|
Cognitive behavior therapy for schizophrenia: effect sizes, clinical models, and methodological rigor |
Wykes T, Steel C, Everitt B, Tarrier N |
|
|
CRD summary This review concluded that in comparison with control, cognitive behaviour therapy for psychosis had a beneficial effect on positive symptoms in patients with schizophrenia. Further methodologically sound research was required to confirm the effects. Overall, the cautious findings appeared to be supported by the data presented and were likely to be reliable. Authors' objectives To examine the effect sizes for targeted and non-targeted symptoms and modes of action in trials of cognitive behavioural therapy for psychosis (CBTp) in patients with schizophrenia and how these are affected by the methodological rigour of the trials. Searching EMBASE, MEDLINE, Current Contents, Web of Science, PsycINFO and CCTR were searched. Search terms were reported, but dates were not. The reference lists of retrieved articles, reviews and meta-analyses, in addition to conference presentations (2004 to 2006), were checked for further studies. Relevant research groups and the Beck Psychosis Networks were also approached for information about additional studies. Published and unpublished studies were eligible for inclusion in the review. Study selection Studies comparing CBT with a control group in study populations where the majority of participants were diagnosed with schizophrenia were eligible for inclusion in the review. Both intervention and control groups had to receive treatment as usual (TAU). Participants had to be allocated to a study group using a specified allocation procedure. CBT had to be targeted at positive or negative mood symptoms of psychosis, functioning, hopelessness/suicidality or social anxiety. Studies where insufficient data were available for inclusion in the meta-analysis were excluded from the review.
Most included studies assessed chronic CBTp aimed at individual patients with chronic schizophrenia undergoing treatment in the community. Of the remaining studies, the majority assessed acute patients and those undergoing group CBTp. The clinical emphasis of the studies varied. Almost two thirds of the included studies were conducted in the United Kingdom. All of the studies used continuous measures of outcome for the targeted symptom of interest and nearly all considered positive symptoms.
Two reviewers independently assessed the studies for inclusion in the review. Assessment of study quality Study quality was assessed using the Clinical Trial Assessment Measure (CTAM), which includes 15 criteria assessing aspects of the sample characteristics, allocation to treatment, assessment of outcome, use of control groups, description of control and intervention treatments, and the use of adequate statistical analyses. Each study was awarded a score up to a maximum of 100.
Data were extracted independently by two reviewers and any discrepancies were discussed until a consensus was reached. Data extraction Effect sizes (ES) were calculated for each outcome using the means of the control and intervention groups, and the standard deviation (SD) of the control group post-intervention. All effect sizes were recalculated from the study data. Only the effects after treatment and not at follow-up were considered. Study authors were contacted for additional data.
Three authors assessed the validity of the studies and any discrepancies were discussed until a consensus was reached. The assessments were also sent to the study authors to check their accuracy. Methods of synthesis Studies were grouped according to outcome. Pooled mean weighted effect sizes were calculated with 95% confidence intervals using a random-effects model. Statistical heterogeneity was assessed using the Χ2 statistic with significance at the 5% level. Significant heterogeneity between studies was investigated in further analyses. Analyses were conducted to investigate the effects of overall study quality, specific aspects of quality and the clinical emphasis of studies. Correlations between the different outcome measures were also investigated. Publication bias was assessed using a funnel plot. Results of the review Thirty-four published and unpublished studies were included in the review. The study designs and numbers of participants were not reported. Sample sizes were reported to range from 11 to 353 (average 58.2). The mean CTAM score was 61.2 (SD 18.1) out of a maximum of 100 (range 27 to 100). All but four studies used random allocation; five studies did not use independent outcome assessment; few studies (number not stated) described blinding and verified blinding at the end of the study; seven studies had a dropout rate of more than 25 per cent; and more than half of the studies did not adequately account for loss to follow-up in their analyses. Overall, loss to follow-up ranged from 0 to 45% (median 14.5%).
There were statistically significant effects in favour of CBTp versus control for target symptoms (ES 0.40; 95% CI: 0.25, 0.55; 33 studies); positive symptoms (ES 0.37; 95% CI: 0.23, 0.52; 32 studies); negative symptoms (ES 0.44; 95% CI: 0.17, 0.70; 23 studies); functioning (ES 0.38; 95% CI: 0.15, 0.60; 15 studies); mood (ES 0.36; 95% CI: 0.08, 0.65; 15 studies); and social anxiety (ES 0.35; 95% CI not reported; two studies). There were no significant differences in hopelessness. However, all outcomes with the exception of social anxiety showed evidence of significant statistical heterogeneity. Where only higher quality studies were considered, no significant differences between control and CBTp were reported except for positive symptoms. Further analyses suggested that there were significant correlations between a number of the different outcomes (further details reported in the review). Relationships between study quality and effect sizes were evident for each symptom area, with studies with lower CTAM scores reporting larger effect sizes. No significant association was reported between effect size and whether group or individual CBTp was used, however, those studies with a more behavioural clinical emphasis tended to produce greater effect sizes.
No evidence of publication bias was detected. Authors' conclusions CBTp had a beneficial effect on positive symptoms in patients with schizophrenia, but further methodologically sound research was required to confirm these effects. CRD commentary This review assessed a clear research question and included both published and unpublished studies, suggesting that the risk of publication bias was low. Some attempt to reduce the risk of reviewer error and bias was taken when selecting studies, but it was not reported whether similar care was taken during the rest of the review process. Study quality was assessed and its effects on outcome effect sizes analysed. Similarly statistical heterogeneity was assessed and its effects investigated in further analyses. Some details of the included studies were reported, but it was unclear what types of study designs were included and what their sample sizes were. Details about the baseline characteristics of the included participants were not reported. Despite these concerns, the authors' cautious findings appeared to be supported by the data presented and were likely to be reliable. Implications of the review for practice and research Practice: the authors stated that their findings supported the inclusion of CBTp in guidance for the treatment of schizophrenia.
Research: the authors stated that further methodologically rigorous trials of well-defined CBTp treatment programmes were required to assess the long-term effectiveness and durability of treatment effects. Such trials should include measures of the treatment process that allow an estimate of the effective dose of treatment other than just the number of sessions. Trials should also use outcome measures that were acceptable to clinicians, researchers and patients, and assess different models, staff training methods and background service provision. Bibliographic details Wykes T, Steel C, Everitt B, Tarrier N. Cognitive behavior therapy for schizophrenia: effect sizes, clinical models, and methodological rigor. Schizophrenia Bulletin 2008; 34(3) : 523-537 Indexing Status Subject indexing assigned by NLM MeSH Cognitive Therapy /methods; Humans; Psychometrics /standards; Schizophrenia /diagnosis /therapy AccessionNumber 12008103475 Date bibliographic record published 23/12/2008 Date abstract record published 22/04/2009 Record Status This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn. |
|
|
|