Methadone use for acute opioid withdrawal in Tshwane shelters during the COVID-19 lockdown

Abstract

Background: Temporary shelters were established for street-based people during the national level 5 coronavirus disease 2019 (COVID-19) lockdown. However, street-based substance users’ need to access substances was not addressed, resulting in large numbers of people experiencing withdrawal. The Community Oriented Substance Use Programme (COSUP) in Tshwane provided methadone to manage opioid withdrawal. Methods: A cross-sectional, descriptive study was conducted using the daily methadone dosing records from shelters in Tshwane between March 2020 and September 2020. Results: The final analysis included 495 participants, of which 64 (12.9%) were initiated on 20 mg – 30 mg of methadone, 397 (80.2%) on 40 mg – 50 mg, and 34 (6.9%) on 60 mg – 70 mg. A total of 194 (39.2%) participants continued their initiation dose for 1–2 months, after which 126 (64.9%) had their doses increased, and 68 (35.1%) had their doses decreased. Approximately 12 (2.4%) participants were weaned off methadone after 1–3 months and 46 (9.3%) after 4–6 months. In all, 100 (20.2%) participants left the shelter prematurely and did not continue with methadone. A total of 126 (25.5%) participants continued to stay in the shelters and received methadone for 6 months, with 125 (25.3%) participants leaving the shelter with continued follow-up at a COSUP site. Conclusion: This study demonstrates variability in methadone dosing regimens among shelter residents. As the lockdown measures eased, many chose to leave the shelters, while others remained to receive methadone and other services. The COSUP appears to be effective during periods of increased vulnerability, since a large number of participants were successfully followed up. Contribution: Opioid dependence is a persistent, lifelong disease. It is multifaceted with complex environmental and individual determinants. This study highlighted the use of opioid substitution therapy during a period of increased vulnerability.