The psychometric properties and treatment outcomes associated with two measures of youth therapeutic alliance using naturalistic data

Purpose: This is the second of two articles that examine the psychometric properties and treatment outcomes associated with two measures of the therapeutic alliance in naturalistic routine settings. Methods: Data were taken from the ACORN database for youth, and, where available, parent/carer, attending for psychotherapy treatment in naturalistic settings (N = 12,573). The sample, the largest to date, included only those completing both an alliance measure and an outcome measure at every session. Two sets of three different alliance items are used across the two populations in routine practice. Results: Analyses revealed that the alliance explained no more than 3% of the variance in outcomes. Alliance measures exhibit ceiling effects, and as such, drawing conclusions about correlations with outcomes can be difficult. Any drop in alliance score as rated by both youth and parent/carer is predictive of outcomes, with parent/carer ratings being marginally more predictive. Where the alliance is rated as better by youth or parent/carer in comparison with ratings as worse, effect sizes are up to 50% better for the youth. Conclusion: The therapeutic alliance remains an important non-specific treatment component; however, measures of the alliance have ceiling effects. Both youth and their parent/carer can provide important feedback to the therapist, and any drop in alliance is predictive of clinically meaningful change. As such, therapists should monitor the alliance with both youth and parent/carer. Implications for practice, training and research are considered.