Skip to main content
. 2023 Jul 26;39(4):1699–1721. doi: 10.1007/s10899-023-10240-z

Table 2.

Screening tools

Tool Items Purpose Setting Study Outcome
4Es 40 items Predict gambling risk based on factors of Escape, Esteem, Excess and Excitement; distinguish gambling risk from alcohol harms Non-treatment seeking Rockloff and Dyer, 2006 In sample of 2,577 (Study 2): Escape, excitement and excess independently predict PGSI gambling problems. Escape and excess independently distinguish gambling from substance use.
The Alfred Screening Tool for Problem Gambling 4 items Screen for problem gambling (self or other) in mental health service Mental health de Castella, 2011 Identified gambling harms in mental health treatment seeking population
ASI-G 5 items Assess severity of gambling problems Gambling treatment; opioid use treatment Petry, 2003 In sample of 598: adequate to good Cronbach’s α (0.90); significant correlations between ASI-G scores and other indices (SOGS, DSM, TLFB)
BBGS 3 items Screen for gambling disorder (DSM-5) Methadone maintenance treatment (MMT) outpatient clinic Himelhoch et al., 2015 In sample of 300: sensitivity (0.909); specificity (0.865); positive predictive value (0.821); negative predictive value (0.933)
Substance use treatment Rowe et al., 2015 Inclusion in treatment manual for ATOD workers
CHI-T 15 items Measure transdiagnostic compulsivity Non-treatment seeking Chamberlain and Grant, 2018 The CHI-T had good convergent validity, with total scores correlating significantly with gambling disorder symptoms and with obsessive-compulsive symptoms. CHI-T total scores were also significantly elevated in participants who had a current substance use disorder versus those who did not.
CHAT 9 risk factors Develop multi-item general practice tool Primary health Goodyear-Smith et al., 2004 In sample of 2,543 patients across 20 GP clinics, 2.8% self-assessed drug use harm, 3.2% gambling harm and 10.8% alcohol harm.
Screen for health risk factors for follow-up in primary care Goodyear-Smith et al., 2009 In sample of 755: alcohol: sensitivity 80%, specificity 85%; drug use: sensitivity 64%, specificity 98%; gambling: sensitivity 80%, specificity 98%.
Testing of online eCHAT tool in primary care Goodyear-Smith et al., 2013 In sample of 196 patients: 97% found iPad-based eCHAT easy to use; 9% had concerns about data privacy.
Testing of CHAT tool in veteran health care Goodyear-Smith et al., 2021

In a sample of 34: VeCHAT tool proved acceptable to veterans and

Veterans’ Affairs staff

Feasibility study of the CHAT Elley et al., 2014 In sample of 107: CHAT found to be acceptable. 2 objections to alcohol questions; no other objections recorded.
EIGHT 8 items Screen for pathological gambling or problem gambling (DSM-IV) Substance use treatment Sullivan, 2007 In sample of 676 in AOD setting: high correlation with SOGS (83.9%)
GDSQ-P 27 items Screen for gambling disorder (DSM-5) Substance use treatment (residential) Maarefvand et al., 2019 In sample of 503: high sensitivity (0.99); specificity (0.98) and accuracy (0.98)
ICB 33 items Measure impulsive-compulsive behaviors Non-treatment seeking Guo et al., 2017 In sample of 687: Impulsive-Compulsions and Compulsive-Impulsions yielded very good Cronbach’s α of 0.89 and 0.84 respectively. The ICB Checklist is best utilised by examining each item endorsed and corresponding severity rating. The total sum of each subscale may indicate whether there is a stronger inclination towards impulsive or compulsive behaviors.
Lie/Bet 2 items Screen for pathological gambling (DSM-IV) MMT outpatient clinic Himelhoch et al., 2015 In sample of 300: sensitivity (0.942); specificity (0.657); positive predictive value (0.651); negative predictive value (0.944)
mASI 27 items (approx. 1 h completion) Secondary screen or assess for gambling problems and substance use problems Substance use/ non-substance use treatment (outpatient) Denis et al., 2016 In sample of 833: The Cronbach’s α ranged from 0.63 to 0.87 and could be considered good for medical, alcohol, and gambling domains; acceptable for employment/ support, drug, tobacco, and psychiatric domains; and questionable for legal and family /social domains.
NODS 17 items Screen for risky gambling, harmful gambling, and pathological gambling (DSM-IV) Substance use treatment Wickwire et al., 2008 In sample of 157: good Cronbach’s α (0.88); positive correlation with SOGS (r = .85, p < .001)
NODS-CLiP 3 items Screen for Loss of Control + Lying + Preoccupation items of NODS MMT outpatient clinic Himelhoch et al., 2015 In sample of 300: sensitivity (1); specificity (0.539); positive predictive value (0.596); negative predictive value (1)
Individuals recruited through substance use settings Volberg et al., 2011 In sample of 375: 3-item NODS-CPR had higher diagnostic efficiency (90.1% compared to 86.4% for NODS-CLiP)
NODS-PERC 4 items Screen for Preoccupation + Escape + Risked Relationships + Chasing items of NODS MMT outpatient clinic Himelhoch et al., 2015 In sample of 300: sensitivity (1); specificity (0.573); positive predictive value (0.614); negative predictive value (1)
PGSI 9 items Screen non-problem gambler, low-risk gambler, moderate-risk gambler, problem gambler (self-assess) Substance use treatment Rowe et al., 2015 Inclusion in treatment manual for ATOD workers
SPQ 160 questions (10 × 16 domains) Assess addictive behaviors Addiction treatment (residential) Christo et al., 2003 In sample of 497 (clinical)/ 508 (non-clinical): SPQ alcohol had high correlation with CAGE, SADQ, SMAST; SPQ recreation drugs scale had strong correlation with SODQ and SDS; SPQ prescription drugs had correlation with 6/8 validated scales; SPQ gambling had strong correlation with SOGS and correlation with 2 validated drug use scales. Cronbach’s α coefficient was high (between 0.82 and 0.98) across subscales.
University students MacLaren and Best, 2010 In sample of 948 university students: adequate inter-item reliability using Cronbach’s α (between 0.81 and 0.96).
SSBA

40 questions

(4 × 10 domains)

Screen for addiction problems Non-treatment seeking Schluter et al., 2018 In sample of 6,000: AUC values were moderate-to-high demonstrating overall ability of each subscale to discriminate between individuals who did/ did not self-report problematic engagement in a target behavior
SSOGS 7 items Screen for pathological gambling (DSM-III) Substance use treatment (residential) Nelson and Oehlert, 2008 In sample of 316: the SOGS internal consistency (coefficient α) for all items was 0.91. The 7-item SSOGS data yielded an internal consistency of 0.79.
Vices’ questionnaire

44 questions

(22 × 2)

Screen for drug and addictive behaviors (‘vices’); distinguish ‘wanting’ from ‘liking’ Non-treatment seeking Dale et al., 2016 In sample of 479: the ‘liking’/’wanting’ version of the survey had higher reliability and validity than the simpler ‘desire’ assessment.