search

oxford-pharmacoepi / MegaStudy

4 stars 2 forks source link

DUS Shiny: drug exposure #49

Closed michaelsfr closed 6 months ago

michaelsfr commented 7 months ago

Hello everyone, when checking the content of the DUS shiny code, we noticed that there are no or only partial meaningful values under "Charact. incident patients > Drug use > Number exposures" (see screenshot).

image

The same applies to initial quality /cumulative quality. In addition, under the Indication tab, there are also only 0 values. Could this happen due to a mapping problem?

I can provide out results in a dedicated directory in the EHDEN Workspace if necessary.

tiozab commented 6 months ago

@michaelsfr thanks for reaching out. To explain how these numbers come about: we looked at an incident exposure (a single exposure) and this single exposure was only joined with other exposures if there was no gap in between the two exposures (this was our definition of "an exposure"). We had to be very strict because of the variety of drugs in the study. Sometimes it would makes sense to allow a grace period, but in this case, we were really only interested in the first exposure, therefore, to see the no. 1 most of the times makes total sense. Does that explain?

michaelsfr commented 6 months ago

@tiozab Got it, thanks for the explanation!