I think all of groups B through D are meant to represent general population in some fashion - anyone who’s ever turned up in the emergency room (basically could be anyone), anyone who’s ever turned up in the emergency room or had surgery on their junk for reasons unrelated to troonery (another flavor of “could be anyone”), and anyone who’s ever gone to the doctor for a sore throat (third flavor of “could be anyone”).
The reason why the control groups aren’t pure general population (without these criteria about ER visits, etc.) is because the database they’re analyzing is a database on health insurance claims, so people only show up in the database if they’re visiting a doctor. So the control groups are more or less constructed as “people who have gone to the doctor for the most generic reason possible”, to more or less represent gen pop.
And yep, it was 5-year suicide attempt risk rather than 1-year. I misread that.
I think all of groups B through D are meant to represent general population in some fashion - anyone who’s ever turned up in the emergency room (basically could be anyone), anyone who’s ever turned up in the emergency room or had surgery on their junk for reasons unrelated to troonery (another flavor of “could be anyone”), and anyone who’s ever gone to the doctor for a sore throat (third flavor of “could be anyone”).
The reason why the control groups aren’t pure general population (without these criteria about ER visits, etc.) is because the database they’re analyzing is a database on health insurance claims, so people only show up in the database if they’re visiting a doctor. So the control groups are more or less constructed as “people who have gone to the doctor for the most generic reason possible”, to more or less represent gen pop.
And yep, it was 5-year suicide attempt risk rather than 1-year. I misread that.