Reducing the risks of opioid detox

When discussing the goal of abstinence for opioid use disorder, it sometimes comes up that it’s much safer to stay in medication assisted treatment (most often methadone or buprenorphine) than to detox. I agree, but I would never advise a patient just to detox. Detox is a procedure, not a treatment as such. If all you do is offer detox, or it’s the only part of the package the client will take, then the outcomes are not only bound to be poor, they are also bound to be fraught with danger. Detox is not enough.

We need to exercise caution with such requests – relapse rates are very high after detox. Loss of tolerance to opioids occurs quickly and relapse then leads to risk of overdose and death. There can be a reluctance or even refusal to look at supporting people to move on from MAT. But here are some considerations for clinicians:

  • In the UK there are thousands of people in long term recovery from opiate dependence, so clearly some people have managed this safely
  • There are people on MAT who want to try to move on
  • In some areas of Scotland this option is already well established

In my experience, some (if not many) patients want a stand-alone detox and have unrealistic expectations of the outcome and discount the risks. Detox may be part of a process that leads to healing, but detox itself does not heal. Thinking of detox in isolation from the other factors that promote recovery, and simultaneously reduce risk of relapse, is not a good idea.

So what are the outcomes from detox?

A study from Geneva of 73[i]  patients with dependent opioid use followed up at one and then six months after detox found 35% abstinent at one month and 37% abstinent at six months. Residential treatment following detox was associated with increased likelihood of abstinence. Not all the lapsers/relapsers were using dependently again, but the relapse group was clearly significant in numbers. Using cocaine after detox was a risk factor for relapse.

In a 14-month follow-up study from Ireland[2] involving 143 patients detoxed from opiates, the participants were divided into three treatment types: no formal aftercare; outpatient aftercare and residential rehabilitation. The average methadone dose prior to detox was around 77mg in the aftercare groups and 69mg in the no formal aftercare group. The patient group were found to be representative of opiate replacement patients generally in Ireland. All engaged in prior preparatory work. The primary outcome was abstinence. They had a good follow-up rate of 75%.

The residential group had the lowest relapse rate. Those participants who chose outpatient aftercare relapsed at a 52% higher rate than the residential patients. The no formal aftercare group relapsed most and fastest. Interestingly, the intention to attend residential treatment post-detox had a statistically significant effect on abstinence. In the longer term the differences between the outpatient aftercare group relapse rate and the residential treatment aftercare group relapse rate began to close raising questions about cost effectiveness for the authors

The authors make the point that even with aftercare, patients were more likely to lapse/relapse than stay abstinent which raises concerns for safety. No deaths or overdoses were reported in the paper, though not everyone was followed up. The authors call for risks for lapsers and relapsers to be managed with appropriate supports.

This study both supports the proposition that achieving abstinence following opiate detox is an achievable goal for some and at the same time represents a risk for others. More than half of those in the group had undergone detox before, suggesting for some that several attempts may have to be made – again seemingly increasing the risk.

I believe there are ways to decrease the risk. Longer times in residential treatment are associated with better outcomes (some of the sample completed only 8 weeks, which is probably an insufficient ‘dose’). There is no mention in the paper of mutual aid which can have significant mitigating effects through social networking using assertive linkage to mutual aid groups like NA, CA and SMART Recovery. Nor was there reference to family therapy or even simply involving families in planning and support.

It’s also not clear what harm reduction advice was given at the outset, and again during treatment and on discharge (overdose prevention, resuscitation training, take home naloxone etc.). Nor is it clear what pathways were available for early re-entry into MAT for those who had relapsed or what was the availability of re-titration for those who wanted to leave early from treatment. The quality, intensity and duration of aftercare, and whether mental health and psychological issues are addressed in treatment and in aftercare, are likely to be important variables which could reduce risk.

I think that harm reduction interventions like overdose prevention, take-home naloxone and early MAT re-entry on relapse are crucial in reducing risk, but that the aforementioned psychosocial elements of recovery management may well be as important in risk mitigation. We have a lot of knowledge and faith in the former, but need to really work on the latter.

Detox from opioids should only normally be offered in the context of a robust treatment and aftercare package and not as a stand-alone intervention. We would not offer chemotherapy to patients on the basis of removing some of the elements that promote success and then offering no follow-up, nor would we neglect the important psychosocial factors that reduce risk of recurrence. Treatment of opioid use disorder deserves at least the same level of care.

Continue the discussion on Twitter: DocDavidM

[i] Broers B, Giner F, Dumont P, Mino A. Inpatient opiate detoxification in Geneva: follow-up at 1 and 6 months. Drug Alcohol Depend. 2000 Feb 1;58(1-2):85-92. doi: 10.1016/s0376-8716(99)00063-0. PMID: 10669058.


[2] Ivers JH, Zgaga L, Sweeney B, Keenan E, Darker C, Smyth BP, Barry J., 2018. A naturalistic longitudinal analysis of post-detoxification outcomes in opioid-dependent patients. Drug Alcohol Rev. 2018 Apr;37 Suppl 1:S339-S347.