Category: Addictive Behaviors
Dropout poses a significant problem when treating disordered gamblers. A growing body of research has investigated when treatment dropout occurs. However, the session-by-session rates reported in these studies may be misrepresented by their duration-based operationalizations because they do not account for disordered gamblers’ symptom improvement. The purpose of the present study was to examine session-by-session dropout rates using an operationalization of reliable change (Swift, Callahan, & Levine, 2009) among a sample of treatment-seeking disordered gamblers.
Participants were composed of outpatients (n = 334) who sought treatment for gambling-related problems at a private practice between 1998 and 2008. The mean age was 45.92 (SD = 11.47), and 61% were women. The majority (83%) were Caucasian, 8% Hispanic American, 3% Asian American, 3% African American, and 3% were other races or multiracial.
At intake, all participants completed the BDI-II and a structured interview for the diagnostic criteria for disordered gambling. Prior to each subsequent treatment session, participants were also required to take the BDI-II to track general symptom improvement. The average BDI-II score at intake was 24.12 (SD = 12.49), and 97% met diagnostic criteria for disordered gambling. The mean number of sessions attended was 4.91 (SD = 3.70; range = 1-17).
We conducted a discrete-time survival analysis to examine rates of dropout and symptom improvement over the course of treatment. Clients who failed to achieve reliable change on the BDI-II were considered treatment dropouts. Overall, 48% (n = 160) of clients dropped out of treatment, and 52% (n = 174) reliably improved in their psychological symptoms. The largest number of participants dropped out of treatment after completing the initial intake and before session 2 (n = 55) and after completing session two and before session 3 (n = 36). Twenty-one clients (6%) achieved reliable change during this same period. Dropout rates remained generally constant through session 7 and plateaued by session 8. Rates of reliable change were constant through all 17 sessions.
Our results indicated that dropout rates remained highest in the early stages of treatment, even when considering disordered gamblers who improved in their psychological symptoms. A small minority will evidence reliable improvement in psychological symptoms during that time. Future research should investigate characteristics of disordered gamblers who drop out early versus late in treatment.