RESEARCH QUESTION
Can a brief screening measure detect problem and pathological gambling in clinical settings?
PURPOSE
Problem gambling often co-occurs with substance use problems and psychiatric disorders. Routine screening for problem gambling, substance misuse, and mental health issues by health professionals could assist with prevention and interventions efforts. A routinely used problem gambling screening tool needs to be short, easy to use, and accurate. One brief measure performed well in detecting problem gambling in the general population of adults. However, it was not clear whether the measure worked in clinical settings such as addiction programs. This research was performed to test a brief measure of problem gambling for use in clinical settings.
HYPOTHESES
None stated.
PARTICIPANTS
Participants 375 adults (average age = 42 years; 59% male) recruited from inner cities in Connecticut. All participants had gambled in the past 2 months and were part of a larger gambling intervention study.
PROCEDURE
Recruitment took place through advertisements at medical and substance use clinics. Through an interview, participants completed two measures of problem gambling. Then, the researchers examined how the brief measure performed at identifying problem and pathological gamblers as compared to the longer measures. The researchers also tested other short combinations of questions for success at assessing problem gambling risk.
MAIN OUTCOME MEASURES
Problem gambling severity was measured using the self-report South Oaks Gambling Screen (SOGS). Problem and pathological gambling was also measured using the self-report NODS scale, based on DSM-IV diagnostic criteria. The NODS-CLiP is a subset of NODS questions, and contains items about dependence, lying, and preoccupation with gambling. Other NODS items involve chasing losses, gambling to escape, and risked relationships.
KEY RESULTS
To be useful, the brief measure needed to have the sensitivity to correctly detect problem and pathological gamblers in clinical settings. The NODS-CLiP correctly identified 98% of problem and pathological gamblers. However, this brief measure also identified many participants who did not have serious gambling problems, with 43% of low risk gamblers and 77% of at risk gamblers mis-identified as problem gamblers. Instead of the NODS-CLiP items, a combination of NODS items related to chasing losses, preoccupation, risked relationships, and gambling to escape appeared to perform better this clinical sample (i.e., the NODS-PERC). This combination of items performed better across all demographic groups, and was better ability to correctly distinguish problem and non-problem gamblers.
LIMITATIONS
All participants were seeking help for their gambling problems and may not reflect the broader clinical population. In addition, this non-random sample was taken from a limited geographical area, and these findings cannot necessarily be generalized to gamblers living in other locations. The use of multiple gambling measures may have influenced the results, and more research is needed to examine how brief measures operate in isolation.
CONCLUSIONS
These findings suggest that brief measures of problem gambling can be useful in a clinical setting. However, more research is needed to confirm the best combination of questions to quickly screen for problem gambling. Both the NODS-CLiP and NODS-PERC performed relatively well at identifying problem gambling, but the NODS-CLiP may lead to higher false positive rates.
