If a patient doesn't believe a particular course of action (new diet, stop drinking, exercise, etc) will work, or they don't have the necessary motivation, then it probably won't, and they'll end up falling back on pills in the end anyway.
You're a physician. You have 10 minutes of face-to-face time to make a diagnosis and prescribe something to help that person, while trying to tell the patient that the 16hrs a year of prescription pill advertising they're exposed to each year (in the U.S. at least) is actually not accurate. It's really not a great situation to be in. If you tell them to go away and exercise/diet/whatever, some of them will, but the rest may either complain you didn't take them seriously, go to another doctor, or do nothing and end up coming back still depressed in 3 months time.
This is what I mean when I say I don't think it's a question of blaming anyone - it's not the doctors' fault or the patients' fault, it's just a product of how the current system operates. Now, the solution is debatable... it could be better public education, better physician training, or banning adverts for prescription drugs like virtually every other country has.
The last thing to remember though is that anti-depressants *do* work, at least short term, and for some people longer-term. My worry is whether it's just masking problems instead of dealing with them.
Anyway, I may return to this when I have a bit more time to sift through the relevant papers.
Martin is the editor of layscience.net.
Follow Me!
RSS | Twitter
Right on KJ! I am the author of that rather lengthy post above (aka Matthew’s mom.) At no time during my long odyssey through the healthcare system did I ask for any specific drug or any drug at all. As a matter of fact, I am rather drug adverse and submit to taking them only when taking care of myself is not working. And I expect that the drugs I do take have a scientifically studied clinical efficacy.
I had no preconceptions about my illness. I just wanted a diagnosis and a treatment protocol. That the diagnosis was depression, and let’s not deceive ourselves, when you are diagnosed with “fibromyalgia,” read “depression,” I was shocked. I did not feel well, that was obvious, but I did not meet the standards of depression as I understood them. Could I be depressed and not know it? As I admitted, I questioned my mental health after receiving this diagnosis, and sought out a mental health professional who “cleared” me.
As long as the doctors were convinced my symptoms were psychosomatic, no other options were considered. A simple ultrasound would have found the gallstones in a heartbeat, but no ultrasound was ordered, that is until I showed up at the local emergency room where my medical records were not immediately available. That brings up another interesting question. If the ER had had electronic medical records available that day, would they have seen “she’s a nut case” on my record and sent me home? Do we now need to see the psychiatrist first to certify that we are mentally healthy enough to receive physical care?
I mean no disrespect to those who suffer depression. I am not questioning the reality of depression or the devastation is can cause in one’s life. Depression sufferers deserve to be treated with respect. To hand over a few samples of an antidepressant to a true depression sufferer without more support is both wrong and dangerous. And the over-diagnosis of depression by doctors, whose specialty is not psychiatry or psychology, and the over-prescription of antidepressants as the panacea for all ailment is also wrong and dangerous.
Thank you. A couple things you just said were a little enlightening. It sounds like you have impression that some doctors at least make assumptions about what patients might do with the information they're given so they don't bother giving the information on alternatives, etc. I know I rarely get any helpful information during doctor visits. I even point blank asked a nurse once how to lower my blood pressure- if there was anything other than using less salt, because that was all I knew about it- and she told me to look it up online. (I've learned so much more on my own, particularly from this physiology course I'm taking that I almost want to learn how to diagnose myself!) I'm willing to do that, but isn't patient education part of the job? And why only ten minute visits? Why no comprehensive checklist of symptoms the person could fill out at home or something? Maybe doctors could get together on some doctors-only site and share what they know and make sure everyone is up to speed regarding diagnosing, drug knowledge, psychological skills, etc. I don't want to insult the profession, but patients do need more than they're getting, I think. Thank you for answering me and thank you very much for your patience, no pun intended.
The issue of studies, their interpretation and the training of medical people seems more important than ever.
Thanks! I'll go reread it.
I apologize for all the unfair questions. I just reread my post and it seems like I got a little hot under the collar.
No problem.
KJ said: "It sounds like you have impression that some doctors at least make assumptions about what patients might do with the information they're given so they don't bother giving the information on alternatives"
Not exactly. I'd prefer to phrase it that (the vast majority of) doctors do the best they can in the circumstances they find themselves in, but that sometimes necessitates taking short-cuts based on experience.
KJ said: "I even point blank asked a nurse once how to lower my blood pressure- if there was anything other than using less salt, because that was all I knew about it- and she told me to look it up online."
If I was a nurse I can imagine doing the same. The problem is that (here in the UK at least) giving out actual medical advice if you're not a doctor puts you on very thin ice - if she was wrong she could be held liable. Chances are she probably felt the best she could offer would be to direct you to another resource.
KJ said: "isn't patient education part of the job? And why only ten minute visits?"
Well if the visits were longer you couldn't get as many patients off waiting lists in a day. As for education, no, it isn't really at all - they don't have time for starters. Time is the big issue - doctors often work very long hours as it is, and the only way to reduce that would be to employ more doctors, which would quickly drive up the cost of medical care.
KJ said: "Maybe doctors could get together on some doctors-only site "
In the U.K. we have NHS Direct online, which is actually a pretty good resource.
I have to say I agree with a lot of the issues you raise. I don't think it's at all the fault of doctors or the medical profession, I think it's a function of how the media/industry/government as a whole works.
One of the reasons I started this blog was because the media reporting of science is abysmal, and what you're talking about is a prime example. The public have many misconceptions created almost entirely by the media that doctors then have to spend time correcting - imagine how many times doctors have had to persuade parents that the MMR vaccine is perfectly safe, for instance. This also leads to problems such as the rise of quack medicine, and patients convincing themselves they have diseases they don't.
The general consensus in the research that's been done on the problem is that we need better patient education, and better training for doctors in dealing with common communication issues Patients will never be able to safely diagnose themselves - there's a reason doctors need so many years of Med School, but if people were better educated in issues relating to drugs, medicine, and just basic critical thinking it would be a start. There's also a lot you could do administratively to help, like allowing experienced nurses to give certain medical advice perhaps.
I guess in summary I'd say that patients need more than they're getting, and so do doctors and nurses.
Martin is the editor of layscience.net.
Follow Me!
RSS | Twitter
You're fine, it didn't read badly at all
Martin is the editor of layscience.net.
Follow Me!
RSS | Twitter
That was helpful. Thank you very much. We do need more doctors in th U.S., too!
That's a very interesting perspective and I think you are right on the placebo thing but there is something else you didn't approach in your study. The risk of addiction on taking antidepressants, I think this aspect stirs even greater controversies and I would love to see your perspective on this.
Darry, addiction to oxycontin puts lives in danger