Recovery from chronic fatigue syndrome: a reflexive thematic analysis of experiences of people before, during and after treatment, 2025, Chalder+

[MrMagoo]

I would add a 4th concept: the claims made by the authors are not only unsupported by the data reported in their study, and the other studies they cite, they actively contradict it.

Tuller-ception!

 

[dratalanta]

A paper shall pass the test if it has:
- objective outcomes
- at six months or later
- registered in advance
- which are clinically as well as statistically significant
- with <20% cumulative drop outs / lost to follow up at any point
- and which is accurately summarised in the abstract and conclusion
- incorporating rigorous analysis of the direction of causation
- and acknowledgement of potential attrition bias.

 

[Sean]

Suggested modifications:

- objective and/or adequately controlled (i.e. blinded) subjective outcomes
- final outcome assessment ≥ 6 months after the last treatment session, preferably ≥ 12 months
- registered in advance, with no deviation from the registered protocol for the initial reporting and analysis
- with <20% cumulative drop outs / lost to follow up at any point, and
- reporting of reasons for any missing data, with a sensitivity analysis of its potential consequences.

 

[Midnattsol]

- registered in advance, with no deviation from the registered protocol for the initial reporting and analysis

I find this difficult, you don't necessarily know if the analytic approach you planned can be done with the data you end up having. It might not fit the planned statistical tests assumptions, which may or may not break the analysis if changes are not made.