In the American Journal of Psychiatry (7/13), B.R. Rutherford and S.P. Roose reviewed the results of published antidepressant clinical trials.
Less than half the studies found the antidepressant medications superior to placebo. The overall indication remained, however, very strongly in favor of the use of these medications for many depressed people. Depression takes a devastating toll on the lives of afflicted individuals and on the functioning and quality of life of our population as a whole. Medication remains one of our strongest resources in treating it. Studies show that many depressed people will be greatly benefitted, that many will no longer be depressed.
The studies that don’t seem to show those benefits run into an unusual problem when the benefits are “compared to placebo”. The research problem is often not that the people who get medication don’t do well, but that those who get placebo also do well! Why should this be? When both groups do well, we don’t see much difference, and the advantage of the medication is less apparent. Even in the studies that do demonstrate a substantial advantage in the medications, the size of the difference is often less than we’d expect. The research results seem less impressive than the results doctors see when actually treating patients. Why this “placebo issue”?
The studies consist of giving one group of depressed people the medication studied (“the active medication”), while giving another group pills (placebos) which look like the active medication but contain no active medication. The researcher then measures the difference between the results with active medication and the results with placebo. In these antidepressant studies, the high figures when using placebo have been unexplained and a cause of confusion. In contrast, for example, in studies of antipsychotic medications, active medication figures are higher and, more to the point, placebo figures are lower. Understanding the reasons for these depression results seems critical to our understanding of the medications and to our understanding of depression.
The studies don’t indicate which depressed people will be helped and which will not. This may be a key to the answer.
In research studies, people receiving placebo are not receiving nothing. The researcher pays attention, listens, is “there”, acknowledges and more often agrees with what’s said. He or she may be engaging, may be empathic. In giving information about the study, information is often given about depression and about symptoms. Tests may be done. This all may involve more visits than many people get when actually seen in treatment in a doctor’s office. The researcher may convey enthusiasm about the study. There is a hopefulness factor- maybe the subject is in the group that’s getting the active medication and maybe the medication is great. But why should the placebo factor be so much greater in depression studies than in other studies, e.g. schizophrenia studies? The answer is important: we need to know as much as we can about when to give the medication and when not to.
The answer may lie in how the studies define depression. There’s evidence, for example, that the difference in effect between active medication and placebo is greater in cases where the depression is more severe. It may also be that the depression defined in the study is not a single disorder but a collection of different disorders which react differently to medication and may react differently to placebo. A classical depression (often called “monopolar”) may differ from the depression which occurs as part of a “bipolar” disorder. These may differ from the depression which accompanies (is “comorbid” to) a different disorder, for example, a traumatic disorder. All three may differ from the depression which is not a “disorder” but a “feeling very bad” reaction to “real” setbacks or losses. There are numerous other differences as well. Histories differ, personalities differ, family systems differ. In contrast, studies of schizophrenia may more clearly define the target of the antipsychotic medication. The more “depression” is a collection of different conditions, the more we might expect medication results to differ and, perhaps, placebo effects as well. Perhaps medication benefits the more “clinically” depressed and placebo benefits many unhappy people with less “clinical” depression. This would narrow the gap between medication and placebo results.
If this is true, it’s a strong indication for better definition in research studies. The Hamilton and other scales and DSM itself may poorly define depression. They may lack “specificity”. They may too roughly detect some assembled group of disorders and the result may be rough and not nearly helpful enough. Studies can surely be done that would better differentiate, better define. These would not only teach us about medication and its best use, but teach us more of what depression is really all about.
All of this surely tells us that an able psychiatrist, when treating his very specific individual with depression, must do a better evaluation and individualize both the treatment and the ways in which he or she works with that person. When research doesn’t tell us enough, we must use any and all the ability, training, and experience we have, to form our approach to the depressed person we work with. When our efforts are under way, we must see him or her often enough and use all we can gather in each visit to learn how to go further.
We may decide to medicate or not. In either case, we must have the ability, the personal professional resources, to work with the person in our office as an individual with his or her own personality and history and a member of his or her own family system.