|
The efficacy of problem-solving treatments after deliberate self-harm: meta-analysis of randomized controlled trials with respect to depression, hopelessness and improvement in problems |
Townsend E, Hawton K, Altman D G, Arensman E, Gunnell D, Hazell P, House A, Van Heeringen K |
|
|
Authors' objectives To determine the efficacy of problem-solving treatments on depression, hopelessness and improvement in problems following deliberate self-harm.
Searching MEDLINE, EMBASE, PsycLIT, the Cochrane Controlled Trials Register (CENTRAL), and the Cochrane Collaboration Depression, Anxiety and Neurosis Controlled Trials Register were searched. The authors also reported handsearching the worldwide literature on deliberate self-harm; details of the sources and dates searched were provided in the Cochrane Review (see Other Publications of Related Interest).
Study selection Study designs of evaluations included in the reviewRandomised controlled trials (RCTs) were eligible for inclusion. The treatments in the included studies lasted from 3 weeks to 3 months (duration was not reported in one study). Four of the included studies were conducted in England and one in Ireland (Cochrane Review, see Other Publications of Related Interest); the country was not reported for the other study.
Specific interventions included in the reviewStudies of any problem-solving therapy compared to any control were eligible for inclusion. The interventions included in the review were crisis-oriented, time-limited task-centred social work at home; interpersonal problem-solving skills training; out-patient problem-oriented therapy; cognitive-behavioural problem-solving treatment at home; and Manual Assisted Cognitive Behavioural Therapy. The control interventions included routine treatment (GP, psychiatric or other referral), standard psychiatric treatment, individual therapy with a clinical psychologist, GP care, and standard aftercare (brief problem-solving therapy). Further details of the intervention and control therapies were given in the report.
Participants included in the reviewStudies of deliberate self-harm patients were eligible for inclusion. Studies that included suicide-ideators without self-harm, or people with depression for whom deliberate self-harm was an outcome variable, were excluded. The participants in the studies included in the review were males and females, aged 15 to 65 years (age was not reported in one study), who had undertaken deliberate self-injury or self-poisoning shortly before being recruited into the study. Self-harm repeaters were included.
Outcomes assessed in the reviewThe outcomes of interest were depression, hopelessness and improvement in problems. The measures of outcome analysed in the review were the Beck Depression Inventory, Hospital Anxiety and Depression Scale (Depression score), the Beck Hopelessness Scale, and improvement in defined target problems or problems in personal relationships.
How were decisions on the relevance of primary studies made?Abstracts identified from the electronic database searches were screened by one reviewer to select potentially relevant studies. Two reviewers then independently applied the inclusion criteria to those studies. Details of the handsearching procedure were not reported.
Assessment of study quality Concealment of the allocation sequence was rated as adequate, inadequate or unclear. Two independent reviewers applied the quality criteria whilst blind to the study authors. Any disagreements were resolved by consensus with the help of a third reviewer.
Data extraction Two reviewers independently extracted the data and any disagreements were resolved by consensus with a third reviewer. Data on the three outcome variables that were reported in more than one study, in a way that could reasonably be combined, were extracted for use in the analysis of depression, hopelessness and improvement in problems. Missing data (means and standard deviations) were obtained from the study authors or imputed from the reported information. This was described fully in the review.
Methods of synthesis How were the studies combined?The studies were combined using both fixed- and random-effects meta-analyses. Continuous data (hopelessness and depression) were combined using the weighted mean difference (WMD) for hopelessness or the standardised mean difference (SMD, or Cohen's d) for depression, along with 95% confidence intervals (CIs). Dichotomous data (improvement in problems) were combined using the Mantel-Haenszel odds ratio (OR) with 95% CI.
How were differences between studies investigated?A chi-squared test for heterogeneity was used in the meta-analysis.
Results of the review Six RCTs (n=583) were included.
Five of the six studies reported adequate concealment of the allocation sequence, indicating reliable randomisation.
Problem-solving therapy showed a statistically-significant reduction in depression based on four studies (n=258). The SMD was -0.36 (95% CI: -0.61, -0.11) when using a fixed-effect analysis. There was borderline evidence of statistical heterogeneity between the studies (P=0.1); however, the treatment effect was still statistically significant in a random-effects analysis (SMD -0.49, 95% CI: -0.89, -0.08).
Problem-solving therapy showed a statistically-significant reduction in hopelessness based on three studies (n=63). The WMD was -2.97 Hopelessness Scale points (95% CI: -4.81, -1.13) when using a fixed-effect analysis. The test for statistical heterogeneity between the studies was not significant (P=0.2).
Problem-solving therapy showed a statistically-significant improvement in problems based on two studies (n=211). The OR was 2.31 (95% CI: 1.29, 4.13) when using a fixed-effect analysis. The test for statistical heterogeneity between the studies was not significant (P=0.3) and the treatment effect was almost identical in the random-effects analysis. Three studies using other measures of problem-solving, which were not included in the meta-analysis, also showed improvement.
Authors' conclusions The authors concluded that problem-solving therapy appears to have been better than control treatments for the outcomes depression, hopelessness and problems among deliberate-self-harm patients.
CRD commentary This review had a clear research question and explicit inclusion criteria. Steps were taken to minimise bias and errors in the study selection, validity assessment and data extraction processes. Various sources were searched to identify relevant studies. The validity of the included studies was assessed using a criterion (allocation concealment) that has been shown empirically to be associated with bias in RCTs. This provides some reassurance of the reliability of the design of the primary studies. The method used to combine dichotomous data (improvement in problems) was appropriate and the authors acknowledged the limitations of their pooling of the continuous data. A test for statistical heterogeneity (differences) between the studies was applied in the meta-analysis, but the number of studies in each analysis was small, limiting the power of the test. However, the more conservative random-effects estimates of the pooled effect sizes did not change the findings. The authors' tentative conclusions reflect the strength of the evidence presented in terms of the small number of studies and participants, the nature of the continuous data in particular, and the differences between the content and context of the interventions. In terms of applicability, most of the included studies were conducted in England and Ireland.
Implications of the review for practice and research Practice: The authors stated that those who provide services for deliberate self-harm patients should investigate the provision and evaluation of problem-solving therapy.
Research: The authors stated that a large trial of problem-solving therapy is needed, which should assess mood, hopelessness and improvement in problems (as indicators of well-being, quality of life and risk of repetition) as well as repetition of deliberate self-harm.
Funding South East Region NHSE Research and Development Committee; the former Anglia and Oxford NHSE Research and Development Committee.
Bibliographic details Townsend E, Hawton K, Altman D G, Arensman E, Gunnell D, Hazell P, House A, Van Heeringen K. The efficacy of problem-solving treatments after deliberate self-harm: meta-analysis of randomized controlled trials with respect to depression, hopelessness and improvement in problems. Psychological Medicine 2001; 31(6): 979-988 Other publications of related interest Hawton K, Townsend E, Arensman E, Gunnell D, Hazell P, House A, Van Heeringen K. Psychosocial and pharmacological treatments for deliberate self-harm (Cochrane Review). In: The Cochrane Library, Issue 4, 2003. Oxford: Update Software.
Indexing Status Subject indexing assigned by NLM MeSH Cognitive Therapy /methods; Depressive Disorder, Major /therapy; Follow-Up Studies; Humans; Problem Solving; Randomized Controlled Trials as Topic; Self-Injurious Behavior /psychology /therapy AccessionNumber 12001002043 Date bibliographic record published 31/01/2004 Date abstract record published 31/01/2004 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. |
|
|
|