Personalising outcome measurement in substance misuse treatment: the feasibility and psychometrics of two individualised outcome measures

Alves, Paula; Faisca, Luis; Sales, Célia M. D.; Ashworth, Mark

http://hdl.handle.net/10400.1/14762

Use this identifier to reference this record.

Name:	Description:	Size:	Format:
Alves2020_Article_PersonalisingOutcomeMeasuremen.pdf		366.37 KB	Adobe PDF	Download

Send Feedback

Authors

Abstract(s)

Individualised information in substance misuse treatment complements standardised outcome measures. However, few studies investigate the use of individualised measures and their robustness in terms of quantifying outcomes. In this study, we analysed the psychometrics and feasibility of two individualised outcome measures (PQ and PSYCHLOPS). We followed a cross-sectional methodology, administering the individualised measures and three additional standardised measures (TOP, a measure of psychological health within addiction services; PHQ-9; CORE-OM) to a sample of 93 patients entering substance misuse treatment in four clinical services. The results showed high levels of patient acceptability of the two individualised measures (response rates > 95%). The internal reliability was good for both PQ and PSYCHLOPS (Cronbach’s alpha, .79 and .72, respectively). Convergent validity of PQ with standardised measures was weak: Pearson’s r values for TOP (psychological health), PHQ-9 and CORE-OM were .21, .22 and .27, respectively. In contrast, convergent validity of PSYCHLOPS was moderate: r = .40, .39 and .50, respectively. Convergence between PQ and PSYCHLOPS was weak (r = .28). Experience of previous treatment episodes was associated with higher PQ and PSYCHLOPS scores; PSYCHLOPS but not PQ scores were higher among those opting to complete the questionnaires in written rather than verbal format. Our findings demonstrated that PQ and PSYCHLOPS are reliable and feasible individualised outcome measures for use in substance misuse treatment units, although the lack of strong convergent validity indicates that they may be measuring different underlying constructs. Optimal outcome measurement may involve combining individualised and standardised measures.