Is haloperidol a first-line augmentation strategy for OCD? Wait a minute…
STUDY: Van Ameringen M et al, Journal of Psychiatric Research 2026
STUDY TYPE: Expert consensus guideline (systematic review with graded evidence)
FUNDING: Internal CANMAT funds
Background
Experts from the Canadian Network for Mood and Anxiety Treatments (CANMAT) teamed with the International College of Obsessive-Compulsive Spectrum Disorder (ICOCS) to create their first joint guidelines on OCD. Read on for some friendly controversy.
Summary
First line: SSRIs and CBT (exposure-response prevention) are equally effective first-line options, with no evidence favoring one over the other.
Second line: Clomipramine, despite strong efficacy, is second-line because of its side-effect burden and cardiac risks.
For non-responders to first-line options, they recommend transcranial magnetic stimulation (TMS). TMS performs better in patients who haven’t yet failed multiple treatments, so they suggest using it earlier rather than holding it as a last resort.
Augmentation
For augmentation in treatment-refractory cases, they recommend:
First line: Augmenting an SSRI with aripiprazole, risperidone, or haloperidol, or adding CBT/ERP.
If that fails…
Second line: Lamotrigine, memantine, or a high-dose SSRI trial are next, each needing at least eight weeks before judging benefit.
Third line: Ondansetron and others below…
Deep brain stimulation and ablative neurosurgery are reserved for the most severe, refractory cases, with both showing similar response rates, around 50 to 60 percent.
A Critical Flaw
For refractory OCD, they recommend haloperidol and other antipsychotics first-line, while relegating ondansetron third-line. I disagree, and will argue both sides here.
The Pro-Antipsychotic View
- Supported as SSRI-augmentation by RCTs involving 434 subjects
- More effective for OCD with tics
OK, but why haloperidol? The authors admit the support is limited, citing a single RCT of 34 subjects. But they move it to first-line because:
- It is inexpensive and widely available internationally
- It is “well tolerated” among the antipsychotics
The Skeptical View
To argue that haloperidol is well-tolerated, they cite a meta-analysis in schizophrenia that found it had low rates of “all cause discontinuation.” However, in schizophrenia trials this metric usually reflects efficacy, not tolerability, and haloperidol was one of the most efficacious antipsychotics for schizophrenia in the analysis.
But haloperidol has a high risk of tardive dyskinesia, a warning that never appears in the OCD guidelines. And the schizophrenia meta-analysis ranks it higher for:
- Parkinsonism, akathisia, anticholinergic effects, QTc prolongation, and prolactinemia.
An Argument for Ondansetron
I’m biased here, as I’ve used ondanstron successfully in hundreds of patients with OCD, both as monotherapy and augmentation (although the studies are mostly monotherapy).
But let’s look at the data. Lumping ondansetron in with another serotonin 5HT3 antagonist, granisetron, we find:
- Support from 7 RCTs involving 434 patients, and a positive meta-analysis
That’s nearly the size of the antipsychotic data (n = 491). So why is ondansetron third-line?
The authors argue that it is too risky, citing QTc prolongation. But they do not mention that antipsychotics also prolong the QTc, and here the prolongation is more dangerous:
- Antipsychotics are linked to fatal arrhythmias (torsades de pointes) from QTc prolongation.
- With ondansetron, the prolongation did not progress to torsades de pointes or ventricular tachyarrhythmia in a meta-analysis of 170 randomized trials.
The authors also neglect to mention other rare cardiac risks with antipsychotics: Sudden cardiac death, increased cardiac mortality in depression and elderly, dyslipidemia, hypotension, and, with some agents, myocarditis and cardiomyopathy.
Rare, yes, but they are talking about adding them to SSRIs, which can lead to toxic antipsychotic levels due to CYP inhibition with fluoxetine, sertraline, paroxetine, and fluvoxamine.
Availability
As for the argument that haloperidol is the only option in some countries, I have doubts. Ondansetron is low-cost and widely available in developing countries, described by the Cochrane group as the “most used gold standard antiemetic drug worldwide.”
Safety and Tolerability
Psychiatrists are comfortable with antipsychotics, and uncomfortable with non-psychiatric medications like ondansetron. But patients are not as comfortable with antipsychotics. Here’s why:
- Ondansetron: Headache, fatigue, and constipation; but reduces gastrointestinal side effects on SSRIs.
- Antipsychotics: Akathisia, parkinsonism, sexual dysfunction, fatigue, weight gain, dystonia, anticholinergic effects, and possibly cognitive impairment.
When it comes to safety, the differences are even more extreme. Besides the cardiac risks above, antipsychotics cause tardive dyskinesia (around 25% over 10 years with second generation agents, and much higher with haloperidol), and:
- Diabetes, metabolic syndrome, falls, hyperprolactinemia, neutropenia, temperature imbalance, and neuroleptic malignant syndrome.
Also, antipsychotics can cause OCD, something common on clozapine but reported with most of them. Ondansetron was developed in the 1960’s and has lit up no safety risks besides QTc prolongation without further cardiac complications.
Practice Implications
- Start with SSRIs and CBT for OCD.
- Use more TMS. It is well-tolerated and about twice as effective as medication.
- We have a long list of augmentation agents. None stand out as robustly evidence based, so best to choose them based on tolerability rather than elevating one to first-line.
- You can also personalize the augmentation strategy: lamotrigine for bipolar, topiramate for migraines, ondansetron for binge-eating or nausea, antipsychotics for tics.
—Chris Aiken, MD
Director, Psych Partners
Editor in Chief, Carlat Psychiatry Report
