Families tend to relax when someone goes into treatment and worry again if something goes wrong months later. The risk profile runs the other way.
The transition is the dangerous part, measurably
Sordo and colleagues (2017) measured mortality among people receiving opioid substitution treatment, and found risk is not spread evenly across an episode. It is markedly higher in the first four weeks of treatment than across the rest of it — and higher again in the first four weeks after leaving treatment than later on.
Both transitions are dangerous. The second one is the one families are least prepared for, because it looks like the end of the problem.
Scott and colleagues (2005) measured how unstable that period is in more ordinary terms. Following 448 people with quarterly interviews across two years, they found about a third moved between relapse, treatment re-entry and recovery in any given quarter. Over the two years, 82% changed state at least once and 62% more than once.
That is not a population settling into a stable outcome. It is a population in continual motion — which means a single reassuring phone call in month two tells you very little about month five.
The mechanism: stress, not weakness
Sinha’s work on stress and addiction explains a good deal of why the transition specifically is hard.
Stress-induced craving measured in a laboratory significantly predicted how soon someone relapsed afterwards. Not their stated intentions, not their insight, not how well they did in treatment — their physiological response to stress.
More usefully, stress and drug-cue exposure appear to produce the same state: rising distress plus a compulsive pull toward the substance. It is that state which carries the relapse risk, regardless of what triggered it. And the pattern holds across substances — alcohol, cocaine, nicotine.
Now consider what leaving treatment involves. Returning to the house, the commute, the job, the relationship and the specific difficult hour of the evening — every environmental cue, restored at once — while simultaneously losing the routine, the company and the supervision that were absorbing the stress.
Maximum cue exposure and minimum support, arriving on the same morning. The research would predict exactly the risk profile we observe.
Why the plan usually fails at exactly this point
Because it is not really a plan.
Most discharge plans consist of a weekly outpatient appointment, a meeting suggestion and a phone number. Between those points the person has roughly 165 hours a week alone with the situation, and the support that exists is available rather than present — it requires them to recognise they are in trouble and choose to reach out, using judgment that is not at its best.
Scott’s team tested the alternative: proactive check-ins at intervals rather than waiting for someone to re-present in crisis. Those who got them returned to treatment significantly sooner and received more of it.
The principle is simple and it is the whole argument. People in trouble do not reliably reach out. The system has to reach in.
What actually reduces the risk
Treat weeks one to four as the highest-risk period they are, not as the recovery lap. Whatever support exists should be at its most intensive here.
Reduce cue exposure deliberately and in advance. What is in the house, which route home, which events are declined for the first month. Easy to arrange two weeks ahead; nearly impossible to negotiate on day 32.
Build stress capacity before it is tested. Sleep, structure, food, exercise — unglamorous, and directly relevant given that stress reactivity is what predicted relapse.
Make contact proactive. Somebody checking in on a schedule and noticing when contact stops.
Step down, don’t stop. Support that ends abruptly recreates the same cliff, just later. Dennis and colleagues found a median of nine years between a person’s first treatment episode and their last use — against that timescale, tapering over months is not excessive.
The reframe worth holding onto
A relapse in week three is not evidence that treatment failed or that the person did not mean it. It is the predictable consequence of removing every protective structure at once, at the precise moment that stress reactivity is highest.
Which is good news, in a way: it is a design problem, and design problems can be solved in advance. That is what support through the transition is — someone present during the weeks the research identifies as most dangerous, reducing as the person’s own structure takes over.
References
- Sinha R (2008). Chronic Stress, Drug Use, and Vulnerability to Addiction. Annals of the New York Academy of Sciences. 10.1196/annals.1441.030
- Sordo L, Barrio G, Bravo MJ, et al. (2017). Mortality risk during and after opioid substitution treatment. BMJ, 357, j1550. 10.1136/bmj.j1550
- Scott CK, Dennis ML, Foss MA (2005). Utilizing Recovery Management Checkups. Drug and Alcohol Dependence, 78, 325-338. 10.1016/j.drugalcdep.2004.12.005
- Dennis ML, Scott CK, Funk R, Foss MA (2005). The duration and correlates of addiction and treatment careers. Journal of Substance Abuse Treatment. 10.1016/j.jsat.2004.10.013
Read more about Personal Recovery Assistant.
