Thread by @EinsteinsAttic

Noting that, just like the #AppsLibrary, #DTAC *still* doesn't have a required section on clinical #efficacy - and that #impacts are much more than just (avoiding) #risks - alongside focusing on the #algorithms and #data, have you considered asking a few more basic questions...

https://twitter.com/NHSX/status/1370425661358272516

...such as:

1) Will the proposed intervention have a direct impact on health, mental health and/or wellbeing?

If so, what specifically is the intended impact and how will it be measured? If not, why are you doing it?

2) Will the proposed intervention interact with or have...

...any impacts on social, economic and/or environmental factors that would 'indirectly' affect health?

If so, what are they and how will they be measured - and, if necessary, mitigated? If not, provide evidence of ALL of the potential impacts you have actually considered...

3) Will the proposed intervention affect an individual’s ability to improve their own health and wellbeing?

Goes to agency, consent, dignity, etc.

4) Is the proposed intervention likely to change access to (or demand for) health and/or care services?

Goes to exclusion...

...education, accessibility (and public value), etc.

N.B. "proposed intervention" above = #algorithm, #AI, etc. = #MedicalDevices

__
For those unaware of #DTAC, there's more info here:

nhsx.nhs.uk/key-tools-and-…