JMIR Publications

ABSTRACT

Background:

Gambling disorder carries a disproportionate burden of suicide risk, with a lifetime prevalence of suicidal ideation (SI) of 31% in this population. Clinicians frequently encounter patients with SI at treatment entry, yet no published study has examined whether SI at intake predicts treatment engagement or symptom change in gambling disorder care. Whether patients with SI engage in treatment and improve comparably to those without SI has direct implications for clinical decision-making, triage, and care planning.

Objective:

The purpose of this study was to determine: (1) the prevalence of SI at treatment entry among adults seeking care for gambling disorder; (2) how clinical and demographic characteristics differ between clients with and without SI at intake; (3) if SI at intake predicts differential treatment engagement; (4) if SI at intake predicts differential improvement in gambling-related symptom severity over the course of treatment.

Methods:

This retrospective cohort study used de-identified clinical data from Birches Health, a U.S.-based digital health platform delivering care for behavioral and process addictions. Eligible patients entered care for gambling disorder, completed at least one session, and had baseline PHQ-9 data available (n=1,518). SI was defined as any score ≥1 on PHQ-9 item 9. Baseline characteristics were compared by SI status using t-tests and chi-square tests. Treatment engagement was analyzed using Wilcoxon rank-sum tests and negative binomial regression. Gambling symptom severity trajectories were modeled using linear mixed-effects models with a SI-by-time interaction as the key predictor, adjusting for baseline gambling severity (G-SAS), depression severity (PHQ-8), age, and gender.

Results:

SI at intake was endorsed by 465 patients (30.6%). Patients with SI presented with significantly higher depression severity (PHQ-8 mean 15.2 vs. 8.7, P < .001), higher gambling severity (G-SAS mean 22.7 vs. 19.1, p < .001), and higher rates of depressive disorder diagnosis (28.7% vs. 20.7%, P = .002), but did not differ on age, gender, or other psychiatric comorbidities. After adjustment for baseline characteristics, SI did not predict total appointments attended (IRR = 1.05, 95% CI [0.92, 1.19], P = .475). Gambling symptom severity declined significantly over time across the full sample (b = -6.38, P < .001), and SI status at intake did not moderate the rate of improvement (b = 0.27, P = .765).

Conclusions:

SI at intake is common among adults entering treatment for GD and identifies a clinically distinct subgroup defined by greater psychiatric burden. These findings suggest that SI at intake should not be considered a predictor of poor treatment adherence or response.