did each point of baseline ocd severity cut the odds of ssri response by about 8%?
“every one point increase in baseline severity lowered the odds of response by about 8%.”
a host, a psychiatrist said on The Carlat Psychiatry Podcast, 2026-10-12, at 05:10. hear it in context → the speaker cited: a 2025 study by Cohen and colleagues.
the episode: How Much Do SSRIs Really Help Kids with OCD? (at 05:10). we link the show's own page; we do not re-host audio.
first, the part that holds
the number is right and it is the paper's own number. a 2025 individual-participant meta-analysis of placebo-controlled ssri trials in children and adolescents with ocd reports that for each one-point increase in baseline cy-bocs, the odds of being a responder fell by 8% in the treatment group relative to the placebo group.
what we found when we went and looked
we read the pubmed record for the 2025 paper in the journal of the american academy of child and adolescent psychiatry (64(7):775-785) and the full text of the version of record. the authors pooled raw patient data from 4 placebo-controlled ssri trials (614 patients, 86% of everyone ever enrolled in such trials for pediatric ocd). baseline severity was the only significant modifier of response, defined as a 35% or greater drop in cy-bocs: the full text gives a per-point ratio of 0.92 (95% ci 0.88 to 0.96, p = .0074) and says the odds of being a responder "decrease by 8% in the treatment group compared with the placebo group." so "about 8%" matches. the paper's own caveats matter, though: the effect was absent at a 25% response cutoff, severity did not modify the continuous symptom-change outcome, and the 8% describes how much the drug's advantage over placebo shrinks, not a child's overall chance of improving.
assertion by assertion
| assertion | verdict | what the record says |
|---|---|---|
| the figure comes from a 2025 meta-analysis by the named first author and colleagues | confirmed | the paper matches the speaker's attribution: an individual-participant meta-analysis in j am acad child adolesc psychiatry, july 2025, 64(7):775-785, doi 10.1016/j.jaac.2025.01.001. it pools 4 trials, 614 patients, all from north america, 88% white. [1][2] |
| each one-point increase in baseline cy-bocs severity reduced the odds of treatment response by about 8% | confirmed | the abstract says severity was negatively correlated with the odds ratio for response (p = .0074). the full text gives a per-point ratio of 0.92 (95% ci 0.88-0.96), which is the 8% drop, for a 35% or greater cy-bocs reduction. the abstract prints this as β = -0.92 and the full text as β = .92 with a ci of 0.88-0.96, so the value is a per-point odds ratio and the minus sign in the abstract is a reporting slip. the 8% figure is unaffected. [1] |
| popular reading: children with more severe ocd are 8% less likely per point to improve on an ssri | misattributed | the modifier is an interaction term, treatment versus placebo. it measures how the ssri's odds ratio over placebo shrinks as baseline severity rises. it does not measure the raw chance that a treated child responds. a post hoc split found odds ratios of 2.5 (mild, significant), 1.68 (moderate, ns) and 1.61 (severe, ns), so the point estimates still favored the drug in every severity band. [1] |
| popular reading: baseline severity is an established predictor of ssri benefit in pediatric ocd | overstated | in the same paper, severity did not modify symptom change measured on the continuous cy-bocs scale, and the response effect disappeared at a 25% cutoff. there were only 4 trials, and the severity subgroups were post hoc. the authors call it a modifier, but it is one positive result among several candidate modifiers and outcome definitions. [1] |
both directions
for the speaker: the paper states the 8% per-point figure in nearly the speaker's words. the abstract's results and conclusion both name baseline severity as a negative modifier of response. this is an individual-participant analysis that covers 86% of everyone ever randomized in these trials, with generally low risk of bias, which is the kind of design able to detect patient-level modifiers. the accompanying editorial in the same issue calls the approach a gold standard.
against: nothing in the record contradicts the number. what cuts against it is how it might be heard. the 8% is relative to placebo, so it describes how the drug's edge over placebo narrows. it is not the child's overall odds of getting better. the moderation appeared only at the 35% response cutoff: not at 25%, and not on continuous symptom change. in the post hoc severity bands, the moderate and severe groups still had odds ratios above 1 (1.68 and 1.61) but not statistically significant ones, which is not the same as no benefit. all data come from 4 industry-sponsored north american trials, mostly from the 1990s and 2000, in a mostly white sample.
what this cannot say
four trials is a small base for a meta-regression, and the result rests on one outcome definition. the authors could not test comorbidity, prior treatment including cbt, or socioeconomic status, so severity may be standing in for something unmeasured. the findings cover short-term (10-13 week) trials only. on our side: we read the abstract through the ncbi pubmed e-utilities record. the per-point 0.92 ratio, the 8% sentence and the 25%-cutoff sensitivity result come from the full-text version of record (cc by), which we read in the authors' institutional repository. that host is outside our citation list and the journal site blocked our fetch, so we cite the pubmed record and the doi for those details. we did not hear the podcast segment, so we take the quote and timestamp as supplied.
sources, each one fetched and read
- Systematic review and meta-analysis of individual participant data: randomized, placebo-controlled trials of selective serotonin reuptake inhibitors for pediatric obsessive-compulsive disorder. J Am Acad Child Adolesc Psychiatry. 2025;64(7):775-785. doi:10.1016/j.jaac.2025.01.001. PMID 39799995 (PubMed record) What we read there: "Data were obtained from 4 studies comprising 614 patients... Analysis of response showed an odds ratio of 1.89 (95% CI 1.45-2.45). Of all possible modifiers, severity was correlated negatively with odds ratio for response (β = -0.92, p = .0074)." Full text of the same paper: "for OCD severity with each 1-point increase in baseline CY-BOCS, the odds of being a responder decrease by 8% in the treatment group compared with the placebo group... This effect was absent for ≥25% response." Fetched 2026-10-15 (HTTP 203). https://pubmed.ncbi.nlm.nih.gov/39799995/
- Editorial: meta-analysis of individual participant data from trials of selective serotonin reuptake inhibitors for pediatric obsessive-compulsive disorder: a gold standard approach. J Am Acad Child Adolesc Psychiatry. 2025;64(7):768-769. PMID 40139344 (PubMed record) What we read there: the companion editorial in the same issue describes the study as investigating "the moderating effects of baseline patient variables on treatment outcome" and frames the individual-participant design as a gold standard approach. Fetched 2026-10-15 (HTTP 203). https://pubmed.ncbi.nlm.nih.gov/40139344/
speakers are described by role, not named. one sentence is quoted under fair use, with attribution and a link to the original at the timestamp; we do not re-host audio or video. nothing on this page is a reason to start, stop or change a medication. if you have the file behind anything marked unverifiable, the research board is where people can answer. adam