does nice tell clinicians in england to use cbt rather than antidepressants first?
“groups like NICE in England are saying use CBT as the first line, not drugs.”
a guest, a historian of psychiatry (Andrew Scull) said on Lex Fridman Podcast, 2026-09-21, at 11:07. hear it in context → the speaker cited: NICE.
first, the part that holds
this part is true: england's depression guideline (nice ng222, june 2022) tells clinicians not to routinely offer antidepressants as first-line treatment for less severe depression unless the person makes an informed choice to take them, and cbt is one of the first-line options it recommends. in the evidence review nice commissioned, antidepressants showed no evidence of effect in less severe depression, and group cbt did.
what we found when we went and looked
we read the guideline's recommendations and the network meta-analysis nice built them on. for less severe depression, nice says "do not routinely offer antidepressant medication as first-line treatment ... only offer it if that is the person's informed preference". its review of 676 trials found group cbt beat usual care (smd -1.01) and found no evidence that antidepressants worked. but the claim leaves two things out. first, cbt is not the only first-line option: nice lists guided self-help first and says every listed treatment, including ssris, can be used first line. second, for more severe depression the first option nice lists is individual cbt combined with an antidepressant, and an antidepressant on its own is fourth, still a first-line option. so "cbt, not drugs" is accurate for less severe depression and wrong for more severe depression.
assertion by assertion
| assertion | verdict | what the record says |
|---|---|---|
| nice recommends against antidepressants as routine first-line treatment (for less severe depression) | confirmed | ng222 recommendation 1.5.3: "do not routinely offer antidepressant medication as first-line treatment for less severe depression. only offer it if that is the person's informed preference." a 2025 bjgp editorial quotes the same rule, and a 2026 audit of one english practice used it as its audit standard. the 'unless it's the person's preference' clause is real and matters: it is a default, not a ban. [1][2] |
| nice says cbt is the first-line treatment | overstated | cbt (group and individual) is one of about 11 first-line options for less severe depression, but nice lists guided self-help first and says all options can be used first line, starting with the least intrusive and least resource-intensive (guided self-help). nice says the order reflects its committee's reading of clinical and cost effectiveness. the list also includes ssris. [3] |
| the 'not drugs' advice applies to depression in general | false | for more severe depression, nice's table 2 lists 'combination of individual cbt and an antidepressant' first and antidepressant medication fourth, and says all of them can be used first line. the evidence review behind it found combined individual cbt plus an antidepressant was the most effective class against pill placebo (smd -1.18, credible interval -2.07 to -0.44). snris (-0.32) and mirtazapine (-0.35) on their own were also effective. [3] |
| nice's evidence shows psychological therapy outperforms antidepressants for milder depression | confirmed | the network meta-analysis nice commissioned (676 rcts, 105,477 participants) concluded: "group ct/cbt (and possibly group yoga and self-help) appears efficacious in less severe depression, whereas antidepressants do not show evidence of effect." group cbt vs treatment as usual: smd -1.01 (credible interval -1.76 to -0.06). the interval is wide. [3] |
both directions
for the speaker: for less severe depression, the speaker's description is close. nice made "not drugs by default" an explicit rule, and the reason it gave is the evidence: in its own meta-analysis, psychological therapies such as group cbt worked better than usual care, while antidepressants showed no evidence of effect at this severity. the rule is being applied and audited in english primary care: one 2026 audit scored practices against "antidepressants not routinely offered first-line for less-severe depression" and found 81.8% compliance (9 of 11 cases). a 2025 editorial in the british journal of general practice argued nice should go further and drop the preference exception. that suggests the guideline's direction is the one the speaker describes.
against: the claim drops the severity split, which is the main structure of the guideline. for more severe depression, nice's first-listed option is cbt combined with an antidepressant, not cbt instead of one, and an antidepressant alone is a recommended first-line option. even for less severe depression, nice does not single out cbt: guided self-help comes first, and an ssri is on the first-line list for people who prefer medication after an informed discussion. nice also says any option in either table can be used first line, chosen together with the patient. "use cbt as the first line, not drugs" turns a ranked menu, conditional on severity, into a single rule.
what this cannot say
nice's own website blocked automated access during this check (http 403). we read the recommendations text (1.5.3 and tables 1 and 2) from a copy of nice.org.uk/guidance/ng222/chapter/Recommendations archived in may 2026, served by the uk government web archive's mirror host. that host is not on our citation list, so the exact wording of table 2 rests on that archived copy plus the peer-reviewed evidence review, not on a live nice page. we cannot rule out that nice changed ng222 between may and october 2026. the severity cut-offs ('less severe' roughly matches mild, 'more severe' roughly matches moderate-to-severe on phq-9 cut-offs) come from scale scores in trials, not from clinical diagnosis. the meta-analysis credible intervals are wide. the audit covered 11 cases at one practice and says nothing about national prescribing. guidelines describe what clinicians are told to do, not what happens in practice. this page is not advice about any individual's treatment.
sources, each one fetched and read
- Editorial, 'Should antidepressants be prescribed simply if it is the patient's preference? Why NICE guidelines must be revised', British Journal of General Practice, 2025;76(762):7-9, doi:10.3399/BJGP.2025.0596 What we read there: NICE "currently advises that antidepressants should not be routinely offered as a first-line treatment for less severe depression 'unless that is the person's preference'"; the authors argue the preference clause should be removed. Fetched 2026-10-06 (HTTP 200). https://pmc.ncbi.nlm.nih.gov/articles/PMC13044191/
- 'A retrospective clinical audit on depression management in line with NICE NG222 guidelines at a primary care setting in rural England', European Psychiatry, 2026 What we read there: audit standard "antidepressants not routinely offered first-line for less-severe depression (unless documented patient preference)" met in 81.8% of cases (9/11), below the 90% target. Fetched 2026-10-06 (HTTP 200). https://pmc.ncbi.nlm.nih.gov/articles/PMC13442253/
- NICE guideline technical team and collaborators, 'A systematic review and network meta-analysis of psychological, psychosocial, pharmacological, physical and combined treatments for adults with a new episode of depression', eClinicalMedicine 2024;75:102780, doi:10.1016/j.eclinm.2024.102780, PMID 39246718 (PubMed record) What we read there: aim: "to update NICE guidance on the management of Depression in Adults in England." less severe: group CT/CBT vs TAU SMD -1.01 (CrI -1.76; -0.06); "antidepressants do not show evidence of effect." more severe: combined individual CT/CBT with antidepressants -1.18 (-2.07; -0.44); individual CT/CBT -0.78; mirtazapine -0.35; SNRIs -0.32; "antidepressants alone appear efficacious in more severe depression." Fetched 2026-10-06 (HTTP 203). https://pubmed.ncbi.nlm.nih.gov/39246718/
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