Perspectives

THE COMPARABLE EFFICACY OF GENERAL MEDICINE MEDICATION AND PSYCHOPHARMACOLOGIC DRUGS

Florian Seemller1*, Emanuel Severus2, Richard Musil1, Hans Jrgen-Mller1, Stefan Leucht3
1

Ludwig-Maximilian-University of Munich, Department of Psychiatry and Psychotherapy, Nussbaumstr.7, 80336 Munich 2 Carl Gustav University Dresden, Department of Psychiatry and Psychotherapy, Fetscherstrae 74, 01307 Dresden 3 Technical University Munich, Department of Psychiatry and Psychotherapy, Ismaningerstr. 22, 81675 Munich

Background Not only psychiatry and psychiatric patients but also also psychiatric medications are stigmatized. One line of criticism arises from recent meta-analyses questioning the efficacy of psychopharmacologic compounds. Examples are meta-analyses that have questioned the effectiveness and efficacy (1) and safety of antidepressants (2). With respect to the efficacy, the presumably high placebo effect in psychiatry has also often led to general criticisms regarding the prescription of psychiatric drugs. The British Journal of Psychiatry recently published a systematic review, which is an important contribution in destigmatizing psychiatry and its pharmacological treatments (3, 4). In the review by Stefan Leucht et al., the authors compared the efficacy of psychiatric medications with general medicine medication by comparing effect sizes of meta-analyses (3). For that purpose a total of 94 meta-analyses have been reviewed, of which 48 cover general medical medications in 20 medical diseases, and 16 meta-analyses cover psychopharmacologic compounds for 8 psychiatric disorders. Mean differences and standardized mean differences, as well as absolute and relative risk reductions, were calculated. While there were some medical drugs with impressive high effect sizes, overall it can be concluded that psychiatric drugs generally were not less effective than most other medical drugs (3). For example, in patients with hepatitis C infection, the number of patients with no detectable virus at treatment end was increased from 1% to 38%, which corresponds to a standardized mean difference (SMD) of 2.27. On the other hand, there are well established, but less effective first-line, options in general medicine such as aspirin for the secondary prevention of cardiovascular diseases resulting in 8.2% participants per year with cardiovascular events taking placebo as compared to 6.7% with aspirin (SMD: 0.12). Other drugs for multifactorial diseases, such as, for example, metformin for the indication of diabetes, show a mortality rate of 14.6% against 21.7% with placebo (SMD: 0.27), and a mean fasting blood glucose level reduction
*

Florian Seemller, Ludwig-Maximilian-University of Munich, Department of Psychiatry and Psychotherapy, Nussbaumstr. 7, 80336 Munich, Germany email: Florian.Seemueller@med.uni-muenchen.de

of 1.84 mmol/l (SMD: 0.87). The use of angiotensin converting enzyme inhibitors in primary blood pressure reduces the cardiovascular events and mortality from 18.1% to 14.1% (SMD: 0.11). These data are contrasted by psychiatric medications, such as, for example, lithium in relapse prevention for bipolar disorder resulting in 81.4% relapses with placebo and 36.2% with lithium maintenance treatment (SMD: 1.2). A substantial reduction of relapse rates can also be found in schizophrenia patients with a reduction from 57% relapse with placebo to 22% with an antipsychotic (SMD: 0.92). Less pronounced are effects that can be seen with cholinesterase inhibitors for dementia with 16.8% of patients without a cognitive decline taking placebo compared to 24.4% with an active drug (SMD: 0.26). Overall, sample sizes were smaller in psychiatric indications but still sufficient due to the robust effect sizes. Some limitations There are numerous limitations for this new approach, (most of which are openly discussed by the authors themselves), which should be mentioned. For example, the translation of dichotomous variables into SMDs is only a rough estimate, which was necessary in order to get a comparable measure across the diverse outcomes. A second important limitation is that different outcomes, such as the reduction of "softer outcomes" like psychopathological symptoms assessed with rating scales and mortality outcomes are not easily comparable. However, untreated psychiatric disorders not only can lead to suicide but also may lead to negligence of, for example, a severe comorbid somatic condition, which also implies considerable health risks. In other words, psychopharmacologic compounds may also have beneficial effects which, according to Stefan Leucht, "could accumulate over time". Another limitation of the study is the selection of the reviewed medical conditions. It was consensus based and not complete and this selection might have been biased the outcome. When reviewing meta-analyses, the publication bias must also be considered and side effects, as well as psychotherapy, should be considered. One important metanalysis by Melander et al has not been considered by Leucht s systematic review, showing an impressive response difference of 16% against placebo (5). Furthermore, efficacy trials rarely allow conclusions regarding the effectiveness in day-to-day practice. Thus, as the authors truthfully acknowledge, this review only is observational and qualitative by nature. Despite these limitations, this work raises some interesting and important implications and conclusions for both fields. A comment on placebo in psychiatry and in general medicine One crucial measure that contributes to the judgement of how effective treatment x,y,z might be ( and in Leuchts analysis to the size of the SMDs) is the placebo verum difference. In other words the higher the placebo response rate the lower the difference and finally the efficacy of treatment x,y,z for any condition. A very important issue concerning placebo response rates in both, general medicine as well as in psychiatry is that "placebo effect" and "placebo group" are often used synonymously. Especially in psychiatry a variety of unspecific factors in the placebo group contribute to that what at the end of the study appears as "placebo effect". Such factors include overrating at study entry, regression to the mean, socially desired response bias from the participants, the inclusion of "professional patients" instead of a "health seeking population", supportive talks to keep patients in the study or even specific psychotherapeutic interventions. These all contribute to higher response rates in the placebo group without a specific influence on the placebo effect. Additionally considering the possibility of a ceiling effect with a maximal

achieveable response rate of for example around 70-80%, the chance of finding an absolute placebo verum difference can be significantly impaired especially in psychiatric conditions. A factor that might specifically contribute to the placebo effect may be expectations that patients have concerning the applied treatment on the measured outcome. In other words, not only in psychiatry but also in a variety of common medical conditions, the placebo effect seems to be modulated by the fact that the intake of any drug is embedded in a specific psychosocial context which gives rise to distinct expectations. These expectations may be a crucial moderator of the size of a placebo effect. So, the systematic review of Leucht serves as a good example of how the size of the placebo effect depends on the possible influence that such expectations can have on the observed outcome. Obviously expectations have only little influence on the hepatitis C virus load as they have on the measurable cognitive performance in dementia. By contrast, for the prevention of cardiovascular mortality as for the prevention of relapse in schizophrenia expectations may have greater influence. As a consequence the placebo verum differences were also markedly smaller in the latter examples. Given the fact that both in general medicine and in psychiatry expectations moderate the placebo effect size in can be questioned why this fact may be widely accepted and common sense in medical conditions on the one hand, but leads to criticism with respect to the widespread use of medications in psychiatry on the other. Amongst others, one important reason might be that not only patients with psychiatric illnesses are stigmatized but that there are also specific stigmas concerning psychopharmacology (6). Conclusion To conclude, Leucht's systematic review across the field borders reminds us in the first place that unfortunately the efficacy of pharmacology is still very limited. Secondly, distinct expectations might moderate the placebo effect on outcome in somatic and in psychiatric conditions and consequently the placebo verum differences and standardized mean differences appear quite comparable, which is an important contribution in destigmatizing psychopharmacology.

References 1. Kirsch I, Deacon BJ, Huedo-Medina TB, Scoboria A, Moore TJ, Johnson BT. Initial severity and antidepressant benefits: a meta-analysis of data submitted to the Food and Drug Administration. PLoSMed. 2008;5(2):e45. 2. Stone M, Laughren T, Jones ML, Levenson M, Holland PC, Hughes A, et al. Risk of suicidality in clinical trials of antidepressants in adults: analysis of proprietary data submitted to US Food and Drug Administration. BMJ. 2009;339:b2880. doi: 10.1136/bmj.b2880.:b2880. 3. Leucht S, Hierl S, Kissling W, Dold W, Davis JM. Putting the efficacy of psychiatric and general medicine medications into perspective: review of meta-analyses Br J Psychiatry. 2012;In Press. 4. Seemuller F, Moller HJ, Dittmann S, Musil R. Is the efficacy of psychopharmacological drugs comparable to the efficacy of general medicine medication? BMC Med. 2012;10:17. 5. Melander H, Salmonson T, Abadie E, Zwieten-Boot B. A regulatory Apologia--a review of placebo-controlled studies in regulatory submissions of new-generation antidepressants. EurNeuropsychopharmacol. 2008;18(9):623-7.

6. Moerman DE. Meaningful placebos--controlling the uncontrollable. N Engl J Med. 2011;365(2):171-2.