Adapted cognitive–behavioural therapy required for targeting negative symptoms in schizophrenia: meta-analysis and meta-regression

• Published in: Psychological medicine 2015-02, Vol.45 (3), p.453-465
• Main Authors: Velthorst, E., Koeter, M., van der Gaag, M., Nieman, D. H., Fett, A.-K. J., Smit, F., Staring, A. B. P., Meijer, C., de Haan, L.
• Format: Article
• Language: English
• Subjects: Cognitive therapy; Cognitive Therapy - methods; Humans; Meta-analysis; Psychopathology; Randomized Controlled Trials as Topic; Regression Analysis; Review Articles; Schizophrenia; Schizophrenia - therapy; Systematic review
• ISSN: 0033-2917; 1469-8978
• DOI: 10.1017/S0033291714001147

There is an increasing interest in cognitive-behavioural therapy (CBT) interventions targeting negative symptoms in schizophrenia. To date, CBT trials primarily focused on positive symptoms and investigated change in negative symptoms only as a secondary outcome. To enhance insight into factors contributing to improvement of negative symptoms, and to identify subgroups of patients that may benefit most from CBT directed at ameliorating negative symptoms, we reviewed all available evidence on these outcomes. A systematic search of the literature was conducted in PsychInfo, PubMed and the Cochrane register to identify randomized controlled trials reporting on the impact of CBT interventions on negative symptoms in schizophrenia. Random-effects meta-analyses were performed on end-of-treatment, short-term and long-term changes in negative symptoms. A total of 35 publications covering 30 trials in 2312 patients, published between 1993 and 2013, were included. Our results showed studies' pooled effect on symptom alleviation to be small [Hedges' g = 0.093, 95% confidence interval (CI) -0.028 to 0.214, p = 0.130] and heterogeneous (Q = 73.067, degrees of freedom = 29, p < 0.001, τ 2 = 0.081, I 2 = 60.31) in studies with negative symptoms as a secondary outcome. Similar results were found for studies focused on negative symptom reduction (Hedges' g = 0.157, 95% CI -0.10 to 0.409, p = 0.225). Meta-regression revealed that stronger treatment effects were associated with earlier year of publication, lower study quality and with CBT provided individually (as compared with group-based). The co-occurring beneficial effect of conventional CBT on negative symptoms found in older studies was not supported by more recent studies. It is now necessary to further disentangle effective treatment ingredients of older studies in order to guide the development of future CBT interventions aimed at negative symptom reduction.