Open sebbacon opened 4 years ago
Discussed yesterday with Chris (TPP) and LSHTM. Long standing issues with identifying pregnancy in the record as every time a woman receives antenatal care they have a pregnancy code added on and therefore the length of pregnancy appears artificially long. We could look at those who had births but that would have to be retrospectively cause they haven't given birth yet if they are currently pregnant.
LSHTM have suggested we use:
population prevalence estimates as they have age-stratified population prevalence using CPRD pregnancy register from 5th March 2014
Likely to be a rough estimation. Should we drop? @alexwalkercebm
Code for identifying pregnancy - could be binary (i.e you are pregnant or not) or could be in trimester
Would need to incorporate looking for codes for 'delivery/birth' or 'miscarriage' or 'termination'