Utility of Collateral Informants to Inform Treatment for Gambling Disorder Proceedings (Faculty180)

cited authors

  • Petra, Megan M; Cunningham-Williams, Renee M


  • Background: Disordered Gambling (DG) occurs in ∼1–2% of the US population, but is 6.5 times more likely in those with sub-stance use disorders (SUDs). DG is now classified with SUDs as a DSM-5 addictive disorder, thus better positioning clinicians to screen and treat DG along with SUDs. Though accurate information about gambling behavior is critical to DG diagnosis and treatment-planning, no biological test exists as a “gold standard” in this process. If clinicians knew that collateral informants (CIs)—gamblers’ friends/family who can report on their behavior—could provide ac-curate (concordant) information, they could use CIs’ verification to inform diagnostic and treatment decisions. Objective: Investigate concordance between gambler self-reports and CI reports. Deter-mine if concordance is associated with gambler-collateral relation-ship or influenced by gamblers’ comorbid SUDs or psychiatric disorders. Methods: We used community-recruited adults who had gambled at least five times in their lives for a psychometric study of a computerized diagnostic interview (C-Gam©). Gamblers and CIs were interviewed separately by telephone. Measures included DSM criteria for DG (C-GAM© & GAM-CI©), SUDs (GAM-DA©), depression (CES-D), and personality disorders (SCID-II). Concordance is provided by Cohen’s kappa and interclass correlation coefficient (ICC), with comparisons made by Fisher’s Z transformation. Kappa and ICC are considered fair (0–.2), poor (.2–.4), moderate (.4–.6), substantial (.6–.8), or almost perfect (.8–1.0). Results: Concordance was substantial (.62) for number of DSM-5 DG criteria endorsed and moderate for DG diagnosis (.44). Family members achieve higher concordance (.62) than friends (.24) on DG diagnosis (p < .05), but concordance was unaffected by gamblers’ comorbid SUDs, depression, or personality disorders (p > .05). Conclusions: Treatment providers should use CIs to verify gambler self-reports, as concordance is likely to be moderate to substantial and is unaffected by gamblers’ comorbid conditions. Family members are better to use as CIs than are friends. CIs are a valuable source of information which clinicians can use to inform DG diagnostic and treatment decisions.


publication date

  • 2014

presented at event

start page

  • 194

end page

  • 216


  • 35