Examining concurrent validity and item selection of the Session Wants and Needs Outcome Measure (SWAN-OM) in a children and young people web-based therapy service.

Abstract

Single-session mental health interventions are frequently attended by children and young people (CYP) in both web-based and face-to-face therapy settings. The Session "Wants" and "Needs" Outcome Measure (SWAN-OM) is an instrument developed in a web-based therapy service to overcome the challenges of collecting outcomes and experiences of single-session therapies (SSTs). It provides pre-defined goals for the session, selected by the young person prior to the intervention, on which progress toward achievement is scored at the end of the session. The objective of this study was to evaluate the instrument's psychometric properties, including concurrent validity against three other frequently used outcome and experience measures, at a web-based and text-based mental health service. The SWAN-OM was administered for a period of 6 months to 1,401 CYP (aged 10-32 years; 79.3% white; 77.59% female) accessing SST on a web-based service. Item correlations with comparator measures and hierarchical logistic regressions to predict item selection were calculated for concurrent validity and psychometric exploration. The most frequently selected items were "Feel better" (N = 431; 11.61%) and "Find ways I can help myself" (N = 411; 11.07%); unpopular items were "Feel safe in my relationships" (N = 53; 1.43%) and "Learn the steps to achieve something I want" (N = 58; 1.56%). The SWAN-OM was significantly correlated with the Experience of Service Questionnaire, particularly the item "Feel better" [rs(109) = 0.48, p < 0.001], the Youth Counseling Impact Scale, particularly the item "Learn the steps to achieve something I want" [rs(22) = 0.76, p < 0.001], and the Positive and Negative Affect Schedule, particularly the items "Learn how to feel better" [rs(22) = 0.72, p < 0.001] and "Explore how I feel" [rs(70) = -0.44, p < 0.001]. The SWAN-OM demonstrates good concurrent validity with common measures of outcome and experience. Analysis suggests that lesser-endorsed items may be removed in future iterations of the measure to improve functionality. Future research is required to explore SWAN-OM's potential to measure meaningful change in a range of therapeutic settings.