senegal

[abhidg]

• parsers: add ccp-senegal
• schemas: allow date and start_date to include time
• schemas(observation): add loss of smell or taste

adtl isaric/parsers/ccp-senegal.toml ../isaric-data/ccp-senegal/CVRZVOV_GLOBAL_Changes.csv

table	valid	total	percentage_valid
subject	200	200	100.000000%
visit	200	200	100.000000%
observation	25011	25019	99.968024%

observation

• 6: data must contain ['phase', 'date', 'name'] properties
• 2: data must be valid exactly by one definition (0 matches found)

[pipliggins]

All the drugs administered (antivirals, antibiotics, steroids etc) appear to be stored in tmt_pdhosp_nom{n} variables (of which there are 20ish in the data). Any thoughts about how to include these? We'll have to work backwards from finding names and inferring that antivirals etc. were administered I think.

It's free text data, but appears to have been somewhat cleaned, it's all in caps so as a stop gap we could for example search for 'HYDROCORTISONE' in those fields, and if present then treatment_corticosteroid will be True. It's a long way from catching everything but prior to getting NLP in there that might be as good as we can do.

[abhidg]

Will have a think, best to programmatically generate these mappings somehow. There are also some names in French and spelling errors.

[ekamau]

looks good!