On therapy that makes people worse, the field’s missing adverse-event system, and why the harmed carry it alone.
by Jod(y)ie
In brief: therapy that makes people worse is real and measured — a meaningful fraction of clients deteriorate in treatment, clinicians underestimate it, and psychotherapy has no adverse-event reporting system. The harm goes home with the client, where silence converts it into shame. Naming returns that cost to sender.
Start with a comparison that should be uncomfortable.
If a medication destabilizes you, there is a system. Adverse events get reported, aggregated, studied. Black-box warnings exist. A doctor who dismisses your reaction can be contradicted by a database. The harm has infrastructure — imperfect, but real. Your suffering lands somewhere.
If a therapy destabilizes you — if the treatment itself, not your condition, degrades your functioning for years — there is, effectively, nothing. No reporting system. No registry. No black-box warning on a modality or a practice pattern. The research that exists tells us the phenomenon is real: studies of therapy outcomes have long found that a meaningful fraction of clients deteriorate in treatment — commonly estimated in the mid-to-high single digits, with higher rates in some populations — and that clinicians reliably underestimate their own deterioration rates, often dramatically. The field has known this for decades. It has built almost nothing in response.
So where does the harm go, if there’s nowhere for it to land?
It goes home with the client. That’s the whole answer. This essay is about what that costs, why it stays invisible, and what the harmed are owed — starting with a name.
Why doesn’t anyone see therapy harm? The camouflages
Therapy harm has a set of structural camouflages that other harms don’t, and they stack.
It happens behind a closed door with no witnesses. Two people, one record — kept by the clinician.
One party holds the credential and the vocabulary. In any dispute over what happened in that room, one account arrives wearing a license.
The harm masquerades as the illness. This is the cruelest one. Destabilization produced by treatment looks, from the outside, exactly like the condition the treatment targets. Getting worse becomes evidence you needed more of what was making you worse. The room manufactures its own justification.
The client’s report is discountable by design. The person raising the alarm is, by definition, someone in treatment for something. Every accurate observation can be rerouted into the case file: resistance, projection, symptom. In an earlier essay I called this the unfalsifiable room — a container in which no sentence you say can be received as data about the container. Add it up and you get a harm that is nearly impossible to see from outside and nearly impossible to prove from inside.
And the people who love you can’t help, not because they don’t care, but because culture supplies exactly one script: therapy is help; therapists are safe; if you’re getting worse you must need it more. Every feature above feeds that script. Your people weren’t failing to see. They were reading through a lens with no category for the container is the injury — a category that, until someone names it, does not exist.
The second injury
Here is what I most need the harmed reader to understand, because it’s the part shame will have privatized: the silence around this harm is not incidental to it. The silence is a second injury, distinct from the first, with its own damage profile.
The first injury is what happened in the room. The second is what happened after: searching for your experience in the literature and finding nothing. Telling someone and watching the script activate. Wondering — because there was no name, no registry, no wall of documented cases — whether you were the only one, and what that said about you. Doing the repair work alone, unpaid and unmapped, while the person who caused the injury continued practicing, protected by every camouflage above.
Harm that lands in infrastructure gets metabolized by the system. Harm that lands in silence gets metabolized by the person — as shame. That conversion is the second injury’s signature move: the field’s missing category becomes your private defect. If you’ve spent years believing your therapy harm said something about your brokenness rather than the field’s blind spot, that belief was not a conclusion you reached. It was a cost that was transferred to you.
I’m returning it to sender. That’s what naming does.
What would accountability for therapy harm look like?
I’m not writing as an enemy of therapy. The lineage I read to survive — Herman, Porges, Fisher, van der Kolk, Perry — is clinical to its bones, and the doctrine that would have prevented my own harm was written by the field, decades ago: stabilization before processing, containment before opening — Judith Herman’s phase-oriented doctrine, in Trauma and Recovery. My argument is not that therapy is dangerous. It’s that a profession which can help this powerfully can harm this powerfully, and only one of those capacities has infrastructure.
What the missing half would look like is not mysterious:
An adverse-event culture. Routine outcome monitoring that a clinician cannot opt out of noticing; deterioration treated as a findable, reportable event rather than a client attribute.
An honest literature. Research on therapy-caused harm funded and cited at a scale proportional to its prevalence — not the current afterthought.
Exit-interview data. The clients who quietly leave are the dataset. Nobody collects it.
And a professional vocabulary for “I opened something I don’t have the map for” — spoken as a safety report, not a career risk. The sentence exists. The culture that would reward saying it does not, yet.
Until the field builds these, the record is being kept where it has always been kept: by the harmed, in essays like this one, on walls of 2am search results, in the accounts that keep arriving in my inbox saying this happened to me too. That record is not less real for being unofficial.
It’s just early.
To the one carrying it alone
You were harmed, and there was no place to put it, so you carried it. Both halves of that sentence are facts, and neither is about your character.
What you’re owed — acknowledgment, accounting, in some cases restitution — may or may not be recoverable from the people and systems that owe it; that’s a case-by-case fight, worth fighting where it’s winnable. But the portion of the weight that was never yours — the shame that grew in the silence — that part you can put down today, on the strength of a single reframe:
The absence of a name for what happened to you was never evidence that it didn’t happen.
It was evidence that you got there before the map did.
Related on this shelf:
– The Client Becomes the Cartographer — the unfalsifiable room, from the inside
– Your Therapy Opened Something. Nobody Gave You the Map. — when the unnamed thing is an awakening, not an injury
Sources & lineage
- Lambert, M. J., & Ogles, B. M. — deterioration rates and clinician blind spots, Bergin and Garfield’s Handbook of Psychotherapy and Behavior Change (Wiley)
- Lilienfeld, S. O. (2007). Psychological treatments that cause harm. Perspectives on Psychological Science, 2(1)
- Judith Herman — Trauma and Recovery (Basic Books, 1992)
- Crawford, M. J., et al. (2016). Patient experience of negative effects of psychological treatment. British Journal of Psychiatry, 208(3)
This essay orbits the cornerstone: Your Therapy Opened Something. Nobody Gave You the Map. For the full framework, start with the map.