Does psychotherapy work with school-aged youth? A meta-analytic examination of moderator variables that influence therapeutic outcomes

Forfatter(e)

Fedewa, A. L. Ahn, S. Reese, R. J. Suarez, M. M. Macquoid, A. Davis, M. C. Prout, H. T.

År

2016

DOI

10.1016/j.jsp.2016.03.001

Tidsskrift

Journal of School Psychology

Sider

59-87

Kategori(er)

Atferdsproblemer, antisosial atferd og atferdsforstyrrelser Skoleprestasjoner Skolefravær

Tiltakstype(r)

Kognitiv atferdsterapi, atferdsterapi og kognitiv terapiLeketerapi Psykodynamisk/interpersonlig terapi

Abstract

The present study is a quantitative synthesis of the available literature to investigate the efficacy of psychotherapy for children's mental health outcomes. In particular, this study focuses on potential moderating variables-study design, treatment, client, and therapist characteristics-that may influence therapeutic outcomes for youth but have not been thoroughly accounted for in prior meta-analytic studies.

An electronic search of relevant databases resulted in 190 unpublished and published studies that met criteria for inclusion in the analysis. Effect sizes differed by study design. Pre-post-test designs resulted in absolute magnitudes of treatment effects ranging from -0.02 to -0.76 while treatment versus control group comparison designs resulted in absolute magnitudes of treatment effects ranging from -0.14 -2.39 .

Changes in youth outcomes larger than 20% were found, irrespective of study design, for outcomes focused on psychosomatization (29% reduction), school attendance (25% increase), and stress (48% reduction). The magnitude of changes after psychotherapy ranged from 6% (externalizing problems) to 48% (stress). Several moderator variables significantly influenced psychotherapy treatment effect sizes, including frequency and length of treatment as well as treatment format.

However, results did not support the superiority of a single type of intervention for most outcomes. Implications for therapy with school-aged youth and future research are discussed.