Regina Murphy1, Paul Hutton2. 1. Child & Adolescent Mental Health Service, Kirkwall Health Centre, Kirkwall, UK. 2. School of Health & Social Care, Edinburgh Napier University, Edinburgh, UK.
Abstract
BACKGROUND: Previous meta-analyses have only found small correlations (r = .10 to r = .19) between therapeutic alliance and clinical outcomes in samples of adolescents receiving psychological therapy. Although study-level variables have been found to moderate this, little is known about the impact of therapist variability. The present meta-analysis aimed to address this gap by using patient-therapist ratio as a moderator variable. METHODS: Contrary to previous reviews of adolescent alliance, individual effect sizes were extracted using a preregistered conceptual hierarchy. Controlling for treatment-level confounds, a random effects meta-analysis assessed the moderating effect of patient-therapist ratio on the alliance-outcome relationship in predefined single-predictor and multipredictor meta-regressions. RESULTS: The alliance-outcome relationship was found to be larger than previously thought (k = 28, N = 2,911, r = .29, 95% Confidence Interval 0.21, 0.37; p < .0001, I2 = 80%). When study samples exceeding the adolescent 12-19 age range were removed, the correlation rose (k = 15, N = 1,797, r = .34, 95% Confidence Interval 0.23, 0.45; p < .0001, I2 = 83%). In contrast to research with adults, patient-therapist ratio did not moderate this relationship in either single-predictor (p = .26) or multi-predictor (p = .22) models. CONCLUSIONS: The alliance-outcome relationship for adolescents was larger than previously thought, and comparable to estimates in adult samples. The failure of patient-therapist ratio to moderate its strength, however, challenges the hypothesis that variability in therapist characteristics is an important determinant of the alliance-outcome effect in this age group.
BACKGROUND: Previous meta-analyses have only found small correlations (r = .10 to r = .19) between therapeutic alliance and clinical outcomes in samples of adolescents receiving psychological therapy. Although study-level variables have been found to moderate this, little is known about the impact of therapist variability. The present meta-analysis aimed to address this gap by using patient-therapist ratio as a moderator variable. METHODS: Contrary to previous reviews of adolescent alliance, individual effect sizes were extracted using a preregistered conceptual hierarchy. Controlling for treatment-level confounds, a random effects meta-analysis assessed the moderating effect of patient-therapist ratio on the alliance-outcome relationship in predefined single-predictor and multipredictor meta-regressions. RESULTS: The alliance-outcome relationship was found to be larger than previously thought (k = 28, N = 2,911, r = .29, 95% Confidence Interval 0.21, 0.37; p < .0001, I2 = 80%). When study samples exceeding the adolescent 12-19 age range were removed, the correlation rose (k = 15, N = 1,797, r = .34, 95% Confidence Interval 0.23, 0.45; p < .0001, I2 = 83%). In contrast to research with adults, patient-therapist ratio did not moderate this relationship in either single-predictor (p = .26) or multi-predictor (p = .22) models. CONCLUSIONS: The alliance-outcome relationship for adolescents was larger than previously thought, and comparable to estimates in adult samples. The failure of patient-therapist ratio to moderate its strength, however, challenges the hypothesis that variability in therapist characteristics is an important determinant of the alliance-outcome effect in this age group.
Authors: Ruben G Martinez; Ilana Seager van Dyk; Juliet L Kroll; Natacha D Emerson; Brenda Bursch Journal: Evid Based Pract Child Adolesc Ment Health Date: 2021-10-22
Authors: Frank Doyle; Karen Morgan; Mary Mathew; Princy Palatty; Prashanti Kamat; Sally Doherty; Jody Quigley; Josh Henderson; Ronan O'Carroll Journal: Int J Environ Res Public Health Date: 2019-04-11 Impact factor: 3.390
Authors: Lindsay H Dewa; Emma Lawrance; Lily Roberts; Ellie Brooks-Hall; Hutan Ashrafian; Gianluca Fontana; Paul Aylin Journal: J Med Internet Res Date: 2021-12-17 Impact factor: 5.428