Update names and types/sub-types in TPMA lookup

[matt-dray]

Update the table according to the spreadsheet sent by @swyatt-SU on 2026-06-29 at 11:35. There's some value changes (e.g. TPMA names) and also new columns (e.g. for sub-types) relative to the one that exists here already.

Consider if this might have knock-on impacts. What if a repo is reading this content and then there is logic related to these groupings? In practice this shouldn't be the case: developers should know that the TPMA code is the unique element to join on. Anyway, most repos do not yet source their TPMA information from this repository. We know that nhp_project_information will be affected: The-Strategy-Unit/nhp_project_information#373.

Note that this is a separate task to #7.

[matt-dray]

@swyatt-SU: some questions related to renaming that could have an impact on how we talk about and use TPMAs across the NHP model and apps and beyond.

To set the scene, the TPMA with code AE-AA-006 is currently called 'A&E Frequent Attenders (Adult, Walk-in)'. The renaming for the first part is 'Frequent ED attenders' with a sub-type of 'Adult, Walk-in'.

So:

1. I think a TPMA 'name' is a unique composite of the two elements, so AE-AA-006 would be renamed to 'Frequent ED attenders (Adult, Walk-in)'. The second element is called a 'sub-type', so is the first element a 'type' (though activity avoidance/efficiencies are sometimes referred to as 'type')?
2. Or is it the case that there are no longer 92 TPMAs, but actually 34 (?) with sub-types? 
3. Should we prefer to use 'ED' rather than 'A&E' from now on whenever referencing this activity type? 

[swyatt-SU]

@matt-dray How about this as an example

TPMA name (34) e.g. Interventions with Limited Evidence
Sub-type name (92) e.g. Interventions with Limited Evidence (General Surgery)
Sub-type short name (92) e.g. General Surgery

In the navigation panel, the TPMA name and sub-type short name would be used in the drop down boxes, but we retain 92 full and unique sub-type names for use elsewhere.

[swyatt-SU]

@matt-dray I don't have a strong view about A&E vs ED. I tend to use ED and I think this is becoming more common. It has the advantage of not having a '&' in it, which I guess is problematic in some circumstances.

[matt-dray]

Thanks Steven.

I'll update the lookup in this repository and then I can crack on with the 'TPMA-explorer' app having two dropdowns: 'TPMA' (34) and 'sub-type'.

If you don't mind, I'd like to continue with 'A&E' because it's already in so many of our user-facing products and is built into a lot of underlying code. We can review this again if 'ED' becomes the overwhelmingly popular choice.

I wonder if this information needs to be shared more widely in the unit, like when 'mitigators' were renamed to 'TPMAs'? Just so we retain consistency in our user-facing products and how we talk about these things.

[swyatt-SU]

Thanks @matt-dray . I will do a briefing about the TPMAs once the app is updated.

[matt-dray]

Just to note, @ghobro: I think it's best to remove the standalone mechanism lookup (JSON) and integrate that info into the TPMA lookup proper (I've done this in #11). It feels like a fundamental grouping of the TPMAs. We can use separate lookups for groupings that are more off-piste or specific to a given project.

[matt-dray]

@swyatt-SU: do you think there's any worth in retaining these columns from the existing TPMA lookup file in this repository:

• mitigator_subset, composed of 'Discharged No Treatment', 'Frequent Attenders', etc
• mitigator_grouping, composed of 'Emergency department and acute medicine activity', 'Hospital activity amenable to public health interventions', etc