Review Are [CBT], cognitive therapy, and behavioural activation for depression effective in primary care? A systematic review and meta-analysis 2025 Carey+

[Andy]

Full title: Are cognitive behavioural therapy, cognitive therapy, and behavioural activation for depression effective in primary care? A systematic review and meta-analysis

Highlights​

• CBT for depression in primary care shows a significant difference compared to inactive controls, but no difference compared to active comparators
• There is no evidence of effectiveness of self-help CBT for depression in primary care
• Most studies examining CBT for depression in primary care appear of low quality

Abstract​

Cognitive behavioural therapy (CBT) is a recommended first-line treatment for depression. Evidence mainly derives from studies in secondary care, though most treatment occurs in primary care. This review examined efficacy of CBT, cognitive therapy (CT), or behavioural activation (BA) for depression within primary care. Databases were searched for trials up to 23rd July 2024. Risk of bias was assessed using the Cochrane risk-of-bias tool, version 2.0.44 studies were included.

CBT, CT, and BA significantly reduced depression symptoms compared to inactive controls (k = 40, g = 0.44, p < .001), but not active comparators (other therapies, medication or exercise) (k = 9, g = −0.06, p = .24). Heterogeneity was significant in studies comparing CBT, CT, or BA to inactive controls, but not in studies using active comparators. Most studies were rated at high risk of bias (36 studies, 81.8 %), predominantly due to use of patient-rated outcome measures in non-blinded studies, lack of ITT analyses, and lack of pre-registering protocols, all of which may result in inflated effect sizes.

Although CBT, CT, or BA appears effective for depression in primary care against usual care or waiting list controls, when compared to active comparators no significant difference is seen, likely a result of variability in the quality of the included studies. Large studies of improved quality (including use of blinded observer-rated outcome measures and ITT analyses) may be required to justify guideline recommendations for CBT over other interventions for depression specifically in primary care.

Open access

 

[Utsikt]

It’s good to see that they criticise the low quality of the studies, but some of the positive ratings of bias seem a bit off:

Cully used subjective outcomes, but was still rated as overall low risk of bias because the questionnaires were filled in by «blinded» observers.

https://psychiatryonline.org/doi/10.1176/appi.ps.20220582

Gilbody has the same problem: using subjective self-reported outcomes where the «assessor» (researcher) as masked to the allocation even though the participant wasn’t.

https://www.thelancet.com/journals/lanhl/article/PIIS2666-7568(23)00238-6/fulltext

There are similar issues with Mansour:

https://oamjms.eu/index.php/mjms/article/view/7883/6688

I have not looked further, but the pattern seems clear. The conclusion is that all of the studies provably had high risk of bias.

 

[rvallee]

Faking it until you actually make it usually ends up with a lot of fakery.

Large studies of improved quality (including use of blinded observer-rated outcome measures and ITT analyses) may be required to justify guideline recommendations for CBT over other interventions for depression specifically in primary care.

Why? Because it has to work? Because it's been asserted to work and by the flying spaghetti monster they have to make it actually work? What other reason can there be to keep on pushing this nonsense? It has been used in standard practice for years, this is a deeply unethical way of doing things. It might not meet the usual definition of medical experimentation but this otherwise is as close to medical experimentation as it gets. Millions of people have been subject to this experimentation, and it's a total bust. There is no reason to try to save face here, and every reason to end this cycle of failure.