What treatment-resistant OCD means in practice

OCD is a treatable disorder. For most people, the combination of a serotonin reuptake inhibitor and a course of exposure and response prevention therapy produces meaningful improvement. But a substantial minority does not respond, and that minority is the subject of this article. Estimates from the OCD treatment literature suggest that 40 to 60% of patients fall short of an adequate response to first-line care, and roughly 10% remain severely impaired regardless of what they try.

The label “treatment-resistant” gets used loosely. We try to be precise. Most expert consensus defines treatment-resistant OCD as failure of at least two adequate SRI trials at therapeutic dose for a minimum of 12 weeks each, plus a genuine trial of ERP. Adequate ERP is not three sessions with a generalist who mentioned exposure once. It is structured weekly work with a clinician trained in the protocol, with active home practice between visits.

If you have been through that and the rituals still run your day, you are in the population this article is written for. If you are earlier in the process, our overview article on ketamine for OCD is the better starting point.

Why ERP and SRIs come first, and how to know they really failed

Before we discuss ketamine, it is worth being honest about what should have been tried first. The reason matters: ketamine is a tool that fits into a treatment plan, not a replacement for proven first-line care.

SRIs (the serotonergic antidepressants used at higher doses for OCD than for depression) are the medication standard. Adequate means therapeutic dose, 12 weeks minimum, and at least two distinct medications attempted. ERP is the behavioral standard, and it is uncomfortable by design. The therapy works by deliberately exposing patients to what triggers the obsession while preventing the compulsion that normally follows. Done well, the brain learns that the feared outcome does not actually arrive. Done poorly or in too few sessions, it does not stick.

We mention this because plenty of patients arrive in our consultation room believing they have failed treatment when they have actually never had a complete trial. If your ERP was four sessions of talk therapy with someone who used the word “exposure,” that is not a failed trial of ERP. We will help you sort that out before recommending anything.

If you do need a referral to a properly trained ERP clinician, our guide to finding an integration therapist in Tennessee covers some of the same ground for behavioral therapists in this region.

Rodriguez 2013 in detail

Rodriguez 2013 is the foundational study and the one most often cited in any honest discussion of ketamine for OCD. Carolyn Rodriguez and her colleagues at Columbia published the trial in Neuropsychopharmacology. It was a randomized, double-blind, placebo-controlled crossover design in 15 adults with OCD who had near-constant obsessions and were drug-free at the time of the study.

Each participant received a single 40-minute IV infusion of ketamine at 0.5 mg/kg and, on a separate day at least one week later, a matching saline infusion. Neither the participant nor the rater knew which session was which. The primary outcome was the Yale-Brown Obsessive Compulsive Scale (Y-BOCS), the standard severity measure for OCD.

The result that put this trial on the map: 50% of participants met treatment response criteria at one week after ketamine, defined as a Y-BOCS reduction of at least 35%. Zero participants met those criteria after saline. The effect appeared rapidly, often within hours of the infusion. It was the first randomized controlled trial showing that a glutamatergic agent could reduce OCD symptoms without an SRI on board.

What the trial did not establish is also important. Fifteen patients is a small sample. The drug-free design was deliberate (the researchers needed a clean signal) but it does not match how most OCD patients live. And while the acute response was clear, the durability of that response between infusions remained an open question that the field is still working on.

Ketamine is FDA-approved as an anesthetic; its use for OCD is off-label.

What has happened since: Bloch, Beaglehole 2024/2025, and the durability question

Rodriguez 2013 was a proof of concept. The work since has tried to answer two harder questions: does the signal hold up in larger and more realistic samples, and how long does it last?

A review by Pittenger and Bloch published in the Psychiatric Clinics of North America (2014) is one of the more careful syntheses of the glutamate-OCD connection. The authors argued that glutamatergic dysfunction in the cortico-striato-thalamo-cortical circuit is increasingly central to how researchers understand OCD neurobiology. That mechanistic story is what makes the NMDA-antagonist trials make sense in the first place. The hyperactivity in those circuits, the kind of looping cognitive activity patients describe, lines up with what an NMDA receptor blocker would be expected to interrupt.

A 2024 study by Beaglehole and colleagues, published in the Journal of Psychopharmacology, was a double-blind active-controlled crossover of ketamine in treatment-resistant OCD. The trial confirmed acute Y-BOCS reductions consistent with the Rodriguez signal in a population that more closely resembled the real-world refractory patient. Durability between infusions, again, was identified as the central limitation.

The honest summary: the acute effect is real and reproducible in small samples. The maintenance schedule that keeps it going is empirical. There is no FDA-defined OCD protocol for IV ketamine because no such protocol has been validated in a large trial. Anyone who tells you otherwise is overselling.

Why ketamine pairs with ERP, not replaces it

The most important thing to understand about ketamine in this population is that it is not a behavioral therapy. ERP is the durable piece. ERP rewires the learned association between obsession and ritual, and that rewiring is what carries patients forward over months and years.

What ketamine may do, based on its glutamatergic mechanism and the brain research that supports ketamine’s effects on neuroplasticity, is open a window in which ERP becomes more accessible. Patients who could not previously tolerate exposures often report that the acute distress around triggers is reduced for some period after ketamine. That reduced distress is not the cure. It is the moment when the harder behavioral work becomes doable.

Practically: a thoughtful protocol pairs ketamine sessions with active ERP work in the same window. We do not provide ERP at MCK. We coordinate with your existing therapist or help you find one. Ketamine without behavioral work is a partial intervention.

This is one of the places where treatment-resistant OCD overlaps with the broader picture of ketamine for treatment-resistant depression and ketamine for anxiety. The mechanism is similar, the framing is similar, and the integration with therapy is what separates a useful course of treatment from a series of expensive afternoons.

Who is a candidate

Based on the available evidence, the strongest candidates for ketamine in OCD tend to share a few features. We look for:

If those conditions fit and you have read through this far, a consultation is a reasonable next step. If they do not, we will tell you that directly and point you toward what does fit.

What an MCK OCD consultation looks like

The first step is conversation. We review your full treatment history: medications tried, ERP work completed, current symptoms, current providers, and what an honest definition of success would look like for you. The point of that conversation is not to sell you on ketamine. It is to find out whether ketamine is a reasonable addition to what you are already doing, or whether something else should come first.

If we move forward, Marla Peterson, CRNA, oversees every infusion with anesthesia-level monitoring: continuous pulse oximetry, blood pressure, and heart rate tracking, with a CRNA in the room and available throughout. We use the same monitoring approach described in our how it works overview and our OCD treatment page.

The treatment environment is private and quiet. Comfortable recliners, low light, the option of music, and our therapy dogs Walter White and Wilma if you want them. Patients consistently describe the setting as calmer than what they expect from a clinical infusion. You will need a driver for each session.

Cost is straightforward. Sessions are $475 each at MCK and insurance generally does not cover ketamine for OCD because the use is off-label. We are transparent about that from the first call. If cost is a barrier, we will say so honestly rather than soft-selling the financial picture.

Honest expectations

A few things we want patients to hear plainly before they decide.

If you have been doing the right things for years and OCD still runs your day, we understand the exhaustion. The treatment-resistant population is exactly where the data on ketamine for OCD has the most relevance, and exactly where an honest conversation matters most.