One in six people suffer from some sort of clinical depression during their lifetimes. They will therefore have been distressed to learn about this paper in PLoS Medicine presenting a meta-analysis (at which word alarm bells should ring) of efficacy of various antidepressants and show that they are no better than placebos in all patients except those showing severe symptoms of depression. Cue the usual brouhaha in The Times, The Telegraph and the BBC and who knows where else.
It’s only when you get into the paper itself that you find that the studies were conducted over timescales far too short to have been clinically relevant.
“In order to generalize the findings of the clinical trial to a larger patient population, FDA reviewers sought a completion rate of 70% or better for these typically 6-wk trials. Only four of the trials reported reaching this objective, and completion rates were not reported for two trials. Attrition rates were comparable between drug and placebo groups. Of those trials for which these rates were reported, 60% of the placebo patients and 63% of the study drug patients completed a 4-, 5-, 6-, or 8-wk trial.”
This will affect the conclusions of the study, as every doctor (and patient) knows, antidepressants are drugs for the long haul. It takes weeks for them to have much effect, and this study seems to have had a cutoff before any such effect could be manifested. The results of this study are therefore compromised, and people who have been distressed by it have, I think, been misled.
But don’t take my word for it. People who really ought to know – psychiatrists and other health professionals – have already started to cast doubt on this study.
So, shame on PLoS Medicine for touting what looks like a sensationalist story that grabs headlines on the distress of others; and shame (as usual) on the hog-whimperingly low standards of science news editing and reporting that have failed to pick up on this important flaw.
I think that the spokespeople from GSK et al. have missed a bit of a trick in not actually critiquing the study on the scientific grounds that you’ve used, Henry. All their quotes essentially say “No, we’re right” and without saying why, it sounds a bit too corporate.
I too am amazed to read about these short timescalse. The SSRI class is known to take a few weeks to “kick-in”.
As usual with these issues, the press does not report scientific/medical research with much intelligence (rehash press release anyone?), and the issues just aren’t polarised. Yes, anti-depressants probably are oversubscribed now as older drugs were (valium, Jaqueline Susann et al) in their day. There is just so much grey.
er, make that timescale.
I am not amazed by the fact that there is not a great deal of difference between the placebos and the drug treated patients. Antidepressants target specific molecular entities or cellular signal transduction pathways for giving you the relief that “you expect to get”. What many of us may be overlooking is the fact that our brain functions is a very complex way.
A) Our neurons are sensing the dynamic changes in a number of chemical substances at any point of time. As a result, there occurs short term and long term changes in the brain.
As we learn more and more about our nervous system functions, we understand that within sets of neurons exist subsets and so and so forth where our “feelings” are filed and from where we retrieve the information that we stored based on the cues that we put in.
These neurons constantly communicate with each other. In other words, the deceptively simple term “depression” can be understood in part at the cellular level as resulting from a communication problem between neurons. The second part of the depression may be due to long term changes induced by prolonged short term changes that occurred within the neurons.
B) Many of the drugs available in the market works from a symptomatic standpoint rather than eliminating the causal factors.
C) The placebos here are not similar to the placebos where we either give an antibiotic or a sugar tablet. The placebos could function as psychologists when it comes to treating depression. Unless you have severe clinical depression, a good psychologist will be able to do a wonderful job that psychoactive drugs may not be able to do. There are exceptions to this case. Everything depends on the stimuli for depression.
An interaction of our environment with our mind controls our mood. And and our mood directly and indirectly controls our mind.
It is expectation-effort-experience-reward loop at work!!! The best way to prevent depression may be by expecting less, by developing good adaptability and if possible by keeping away from unhealthy social environments!!!.
I wonder what the drugs companies were doing using such short time scales for their trials in the first place. Was it thought at the time that two months would be long enough to have an effect?
It looks odd to me, but I’m not in clinical trials so perhaps I’m showing my extreme ignorance.
Incidentally, Ben Goldacre has a post about the same topic.
Ben Goldacre’s story says little or nothing about the timescales of the studies, but concentrates on the evil nastiness of drug companies (well, what do you expect from the Grauniad?)
I noticed that, Henry. He seemed to miss the point completely.
I would encourage further commenter’s here to (also) please place their comments here so that they contribute directly to the ongoing lively discussion.
Many thanks.
To be fair, Ben focused his first column on the inaccurate media response to the story. He says he’s keeping his options open to write about the article itself at the weekend – remember also these are themselves columns for publication, so I would guess there are quite strict word limits.
However, even in the piece that’s there, we see the following:
“2. It did not look at all the trials ever done on these drugs: it looked only at the trials done before the drugs were licensed (none of them more than six weeks long),”
So he’s certainly aware of the limitation.
The question is why the paper itself focused only on the studies done before clinical approval was granted (which are necessarily shorter than the studies that can be done once a drug is being actively prescribed). The answer seems to be that this is the only way to avoid publication bias, since it’s only the pre-approval stage that requires all trials to be registered and reported.
It’s a way of making the scientific and political case for the trial registry Ben advocates – i.e. that all studies on drug efficacy should be required by law to be made public, even studies on existing drugs.
Peter – thanks for these excellent points, and the clarification. I had been worrying, increasingly, why it was that only very short trials had been available to the researchers, given that antidepressants presumably need to tested in patients for very much longer intervals, just to break free of the placebo effect and so forth. It’s worrying, though, that only those trials from the pre-approval stage are required to be registered and reported.
>It’s worrying, though, that only those trials from
>the pre-approval stage are required to be registered
>and reported.
Quite! In fact, it’s worse than that – even those trials are only required to be reported to the licensing authorities themselves, not necessarily made publicy available. The researchers on the PlOS paper had to use the Freedom of Information Act just to get their hands on the data in the first place!
There is a key problem with your argument. Many physicians (and most patients) do not appreciate the long-haul nature of treatment with these drugs.
Therefore, the accurate conclusion is really that these drugs are ineffective as prescribed.
Then, Jonathan, the problem is not with the argument but the prescribing physicians!