Decision-making impairment predicts 3-month hair-indexed cocaine relapse.

One of the key outstanding challenges in cocaine dependence research is determining who is at risk of relapsing during treatment. We examined whether cognitive decision-making profiles predict objectively (hair) indexed cocaine relapse at 3-month follow-up. Thirty-three cocaine-dependent patients commencing outpatient treatment in a public clinic performed baseline decision-making assessments with the original and variant versions of the Iowa Gambling Task, and provided a 3-cm hair sample 3 months afterwards. Based on Iowa Gambling Tasks' performance cut-offs, 5 patients had intact decision-making skills, 17 patients showed impaired sensitivity to reward or punishment (impairment in one of the tasks), and 9 patients showed insensitivity to future consequences (impairment in both tasks). Based on a 0.3 ng/mg cocaine cut-off, 23 patients were classified as relapsers and 10 as non-relapsers at the 3-month follow-up. Eighty percent of patients with intact decision-making were abstinent at follow-up, whereas 90% of patients with insensitivity to future consequences had relapsed. The two subgroups (relapsers and non-relapsers) showed no significant differences on drug use, comorbidities, or psychosocial function, and significantly differed on verbal but not performance IQ. A regression model including decision-making scores and verbal IQ predicted abstinence status with high sensitivity (95%) and moderately high specificity (81%). These preliminary findings demonstrate that decision-making profiles are associated with cocaine relapse. Moreover, combined decision-making and IQ assessments provide optimal predictive values over stimulant relapse, yielding significant opportunities for clinical translation.