Enhancing the utility of the problem gambling severity index in clinical settings: Identifying refined categories within the problem gambling category

• Published in: Addictive behaviors 2020-04, Vol.103, p.106257-106257, Article 106257
• Main Authors: Merkouris, SS, Greenwood, C., Manning, V., Oakes, J., Rodda, S., Lubman, D., Dowling, NA
• Format: Article
• Language: English
• Subjects: Adolescent; Adult; Australia; Behavior, Addictive - psychology; Female; Gambling - psychology; Gambling services; Help-seeking; Help-Seeking Behavior; Humans; Latent Class Analysis; Male; PGSI; Problem gambling; Problem gambling severity index; Psychometrics - instrumentation; Reproducibility of Results; Severity of Illness Index; Validity; Young Adult
• ISSN: 0306-4603; 1873-6327
• DOI: 10.1016/j.addbeh.2019.106257

•Refined PGSI problem gambling categories enhance its utility in clinical settings.•PGSI cut-off of 19 distinguishes between low and higher problem gambling severity.•The categories were associated with G-SAS, expenditure and face-to-face help-seeking.•Assist gambling services with treatment planning and improved client outcomes.•Further research required to validate these refined problem gambling categories. The Problem Gambling Severity Index (PGSI) was intended for use in epidemiological research with gamblers across the continuum of risk. Its utility within clinical settings, where the majority of clients are problem gamblers, has been brought into question. (1) Identify refined categories for the problem gambling category of the PGSI in help-seeking gamblers; (2) Validate these categories using the Gambling Symptom Assessment Scale (G-SAS); (3) Explore the relationship of these categories with indices of gambling and help-seeking behaviour. Secondary data analysis of help-seeking problem gamblers from the Australian online gambling counselling/support service (Gambling Help Online [GHO]) from October 2012 to December 2015 (n = 5,881) and trial data evaluating an Australian online self-directed program for gambling (GamblingLess; n = 198). Both datasets included the PGSI, gambling frequency and expenditure. The GamblingLess dataset also included the G-SAS and help-seeking behaviour. A Latent Class Analysis, using GHO data, identified a 2-class solution. Multiple analytical methods identified a cut-off value of ≥ 19 distinguishing this 2-class solution (low problem severity: Median = 16; high problem severity: Median = 23). High problem severity gamblers had increased odds of being categorised in the higher GSAS category, greater gambling expenditure and having sought face-to-face support. The refined categories were not associated with gambling frequency, distance-based or self-directed help-seeking. These findings are consistent with a stepped-care approach, whereby individuals with higher severity may be better suited to more intensive interventions and individuals with lower severity could commence with less intensive interventions and step-up to intensive interventions.