Extended release injectable naltrexone before vs. after release: A randomized trial of opioid addicted persons who are in prison.

BACKGROUND

Usual treatment for persons with opioid use disorders who are in prison is detoxification with referral to treatment after release but failure to engage in treatment and relapse is common. Starting medication treatment before release might improve outcomes.

OBJECTIVES

Determine if administering extended-release injectable naltrexone (XR-NTX) before release (BR) from prison results in less relapse within the first three months after release than when offered by referral after release (AR).

METHODS

The study randomized 1:1 persons who had an OUD, expressed interest in XR-NTX, and met study admission criteria to receive XR-NTX BR or at a local program AR, with continued medication and counseling available at that program.

RESULTS

Four-hundred and two persons expressed interest in the study, 222 consented, and the study randomized 146. Uncertainty about release dates resulted in a time lag between randomization and final disposition during which 60 of the randomized patients were sentenced to other facilities, withdrew consent, or became otherwise unavailable for study treatment, leaving 86 for outcome analyses (38, BR; 48 AR). Missed follow-up appointments on the remaining 86 led to development of a phone-based questionnaire to determine presence/absence of relapse. Using it to supplement other data, we were able to confirm relapse or nonrelapse for 63 of the 86 (73%). All BR and a third of the AR patients received their first XR-NTX dose, however dropout was high and nonrelapse by month three was not significantly different between BR (39.5%) and AR (25%) (Chisq (2) = 3.23, p = 0.20).

CONCLUSIONS

BR patients were much more likely to receive medication and its extended relapse and overdose protection effects in the first weeks after release. Dropout was high and the study detected no significant difference in relapse by month 3; however, the less-than-planned number of patients and missing data make this finding inconclusive.