Safety testing time - evaluations - for users of Hodges' model to be extended

The time to test Hodges' model* is to be extended (in my dreams...).

Or. Do not underestimate 'simple' models.

The time to evaluate Hodges' model:

• for individual users
• cross-curricula
• professional groups (e.g. registered mental health; and general nurses)
• within multidisciplinary / skilled teams / venues (Nursing home, Community Mental Health Team)

Given the universal, multicontextual scope of the model for public and patient safety it was decided to extend the time for testing by a further month to two quarters [ 😉] - six whole calendar months.

When applied it is important that checkpoints are agreed:

• Data gathering - initial interview
• initial formulation / care plan
• care review / evaluation
• discharge - staying well 

There is no agreed standard for testing a generic model in healthcare. Is there?

• Not for the medical / bio-medical model
• Nor, for the biopsychosocial model?

What chance then for a biopsychosociopolitical model?

*Per user and group of users of course.

My source: 
Christina Criddle, Concerns raised after OpenAI slashes time taken to test potential risks of latest models, FTWeekend, 12-13 April, 2025, p.14. (italicised terms discussed).

See also:
Radhika Rajkumar, Editor. OpenAI used to test its AI models for months - now it's days. Why that matters, ZDNET.

https://www.zdnet.com/article/openai-used-to-test-its-ai-models-for-months-now-its-days-why-that-matters/