|
Rational emotive therapy with children and adolescents: a meta-analysis |
Gonzalez J E, Nelson J R, Gutkin T B, Saunders A, Galloway A, Shwery C S |
|
|
CRD summary This review, which assessed the effect of rational emotive behavioural therapy on children and adolescents, concluded that it had the most effect on disruptive behaviours and was more effective in younger children. Limited details of the methods and study results mean that the reliability of the meta-analysis results could not be assessed.
Authors' objectives To assess the impact of rational emotive behavioural therapy (REBT) on children and adolescents, and to identify variables that moderate study outcomes.
Searching ERIC, Psychological Abstracts and Sociological Abstracts were searched from 1972 to January 2002; the search terms were reported. In addition, the authors searched electronic and web-based databases held at the Albert Ellis Institute and reference lists, and handsearched journals known to publish studies on REBT. Only studies published in the English language in refereed journals were included.
Study selection Study designs of evaluations included in the reviewStudies including a control group were eligible. Case studies, narratives and other descriptive studies were excluded.
Specific interventions included in the reviewStudies comparing REBT with a control group in any setting (school- or clinic-based) were eligible. The control groups could include waiting-list and attention-control groups. In the included studies, REBT was delivered by mental health professionals (e.g. counsellors, therapists, social workers and psychologists) or non-mental health professionals (e.g. classroom teachers, students or other paraprofessionals). The content of REBT varied between the studies and included discussion, behavioural rehearsal, problem-solving, cognitive-behavioural training, and uncovering anti-empirical beliefs. The length of REBT ranged from 10 to 120 minutes and the number of sessions ranged from 6 to 85. The control treatments were no treatment, placebo or an alternative treatment.
Participants included in the reviewStudies of any school-age children or adolescents under the age of 18 years were eligible. The participants were classified into school grades: elementary (grades 1 to 6), middle (grades 7 to 9) and high school (grades 10 to 12). They were also classified according to whether or not they had presenting problems: those with no problems but considered at risk, or those presenting with problems such as text anxiety, behaviour problems or considered at risk of school failure.
Outcomes assessed in the reviewStudies providing the necessary statistical results to enable the calculation of effect sizes for relevant outcome categories (e.g. anxiety or disruptive behaviours) were eligible. The outcomes reported by the included studies were grouped into five domains: frequency of disruptive behaviours, self-concept, anxiety, rational and irrational thinking, and grade point average (measured by final school grade).
How were decisions on the relevance of primary studies made?The authors did not state how the papers were selected for the review, or how many reviewers performed the selection.
Assessment of study quality The studies were classed as high or low validity. Studies high in internal validity were those with random treatment assignment, use of psychometrically sound instruments, and with a drop-out rate of less than 15% that was similar between groups; all other studies were classed as having low internal validity.
The authors did not state how the validity assessment was performed.
Data extraction The authors did not state how the data were extracted for the review. Some details (internal validity, grade level, study population, therapist category and comparison group) were extracted by more than one author, with coding disagreements resolved by discussion and consensus. However, it was unclear if all data, particularly the numerical results, were extracted in this way. Effect sizes (Z) were calculated for each study using Fisher z transformed correlation estimators. Other statistical results, such as p-values, t- or F-statistics, were transformed into correlation coefficients and then Z effect sizes using recommended methods.
Methods of synthesis How were the studies combined?The studies were combined in a meta-analysis and the effect sizes pooled, weighted by sample size, for each outcome measure. Average weighted effect sizes and 95% confidence intervals (CIs) were calculated overall and for each type of outcome. If a study used a number of measures of the same outcome, then an average weighted effect size was calculated for that study.
How were differences between studies investigated?Subgroup analyses were used to assess differences between the studies. Average effect sizes were calculated for groups based on study validity (low, high), study population (children presenting with and without problems), type of therapist (health professional, non-mental health or other professional), duration of REBT (low, 60 to 375 minutes; medium, 675 to 770 minutes; or high, 1,200 to 2,115 minutes), control group (no treatment, alternative treatment), and school grade (elementary, middle or high school).
Results of the review Nineteen studies were included in the meta-analysis (1,021 received REBT treatment).
The overall average effect size was 0.50 (95% CI: 0.40, 0.61; from 56 effect sizes). For the different outcome categories, REBT was observed to have the largest positive effect on disruptive behaviours with an average effect size of 1.15 (95% CI: 0.89, 1.42; from 7 effect sizes) and the smallest effect on self-concept (effect size 0.38, 95% CI: 0.34, 0.41; from 15 effect sizes).
For subgroup analyses, there were statistically significant differences between the effects of REBT delivered by health professionals (effect size 0.36, 95% CI: 0.34, 0.38) and non-mental health professionals (0.54, 95% CI: 0.51, 0.57). There was also some evidence that REBT was less effective in middle school students (0.18, 95% CI: 0.16, 0.20) compared with elementary (0.70, 95% CI: 0.62, 0.78) and high school students (0.51, 95% CI: 0.48, 0.55). A short total duration of REBT was significantly less effective than a long duration (0.22, 95% CI: 0.20, 0.25 and 0.59, 95% CI: 0.54, 0.64, respectively). There were no statistical differences between studies of low and high validity, between those in populations with and without problems, and between different types of control treatment.
Authors' conclusions The authors' conclusions appear to have been that: REBT had the most impact on child and adolescent disruptive behaviours; there was little difference between studies of high or low validity, or of those where the client did or did not have a presenting problem; non-mental health professionals produced larger REBT effects than mental health professionals, and longer duration of REBT was also more effective; and children appear to derive more benefit from REBT than older adolescents.
CRD commentary This review stated broad inclusion criteria with respect to the intervention, participants and outcomes. The electronic database searches seemed relevant and attempts were made to locate unpublished articles. However, the restriction to English language papers published in peer-reviewed journals means that some relevant studies might have been missed. It was unclear if the methods of the systematic review (study screening, selection, validity assessment and data extraction) were performed in duplicate, and this might have introduced the potential for error and bias. The authors stated that they assessed the inter-rater reliability of some coding, but this was poorly described and appeared to be only for some data. Study validity was assessed but details for each study were not reported. Effect sizes were pooled within studies and then between studies but, as no details of the numerical results for each study were presented, it was therefore not possible to assess the reliability of the pooled estimates. A number of subgroup analyses were performed and these formed the basis of most of the review conclusions but, again, the limited details presented mean it was difficult to assess their reliability. Overall, the reliability of the authors' conclusions is unclear and caution is advised.
Implications of the review for practice and research Practice: The authors did not state any implications for practice.
Research: The authors stated that future research should be of high quality; link treatment outcomes to demographic characteristics; report the integrity of implementing different aspects of REBT treatment; and use longer follow-up periods to assess the maintenance effects of REBT.
Funding U.S. Department of Education, grant numbers H324X010010, H324D010013 and H325D990035.
Bibliographic details Gonzalez J E, Nelson J R, Gutkin T B, Saunders A, Galloway A, Shwery C S. Rational emotive therapy with children and adolescents: a meta-analysis. Journal of Emotional and Behavioral Disorders 2004; 12(4): 222-235 Indexing Status Subject indexing assigned by CRD MeSH Adolescent; Anxiety Disorders /diagnosis /psychology /therapy; Behavior Therapy; Child; Child Behavior Disorders /diagnosis /psychology /therapy; Emotions; Neurotic Disorders /diagnosis /psychology /therapy; Psychotherapy, Rational-Emotive /methods; Schools; Social Adjustment; Treatment Outcome AccessionNumber 12005006033 Date bibliographic record published 31/10/2007 Date abstract record published 31/10/2007 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. |
|
|
|