You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
we're using a different trino base docker image in ehrQL tests to the one used in cohort-extractor. This is so we can use the memory connector in tests, but we should verify that core ehrQL functionality is consistent with the version of trino that EMIS uses (cohort-extractor is still using a presto db URL and the presto-python-client rather than trino
Data checks: some of the data in EMIS looks similar but different to TPP. In cohort-extractor, there appear to be some assumptions that it is the same (e.g. returning values are specified using the same names, such as nuts1_region_name, but actually contain different value ranges). Ones I noticed while working on Add an EMIS schema and backend #797 are:
patient_all_orgs_v2.english_region_name is used to return nuts1_region_name in cohort-extractor, but values do not match names of NUTS1 regions
MSOA in EMIS contains Scottish and Welsh MSOAs (although registered practices for these patients is in England)
#797 added the main tables to support EMIS as a backend. There is still some work to do before studies can be fully runnable in EMIS:
presto-python-client
rather than trinonuts1_region_name
, but actually contain different value ranges). Ones I noticed while working on Add an EMIS schema and backend #797 are:patient_all_orgs_v2.english_region_name
is used to returnnuts1_region_name
in cohort-extractor, but values do not match names of NUTS1 regionsThe text was updated successfully, but these errors were encountered: