S27 – Adaptations to Contingency Management for Alcohol Use Do Not Increase Efficacy in Individuals who Engage in Pre-Treatment Drinking Consistent with Non-Response

Aim: To determine if adaptations to contingency management (CM) led to increased alcohol abstinence among individuals entering treatment with a pre-treatment urine-ethyl glucuronide (uEtG) alcohol biomarker level shown previously to predict CM-non-response (uEtG >350 ng/mL).

Methods: 307 individuals who met criteria for a serious mental illness and moderate to severe alcohol use disorder submitted urine samples twice weekly over a four-week pre-treatment phase. 158 individuals who had a mean uEtG level ( >350 ng/mL) over the pre-treatment phase were randomized to 16 weeks of treatment-as-usual and one of three prize-draw CM conditions: (1) usual CM (reinforcement criteria: uEtG <150 ng/mL, consistent with alcohol abstinence, $1,686 maximum earnings); (2) high magnitude CM (reinforcement criteria: uEtG <150 ng/mL, $2,982 maximum earnings); or (3) shaping CM (reinforcement criteria: weeks 1-4 uEtG <500 ng/mL, weeks 5-16 uEtG <150 ng/mL, $1,686 maximum earnings). Generalized estimating equations were used to examine differences in uEtG-defined alcohol abstinence between groups over the pre-treatment and treatment periods separately, and differences between the pre-treatment and treatment periods when the groups were pooled.

Results: There were no group differences in uEtG-defined alcohol abstinence during the pre-treatment (p=0.61) or treatment periods (p=0.36). During the treatment period participants had four times higher odds of being abstinent (estimated marginal mean=60% [SD 24%]) than during the pre-treatment period (estimated marginal mean=27% [SD=26%]; OR=4.23 [CI=3.02-5.92], p <0.01).

Conclusions: There were no differences in uEtG-defined alcohol abstinence across groups during CM treatment; however, uEtG-defined alcohol abstinence was higher during treatment, relative to the pre-treatment period. These findings are noteworthy because this population was selected due to a level of pre-treatment uEtG-documented drinking that was consistent with CM non-response in a past study. A lack of group differences could be attributed to the usual CM intervention having a $403 higher maximum reinforcement magnitude, relative to the CM intervention that provided in this prior study.