1 in The Lancet Psychiatry compares the effi cacy of pharmacological and psychotherapeutic treatment strategies for adults with obsessive-compulsive dis-order. The authors should be commended for having systematically reviewed just over 36 years of literature (1980–2016) and for using a new analytical approach— network meta-analysis—which allows for simultaneous comparison of several treatments in a single model and has been previously applied to other mental disorders. 2 The network meta-analysis 1 included 54 randomised controlled trials (RCTs) and 6652 participants. This comprehensive, carefully done study 1 is an example of how obtaining of evidence-based information to guide clinical practice is challenging, even in a well explored research area. Despite the large number of published RCTs, important clinical questions remain unanswered concerning obssessive-compulsive disorder treatment. As properly acknowledged by the authors, several gaps in knowledge remain to be addressed in future research. One of these gaps is that no potential diff erences in the effi cacies of individual SSRIs or between SSRIs as a group and clomipramine could be shown; they seemed to be equally eff ective. The mean Yale-Brown Obsessive Compulsive Scale diff erence was –3·49 (95% credible interval –5·12 to –1·81) for SSRIs and –4·72 (–6·85 to –2·60) for clomipramine. The authors could not examine whether higher doses of the same drug are associated with greater treatment responses. Likewise, when studies that used waiting list control groups were excluded, the eff ects of recommended psychotherapies were similar in magnitude. The mean Yale-Brown Obsessive Compulsive Scale diff erence was –10·41 (95% credible interval –14·04 to –6·77) for behavioural therapy, –9·45 (–13·76 to –5·19) for cognitive therapy, and was patient rated (rather than clinician rated) is even more worthwhile to highlight. The patient's point of view is increasingly sought because, in the end, it is the patient not the doctor who must take the medication. Diff erences between the full versus the per-protocol sample are common because clinically important events, such as poor effi cacy or unpleasant side eff ects, can cause patients to withdraw from a study prematurely, such that the fi nal per-protocol results do not refl ect the important diff erences between the interventions. However, the diff erence in CGI effi cacy index (CGI minus side eff ects at 6 weeks) was signifi cant, suggesting that the index is driven by a side-eff ect diff erence. Further analyses could explore whether the quality-of-life diff erences were mediated by changes in extrapyramidal symptoms. Unfortunately, the subjective well being under neuroleptic treatment (SWN-K) scale (high with FGAs) was not given at each visit, which might have shed light on the mechanisms.
Torres, A. R. (2016, August 1). The road to improvement in obsessive-compulsive disorder. The Lancet Psychiatry. Elsevier Ltd. https://doi.org/10.1016/S2215-0366(16)30118-3