Data cleaning for insulin regimen

[KonradUdoHannes]

Based on information provided, data cleaning aims for the following goal for Insulin regimen.

• Insulin regiment should be split into a main category and a sub category. In newer tracker files these are separated by a ,, whereas in older tracker files sub category does not exist.
• The allowed values for the main category are
  • Basal-bolus (MDI)
  • Premixed 30/70 BD
  • Self-mixed BD
  • Modified conventional TID
• The sub category should have the allowed values
  • For main category Basal-bolus (MDI) it can be Basal-bolus MDI (AN) or Basal-bolus MDI (HI)
  • Nothing for other categories

[KonradUdoHannes]

The mapping of the main category is done by doing a case insensitive matching on "basal", "premixed", "self-mixed" and "conventional" respectively.
In the tested tracker files there seems to be a few cases where sub-category information is mixed with category information and where the category is "Basal-bolus MDI (AN)" whereas the sub-category is only "Basal-bolus". Such a situation currently leads to the "(AN)" information getting lost. It should be clarified whether attempts should be made to handle such edge cases, or whether categories and sub-categories in future tracker fiels will be cleare, i.e. not mixed up.