Testing adaptations to contingency management for alcohol use disorders: A randomized controlled trial

To determine if adults with an alcohol use disorder (AUD), who had a preintervention urine ethyl glucuronide (uEtG) level predictive of nonresponse to contingency management (CM), would respond to two intervention modifications (https://clinicaltrials.gov/ ID: NCT03481049).

One hundred fifty-eight adults (53.2% female) with AUD, serious mental illness, and a mean uEtG ≥ 350 ng/mL over a 4-week induction period were randomized to (a) usual CM (uEtG-negative [<150 ng/mL] samples reinforced with $1,686); (b) high magnitude CM (uEtG-negative samples reinforced with $2,983); or (c) shaping CM (reduced drinking [uEtG < 500 ng/mL] samples reinforced for 4 weeks, then uEtG-negative samples reinforced for 12 weeks with $1,686). The primary outcome was uEtG-negative samples during induction and Weeks 5-16 of CM. The relationship between outcomes and uEtG-defined heavy drinking (≥ 500 ng/mL) immediately prior to randomization was assessed.

CM conditions did not differ in uEtG-negative samples during the intervention period, Wald, χ²(2) = 1.96, p = .46. Participants were 4.2 times (95% CI [3.02, 5.92], p < .01) more likely to submit a uEtG-negative sample during CM, relative to induction. Those with a heavy drinking uEtG result immediately before randomization were less likely to submit uEtG-negative samples during CM, Wald, χ²(1) = 15.33, p < .01.

CM modifications were not associated with lower levels of alcohol use. Participants engaged in less alcohol use during CM, relative to induction. Two patterns of response to CM were observed based on uEtG-defined heavy drinking immediately prior to CM. (PsycInfo Database Record (c) 2025 APA, all rights reserved).