07/20/2026
🤔My 2 🪙🪙
Building on Chris Aiken’s Take on the New CANMAT/ICOCS OCD Guidelines
Chris Aiken’s recent critique of the new CANMAT/ICOCS OCD guidelines (Van Ameringen et al, Journal of Psychiatric Research, 2026) makes a strong case that antipsychotics — haloperidol in particular — shouldn’t have been elevated to first-line augmentation status while better-tolerated agents like ondansetron were pushed to third-line. I want to say yes, and — building on his argument from my own clinical experience.
Where I fully agree. Haloperidol’s promotion rests on a single 34-subject RCT, and the “well tolerated” claim leans on an all-cause-discontinuation metric from schizophrenia trials that actually reflects efficacy, not tolerability — that metric ranked haloperidol among the more effective antipsychotics for schizophrenia, not among the gentler ones.
The guidelines’ silence on tardive dyskinesia risk is a real omission, and Aiken’s point about ondansetron having comparable trial support (7 RCTs, 434 subjects) with a cleaner cardiac profile deserves more weight than it got.
What I’d add. In my own practice, by the time augmentation comes up, my patients have already been through multiple full trials at anti-obsessional SSRI dosing — not just antidepressant dosing. That distinction matters more than most guidelines acknowledge: OCD often needs doses well above what’s standard for depression before a trial can honestly be called “failed.”
Once that groundwork is done, I reach for memantine first, if all else is equal and there are no comorbidities. I’m increasingly using ondansetron as well, with constipation typically the side effect that limits how far I can push the ondansetron dose.
It’s worth underscoring something Aiken’s piece touches on but doesn’t dwell on: memantine and riluzole are both borrowed from other fields — memantine from Alzheimer’s disease, riluzole from ALS — and in OCD, the emerging picture is that they may be pulling more clinical weight than they do in their original indications. That’s a pattern worth watching as the evidence matures.
None of this makes antipsychotics obsolete. They still have a clear place, particularly for tic-related OCD. But tardive dyskinesia is a lifelong risk, and that alone argues for trying better-tolerated options first wherever the evidence allows it — which is exactly Aiken’s point, extended.
For SSRI and clomipramine dosing, as well as augmentation strategies, I’d point psychiatrists to IOCDF
materials, and to clinicians like Dr. Steven Poskar — director of OCD NYC and Co-Chair of the IOCDF’s Treatment Recommendation Guidelines — Dr. Robert Hudak, a leading contributor to the OCD treatment guideline literature, and Dr. Sanjaya Saxena, who sits on the IOCDF’s Scientific and Clinical Advisory Board and formerly directed the OCD programs at UCLA and UC San Diego.
Bottom line: we have a long list of augmentation agents and no single one is robustly superior across the board. Choose based on tolerability and the individual patient — tics favor antipsychotics, migraine history favors topiramate, bipolarity favors lamotrigine (must be mood stabilized before trials of an SSRI), GI symptoms favor ondansetron — after adequate first-line trials, not before.
