Eight questions to ask before they leave treatment.

Plenty of discharge plans end with a phone number and the words “call to schedule.” What happens next is not encouraging. In a study of 18,793 Medicaid patients leaving psychiatric units, people with no history of outpatient care were about three times more likely to show up within 7 days if the appointment was already on the calendar when they walked out.

Showing up matters more than it sounds. Among 10,064 veterans leaving residential addiction treatment, those seen by an outpatient addiction provider within 14 days had 23% lower odds of dying over the next two years.

There are two separate problems here. The first is between providers. In hospital medicine, where this has been studied closely, the discharge summary reaches the next doctor by the first follow-up visit only 12 to 34% of the time, and the two doctors speak directly in 3 to 20% of cases. Addiction treatment has no reason to think it does better.

The second problem is the family. Federal law (42 CFR Part 2) protects addiction treatment records more tightly than almost any other medical record. Without a signed release naming you, the outpatient provider cannot tell you whether your son went to his appointment. Since the 2024 rule change, one consent can cover future treatment-related sharing, so there is no reason to leave the meeting without it.

If opioids were involved, this may be the most important question on the list. In a 10-state Medicaid study of 90,639 residential discharges, only 36% of people received a medication for opioid use disorder in the 30 days after leaving. People who had been on one before admission were far more likely to continue, by 24 percentage points.

For alcohol, the numbers are worse. In 2019, 1.6% of U.S. adults with alcohol use disorder took a medication for it, even though several exist and work. A medication started inside treatment stops the day the prescription runs out, unless someone has already arranged the next one.

“Home for now” is the default, and sometimes it is right. It should still be an actual decision, made out loud, with the team’s input. In a randomized trial, people leaving treatment were assigned either to a recovery residence or to usual aftercare. Two years later, 31% of the recovery-residence group had returned to use, against 65% of the others, and their monthly income was more than double.

For young adults the pattern is even sharper, as Issue 01 covered: about six times the odds of staying continuously abstinent, yet only 20 to 35% go.

If opioids or stimulants were involved, the first weeks home are the most dangerous ones, because tolerance has dropped and the supply is unpredictable. The fix is cheap and proven. Massachusetts communities that trained more than 100 people per 100,000 residents in overdose response had 46% lower opioid overdose death rates than communities that trained nobody.

The person most likely to be there is a family member. In a CDC review of overdose deaths among 10- to 19-year-olds, about six in ten happened at the young person’s own home, and in two-thirds of cases someone else was present.

Sometimes this question comes up because insurance has stopped paying, or because the client wants out. Either way, ask it out loud. In the veterans study of residential programs, people who left within the first week had 49% higher odds of dying over the next two years. NIDA has said for years that treatment shorter than about three months, across whatever levels of care, rarely holds.

Nobody is saying residential has to last three months. The question is what comes after it, and whether the dates are written down.

Plans tend to assume success. Ask the team to plan for the other outcome while everyone is still in the room and calm. In two randomized trials, people who got scheduled recovery check-ups after treatment went back to treatment more often when they needed it (55% against 37%), and got there about a year sooner.

A slip that gets a response in week two is a very different situation from one nobody notices until spring.

Families usually leave treatment with one of two instructions: “be supportive” or “let him do the work.” Neither one tells you what to do on a Tuesday. The research says family involvement helps when it is structured. CRAFT, a training approach for families, gets reluctant people into treatment at two to three times the rate of Al-Anon-style support or a traditional intervention.

A 2020 meta-analysis of 16 trials found that bringing a family member into treatment produced a small but consistent extra drop in substance use, and the benefit lasted up to 18 months. Ask for the structure.