What I learned from research on OCD and PANDAS, and how it shapes my practice today

Between 2004 and 2007, as a resident and then a fellow in the Department of Psychiatry at the University of Florida, I co-authored a series of papers on obsessive-compulsive disorder: its circuitry, its treatment when it resists medication, its relationship to tic disorders, and the hypothesis that some childhood cases are triggered by streptococcal infection. My last paper appeared in 2008, and I now treat adults with depression and anxiety in a community practice. But that research shaped how I think as a clinician more than anything else in my training, and I want to explain how.

What we were working on

The Gainesville OCD group in those years was asking a set of linked questions. In a 2006 paper in Progress in Neuro-Psychopharmacology and Biological Psychiatry, we reviewed the neurocircuitry of OCD and its connection to disgust, the emotion that sits underneath so much contamination-related obsession. In a 2004 review in CNS Spectrums, we surveyed the treatment of refractory OCD from medication through to deep brain stimulation. In a 2007 study in the Journal of Psychiatric Research, we asked whether patients with OCD and comorbid tics responded differently to an eight-week open-label trial of fluoxetine, using a predefined threshold for what counted as a clinically meaningful response.

The other strand was PANDAS, the acronym for pediatric autoimmune neuropsychiatric disorders associated with streptococcal infections. Susan Swedo and colleagues at the National Institute of Mental Health had described the syndrome in the American Journal of Psychiatry in 1998: children with abrupt-onset OCD or tics following strep throat, with a proposed mechanism of antibodies cross-reacting with the basal ganglia, by analogy with Sydenham chorea in rheumatic fever. We reviewed the preclinical and clinical evidence for a central nervous system infectious cause in a 2006 chapter in Advances in Neurology, and the broader question of autoimmunity in OCD and tic disorders in a 2004 paper in the Brazilian Journal of Psychiatry.

Lesson one: OCD is a circuit disorder, not a bad habit

The neuroimaging literature we reviewed converged on a loop: orbitofrontal cortex, anterior cingulate, striatum, thalamus and back again, overactive in OCD and quieting with successful treatment. The insula, which processes disgust, showed up as well. OCD was not a failure of willpower or an exaggerated version of ordinary tidiness. It was a specific brain circuit doing something specific.

That framing changed how I talk to patients. A person who washes their hands until they bleed does not need to be told to stop; they know. What helps is hearing that there is a reason this is happening, that it has a biology, and that the biology responds to treatment. I say a version of that to patients with every disorder I treat now.

Lesson two: adequate treatment means something different in OCD

Our review of refractory OCD made one point that I repeat in clinic weekly: OCD requires higher doses of SSRIs, for longer, than depression does. A trial of fluoxetine that would be adequate for depression at eight weeks and a moderate dose is not adequate for OCD, which often needs the top of the dosing range and ten to twelve weeks before the response is clear. Patients who have been told they are "treatment-resistant" after a depression-sized trial frequently are not. They are under-treated.

The review also made clear that medication alone is rarely enough. Exposure and response prevention, the behavioral therapy that asks patients to face what they fear and refrain from the ritual, has the strongest evidence of any treatment for OCD and works in combination with medication. When I see OCD in my adult practice, I treat the medication and the referral to an ERP therapist as a single prescription.

Lesson three: define success before you start

The 2007 fluoxetine study set a threshold for a clinically meaningful response in advance, then measured against it. That is standard practice in research and rare in clinics, where "better" is usually defined after the fact and in the direction of whatever the clinician hoped. I now set the response criterion with every patient at the first visit: this score, these functional goals, by this date. If we reach it, good. If we do not, the trial failed and we change course. Writing the criterion down before the trial starts removes the temptation to move the goalposts.

Lesson four: ask about tics, and about childhood

Comorbid tics were the variable in our 2007 study because the OCD that travels with tics appears to be a somewhat different illness: earlier onset, more often male, more often familial, and in some studies a different response to standard treatment. I still ask every OCD patient about tics and about the age their symptoms began. In an adult practice this matters because the adults in front of me were once children, and an adult with lifelong OCD that started abruptly at age seven after a sore throat has a different history from one whose symptoms crept in during a stressful year at thirty.

Lesson five: hypotheses need humility

PANDAS was, and remains, controversial. The mechanism was plausible, the case series were compelling, and the analogy to Sydenham chorea was elegant. Confirming it with controlled evidence proved far harder. When we wrote our reviews, the preclinical data pointed in interesting directions and the clinical data were suggestive but not definitive. Two decades later the field has broadened the concept, restricted the diagnostic criteria, and argued at length about treatments.

Working on PANDAS taught me the difference between a good hypothesis and an established fact, and how easy it is to let the first become the second in the mind of a worried patient or an enthusiastic clinician. I treat adults, and PANDAS is a pediatric diagnosis, so I rarely encounter it directly. But I encounter the same pattern constantly: a compelling explanation, a hopeful family, and a treatment that has outrun its evidence. The habit of asking "how do we actually know that" is the one piece of my research training I use every day.

Lesson six: the treatments catch up

When we wrote about deep brain stimulation for OCD in 2004, it was a research procedure in a handful of centers. In 2009 the FDA granted it a humanitarian device exemption for severe, refractory OCD. In 2018 the FDA cleared Deep TMS, the same technology I now oversee for depression at our Palm Beach Gardens center, for OCD as well. Interventions that were at the edge of the literature when I was a resident are now options I can offer.

What stayed with me

I did not become an OCD specialist, and I did not build a research career. I built a community practice focused on depression and anxiety, which is where the need is. But the way I practice came out of those years: the insistence on a biological explanation the patient can hold onto, on adequate trials before declaring failure, on criteria set in advance, on asking about the history that nobody asked about, and on the discipline of separating what we know from what we hope. Those habits are not specific to OCD. They are simply what I learned by studying it.

More from Dr. Husted

• Why I measure depression treatment by function, not just mood2026-06-11
• When to consider Deep TMS: how I decide with my patients2026-07-09
• Spravato in practice: what I tell patients before their first session2026-08-06

Educational content and not a substitute for evaluation. If you are in crisis, call or text 988.