Objective: 1. To determine whether naltrexone-accelerated detoxification with minimal sedation is an acceptable and effective form of induction onto naltrexone. 2. To monitor outcomes of detoxified patients. Design: Observational study. Setting: Medical ward of a general hospital (for detoxification) and a community clinic (for follow-up) in Sydney, NSW, 1998. Patients: 15 heroin users and 15 people seeking withdrawal from methadone. Intervention: Detoxification used naltrexone (12.5 or 50 mg), with flunitrazepam (2-3 mg), clonidine (150-750 mu g) and octreotide (300 mu g) for symptomatic support. Patients remained awake and were discharged when they felt well enough. Follow-up was daily for four days and then weekly for up to three months for supportive care. Main outcome measures: Acute side effects; patient ratings of severity and acceptability of withdrawal; nights of hospitalisation; rates of induction onto naltrexone; retention in treatment over three months; and relapse to opioid use. Results: Acute withdrawal with delirium lasted about four hours. Octreotide was crucial for controlling vomiting; with octreotide no patient required intravenous fluids. There were no major complications. Eighteen patients (60%) reported that it was a "quite" acceptable procedure, 18 (60%) required only one night's hospitalisation, and 24 (80%) were successfully inducted onto naltrexone (defined as taking naltrexone on Day 8). Three months later, six (20%) were still taking naltrexone (with four of these occasionally using heroin) and seven (23%) were abstinent from opioids, including five not taking naltrexone. Eleven had gone onto methadone maintenance, seven had relapsed to heroin use, and one had died of a heroin overdose. Conclusions: Rates of induction onto naltrexone were comparable with those reported for accelerated detoxification under sedation, suggesting that it can be performed successfully with minimal sedation. As in other studies of naltrexone maintenance, retention was low, and relapse to heroin use was common.
