Inputs-data parquets include erroneous 'maternity' mitigators

[matt-dray]

The parquets containing rates, age-sex, procedures and diagnoses data contain these erroneous strategies:

• consultant_to_consultant_reduction_maternity
• convert_to_tele_maternity
• followup_reduction_maternity
• gp_referred_first_attendance_reduction_maternity

These aren't recognised mitigators. Ideally we would prevent these from being generated (unless there's a good reason?).

[tomjemmett]

@yiwen-h I thought the intention on these was that we would not include them as mitigators? IIRC because the thought is this is activity that is likely to grow?

so the question is, do we care that these are being generated (but not used) in the inputs data?

the solution if we wanted to remove would be to insert a filter [https://github.com/The-Strategy-Unit/nhp_data/blob/add_demographics/inputs_data/op/init.py#L39](in the op file) to exclude type == "maternity"

[yiwen-h]

I think we implement this, it's confusing to make things that aren't used

the solution if we wanted to remove would be to insert a filter[https://github.com/The-Strategy-Unit/nhp_data/blob/add_demographics/inputs_data/op/init.py#L39](in the op file) to exclude type == "maternity"