I haven't see the study mentioned in many places yet, but there was a meta-analysis of patient preference for psychological versus pharmacological treatment of psychiatric disorders in the Journal of Clinical Psychiatry. It contained all of the usual buzzwords about evidence based medicine and why this is a hot topic to study because of the possible cost savings and potential for better outcomes if preferences were matched to actual treatments. Interestingly, in the same month a more high tech approach to matching depressed patients with pharmacotherapy versus psychotherapy came out in JAMA Psychiatry where the independent variable was a brain imaging result rather than patient preference.
The authors here looked at a final sample of 34 studies out of 644 studies that were screened. They end up with a chart of effect sizes with confidence intervals for each of the 34 studies. There were a total of 90,483 participants but 78,753 were included in one study. All of the studies are of depression and anxiety. They had tried to include studies on schizophrenia and bipolar disorder and found that they were not published. The authors conclude that their meta-analysis was valid and that there was a consistent preference for psychological treatment in the treatment seeking and non-treatment seeking or recruited patients. From this the authors suggest that patient prefernce should trump other considerations if the efficacy of both treatments are equivalent. They question why medication related treatments have increased and psychological therapies have dimished over the past decade. They suggest that the patients who prefer medication related therapies are non adherent.
In their discussion of the limitations of the study they find there was not enough data to compare combination therapy as a choice, they excluded non-published studies and therefore included potential publication bias, and they were not able to address the question about why psychological treatments were preferred over medication based therapies by a factor of 3:1.
These and other important questions have already been answered on this blog, but don't expect to see any publications on this anytime soon. Managed care has taken the very evidence based treatments that these authors emphasize and stood them on their head. I have written many times about the diagnosis of depression using rating scales and the preferred treatment of antidepressants. If you are using a primary care physician follow up code and a PHQ-9 score result to diagnose depression in ten minutes and treat all of these patients with a generic antidepressant ($4/month) - there is no psychotherapy that compares to that low cost.
All psychiatrists who are actively looking for psychotherapists to treat anxiety and depression encounter the problems of a lack of qualified therapists and more specifically a lack of therapist time in managed care systems. Managed care systems especially those that are actively managed to reduce outpatient mental health treatment has reduced available therapy in many systems to 2 or 3 sessions of crisis management and essentially limited or eliminated additional services like psychological testing that some therapists require to do their work. It is no accident that patients seeking psychological therapy can't get it. It is a conscious business decision.
The second problem is the lack of availability to research proven psychotherapies. Any psychiatrist doing patient evaluations will hear the story that therapy sessions are often very non-specific, lack goals, and often result in the patient losing faith in the process and stopping the therapy. Being seen in a psychological therapy is no assurance of a good outcome. Many patients who are provided with excellent research proven therapy are frustrated with the time commitment and stop because of the cost or number of sessions. Psychotherapy may look a lot better on paper than the reality of the relationship with the therapist and the logistics of getting to and paying for the sessions.
What can be done to improve the situation right now? The decision to take a medication for any reason is never a casual one. Taking that medication reliably is even more significant. Non medication alternatives and combination therapies to reduce exposure to medications should be available in every clinic. Instead of screening everyone for a medication on day one, non-medication alternatives should be presented at that time. There are innovative non-medication therapies such as computer delivered psychotherapy for depression, anxiety and obsessive compulsive disorder. No clinic appointments. The therapy is delivered online or by phone any time of the day or night. With the appropriate implementation, these therapies could be offered as first line treatment to massive numbers of patients. The human cost is so low they could essentially be made available across an entire health plan for free. There is no reason why networks of therapy clinics cannot be linked to primary care clinics who see the majority of patients with depression and anxiety. Any medication alternative can be discussed if the psychotherapy or non-medication intervention works.
From a research perspective if only 34 of 644 studies were suitable for inclusion in a meta-analysis, the problem is clearly not being studied very well. I think it is important to ascertain patient preference for psychological, combination, pharmacological and other (eg. lifestyle change) therapies in all registered clinical trials. In clinical practice, it is all part of informed consent for treatment. I think it is the universal experience of physicians that most people prefer to not take medications. The negative treatment of psychiatry and psychiatric medications in the press create an understandable bias against psychiatric medications relative to others as a potential source of the described phenomenon. There is some evidence that the advertising of these medications is different and potentially stigmatizing. We also need better design of clinical trials. If therapies are in fact equivalent, they need to be tested in actual clinical populations where psychiatrists work. That includes severely ill patients with comorbidity, patients who are acutely agitated and suicidal, women and children and adolescents. Much of the discussion of equivalent therapies is based on extrapolation from populations of people who are mildly depressed and in some cases who have enrolled in a number of studies.
This study highlights the current weaknesses in studying how people actually receive psychiatric treatment and how to best approach that from a research perspective. It points out that we need much better research designs and better patient selection in order to answer even basic questions about the treatment process. It should be apparent that a research design that is not adequate to describe clinical practice is not a commentary on clinical practice.
George Dawson, MD, DFAPA
McHugh RK, Whitton SW, Peckham AD, Welge JA, Otto MW. Patient preference of psychological vs pharmacological treatment of psychiatric disorders: a meta-analytic review. J Clin Psychiatry 2013; 74:6: 595-602.