We currently have two columns for the treatment function in our data extracts. Initially, we took the tretspef column from HES and grouped any non-RTT treatment function into 3 other categories (Medical, Surgical, and everything else).
This column is used in a couple of the steps, such as baseline adjustment: users can set values per treatment function at the grouped level.
But then we needed to have the ungrouped treatment function in some aggregations. As a kludgey fix, we introduced tretspef_raw, which is the unaggregated column.
This is something that should be cleaned up, keeping just a single column in the data extracts. The aggregated column is only really needed during model runtime.
If we change this it will change the name of the aggregation, but that should make things clearer (if anyone tried to lookup the field name in the HES data dictionary it could cause confusion).
See The-Strategy-Unit/nhp_model#258, The-Strategy-Unit/nhp_model#255
This will impact:
We currently have two columns for the treatment function in our data extracts. Initially, we took the
tretspefcolumn from HES and grouped any non-RTT treatment function into 3 other categories (Medical, Surgical, and everything else).This column is used in a couple of the steps, such as baseline adjustment: users can set values per treatment function at the grouped level.
But then we needed to have the ungrouped treatment function in some aggregations. As a kludgey fix, we introduced
tretspef_raw, which is the unaggregated column.This is something that should be cleaned up, keeping just a single column in the data extracts. The aggregated column is only really needed during model runtime.
If we change this it will change the name of the aggregation, but that should make things clearer (if anyone tried to lookup the field name in the HES data dictionary it could cause confusion).
See The-Strategy-Unit/nhp_model#258, The-Strategy-Unit/nhp_model#255
This will impact: