RESEARCH QUESTION
Does the NOR DSM-IV Screen for gambling problems (NODS) and a clinical interview classify gamblers in the same way?
PURPOSE
The scientific community must develop instruments that accurately evaluate the presence and the extent of gambling problems in a population. The purpose of this study is to verify whether the NORC DSM-IV Screen for gambling problems (NODS), an instrument that evaluates the presence of problem gambling, is concordant in comparison to a clinical interview approach and is a valid tool for estimating the prevalence of gambling problems in a population.
PARTICIPANTS
Participants were recruited from a pool of individuals who had taken part in previous studies. Two hundred French-speaking individuals (140 men, 60 women; average age 35) from Quebec participated. Participants were 18 years of age or older and had gambled on video lottery machines at least once during the prior 12 months.
PROCEDURE
Participants were first assessed using the NORC DSM-IV Screen for gambling problems (NODS) to evaluate whether gambling problems were present during the last 12 months. The participants were then interviewed by a clinician and categorized as either a gambler without a problem, an at-risk gambler, or a pathological gambler.
MAIN OUTCOME MEASURES
NORC DSM-IV Screen for gambling problems (NODS).This is an instrument developed in 1998 by the National Opinion Research Center for the National Gambling Impact Study Commission. It includes 17 questions designed to evaluate whether an individual had gambling problems during the past 12 months.
Clinical Interview. Based on the diagnostic criteria of the DSM-IV, the interview was used to assess the severity of each participant’s gambling behaviour and to investigate if the participant’s gambling was better explained by a manic episode.
KEY RESULTS
Results showed that 50 of the 200 participants were not categorized in the same way with the NODS and the clinical interview. Forty-seven participants were identified as having a more severe gambling problem according to the NODS than with the clinical interview (24% of the sample). Of the 71 participants who were identified as pathological gamblers with the NODS, 18 were at-risk gamblers according to the clinical interview and 3 were gamblers without a problem. Among the 66 participants that the NODS categorized as at-risk gamblers, the clinical interview identified 26 as gamblers without a problem and 3 were rated as pathological gamblers. All 63 participants who were identified by the NODS as being without a gambling problem were correctly classified.
The number of criteria endorsed using the NODS was significantly higher than with the clinical interview (NODS = 3.28, clinical interview = 2.38). Overall results showed that out of the 200 participants, 97 obtained a higher score on the NODS than with the clinical interview, 19 obtained a lower score, and 84 participants obtained an equivalent score.
Because the clinical interview is known as the most reliable tool to diagnose problem gambling, results from this study suggest the 10 item NODS does not reliably classify PGs (25% of this sample were incorrectly diagnosed).
LIMITATIONS
The NODS asks dichotomous questions (e.g., “yes” or “no”), leaving participants little choice when rating the criteria. This may have contributed to the discrepancy between classifications of gamblers according to the NODS and the clinical interview. The maximum score on the NODS is 10, meaning that one misunderstood question is sufficient to incorrectly categorize an individual. This could be rectified by increasing the number of questions, resulting in less of an impact on classification when wrongly endorsing a question.
CONCLUSIONS
The two modes of evaluation used for this study provided different classifications and cannot be deemed comparable. It is possible that both the NODS and the clinical interview may be tapping two independent factors (dependence and gambling-related problems). Of the 71 individuals who had been screened by the NODS as pathological gamblers, only 50 were also diagnosed in the same way according to the clinical interview. It appears as though there is currently no single screening questionnaire in existence that adequately reflects the multi-dimensional nature of problem gambling (note this article was published in 2005).
