“They seem better, so the worst is over”
The myth
Once someone seems better, the danger has passed.
Actually
The weeks after a crisis — including after discharge from hospital — are among the highest-risk periods there are.
4 minUpdated 4 August 2026
Of everything on this page, this is the myth most likely to result in a death, because it does not merely mislead — it prompts an action. It tells worried people that the sensible thing to do now is step back.
What actually happens after a crisis
The period following a suicidal crisis is among the highest-risk periods there is. Risk after discharge from a psychiatric inpatient stay is dramatically elevated compared with the general population, and it is highest in the first days and weeks — precisely when everyone involved has just been told things are better.
The same shape appears after an emergency department visit, after a crisis team signs off, and after a course of therapy ends. The support is withdrawn on the schedule of the service, not the person.
The first 72 hours after a hospital stay →
Why "better" can be misleading
Several different things all look like improvement from outside.
Genuine recovery. Common, real, and what everyone hopes for.
Exhaustion passing. A crisis is physiologically expensive. Someone who has slept can look transformed without anything underlying having changed.
Relief at having decided. This is the one people miss. A person who has resolved something — in either direction — often becomes calmer, more settled, sometimes noticeably lighter. It reads as recovery.
Performance. By the second week of everyone watching you, most people have learned exactly what to say to make the watching stop. Being fine is easier than being asked.
You cannot reliably tell these apart from outside, and the person themselves frequently cannot either.
What this does not mean
It does not mean improvement is fake, or that recovery should be treated with suspicion. Most improvement is real. Treating every good week as a warning sign is exhausting for everyone and communicates that you do not believe them, which is its own harm.
The point is narrower: improvement is not a reason to withdraw support. It is a reason to keep going for longer than feels necessary.
What to do instead
Keep the rhythm past the point it seems needed. If you were texting weekly during the bad patch, keep texting weekly for months afterwards. The check-in that matters is the one three months later when everyone else has stopped.
Diarise the drop-offs. Discharge date, last therapy session, end of a sick note. Put a reminder in for a week after each. Those are the predictable moments when the structure disappears.
Say the quiet part. I know things are better. I'm going to keep checking in anyway, and it's not because I don't believe you. That removes the implication of doubt and makes the continued contact easy to accept.
Help them write down their early signs while they are well. This is the single most useful thing that can be done in a good week. Early warning signs worksheet →
For services, and for anyone with a discharge letter
Ask what happens if things deteriorate, get the answer in writing, and get a follow-up appointment booked before leaving the building.
Being discharged from mental health services → · Relapse and setbacks →