Guide
What predicts therapy dropout?
Updated September 15, 2026 · kept current as we cover new studies
The short answer
Three things show up in the research we have covered: who the person is, which therapy they are given, and how often the sessions happen.
On the person: in a study following 85 people through a DBT programme, those with high negative affectivity or anankastia were more likely to drop out, and those with very high negative emotions or psychotic-like traits improved less. Personality functioning moved the most of anything measured, which makes it a candidate marker of progress.
On the therapy and the schedule: across 84 study arms from 66 PTSD trials, patients were somewhat less likely to complete prolonged exposure than cognitive processing therapy overall, but when prolonged exposure was delivered more often than weekly, the odds of completing it doubled. Patients were also more than twice as likely to complete EMDR as CPT. Scheduling, in other words, is not just logistics.
And one finding about the field rather than about patients: only about a third of those trials defined what completing treatment even means, which makes dropout rates harder to compare than they should be.
What this page covers. Drawn from a DBT cohort study and a PTSD meta-regression, plus studies on what early sessions predict. Dropout has been studied far more widely than this; these are the studies we have summarised. Every claim above is drawn from the studies below, and each of those links to the original paper.
Who is more likely to leave
The person-level predictors, measured dimensionally rather than by diagnosis.
Who drops out of DBT — and who improves?
Following 85 people through a DBT program, personality dimensions predicted both improvement and dropout better than symptoms alone.
Which therapy, and how often
Completion differs between the established trauma therapies, and session spacing moves it.
Does seeing clients more often help them finish PTSD treatment?
A meta-regression across 84 study arms asked whether session frequency drives treatment completion.
What early sessions predict
These studies are about outcome rather than dropout directly, but both bear on what a session is doing when it is working.
Crying in therapy predicts better outcomes
Across 58 patients, those who cried more intensely and longer in early sessions reported better results later.
Why does trauma therapy actually work? It may come down to changing beliefs
A meta-analysis pooling 44 studies and over 5,000 participants found that the more a person's trauma-related beliefs shifted during therapy, the more their PTSD symptoms improved (r = .45).
Lowering the barrier instead
If attendance is the constraint, one response is to change what has to be attended.
Does a mental health app still work without a coach checking in?
In a three-arm trial recruiting across 13 countries, 825 adults living with chronic medical conditions used a digital wellbeing program either on their own or with weekly phone check-ins. Both versions cut combined anxiety and depression scores by roughly 3 points versus a waitlist, and the two versions were statistically indistinguishable from each other.
Lifestyle medicine for depression, without a therapist
A three-arm trial of an 8-week program delivered by app, by booklet, or not at all.
Common questions
Which trauma therapy do people finish most often?
In the meta-regression we covered, patients were more than twice as likely to complete EMDR as cognitive processing therapy, and somewhat less likely to complete prolonged exposure than CPT. Completing a therapy is a separate question from how well it works.
Does seeing a therapist more often help people finish?
For prolonged exposure, yes: delivered more frequently than weekly, the odds of completion doubled. For EMDR the frequency effect was not statistically clear.
What personality traits predict dropping out?
In the DBT study, high negative affectivity and anankastia predicted dropout. Very high negative emotions or psychotic-like traits predicted less improvement rather than dropout as such.
Why are dropout rates hard to compare between studies?
Because only about a third of the trials in that meta-regression defined what counts as completing treatment. Without a shared definition, two studies can report very different dropout rates for the same behaviour.
This is an educational summary, not medical or psychological advice, and it is not a substitute for consultation with a qualified professional. Read the full disclaimer.