Open StevenMaude opened 2 years ago
There are multiple caveat emptor warnings around the use of this kind of information in analytic work. General consensus is that it's very hard to draw any conclusions from it at all. If we chose to implement this, we should be sure to document them all thoroughly. @brianmackenna is a person who has a good handle on this.
Consultation_Medium_Used
was implemented for outpatient appointments in https://github.com/opensafely-core/cohort-extractor/pull/644 , following discussion in https://github.com/opensafely-core/cohort-extractor/issues/492
@v1kki213 This has come up frequently in our work with collaborators. Although seemingly straightforward its very complex once we get into it! Briefly there are two main issues
There is a substantial amount of work required in this area and we are keen to collaborate with teams with the skills/experience on it. As an immediate first step @mbfons at NHSX has completed a descriptive project on the use of online consultation codes. This work is going through approvals and @mbfons will share soon!
From https://github.com/opensafely/documentation/issues/430:
@v1kki213 wrote:
Guessing this is
Consultation_Medium_Used
in the TPP database? https://datadictionary.nhs.uk/attributes/consultation_medium_used.html