DSTS - Generalforsamling

Tid og sted

DSTS' generalforsamling afvikles tirsdag den 22. februar kl 17:15.
Institut for Matematisk Fag - Auditorium 10, HCØ, Universitetsparken 5, 2100 København Ø.

Deltag virtuelt via Zoom: https://aaudk.zoom.us/j/69843099224 with passcode: dsts (Meeting ID: 698 4309 9224)

Der vil blive arrangeret middag efterfølgende på en nærliggende restaurant.

Dagsorden

Se vedhæftede dagsorden (link nederst). Forslag til opdaterede vedtægter ligeledes vedhæftet.

Efter generalforsamlingen (og inden spisning) følger nedenstående foredrag.

Talk

Henrik Støvring, DrMedSci, PhD, MSc
Department of Public Health - Department of Biostatistics
Aarhus University

Estimation of prescription duration from dispensing dates – challenges and opportunities with the Waiting Time Distribution

Pharmacoepidemiologic databases record dispensings of medication to patients with information on date, patient and amount dispensed, but not on prescription duration, i.e. the duration of treatment after the dispensing (“the time until the pill glass is empty”). Pharmacoepidemiological data are essential for determining medication exposure in large studies aimed at estimating effects and side-effects of medications in real-world settings, sometimes called post-marketing studies or phase IV studies. It is common practice in these studies to employ simple decision rules regarding occurrence of medication dispensings within a fixed time window to determine prescription duration, although such decision rules are prone to bias and lack theoretical underpinnings. In this presentation, I will review some of the fundamental challenges to estimation of prescription durations before I introduce the parametric Waiting Time Distribution model, which circumvents the need for decision rules. Further, I will show how it allows estimation of prescription durations using regression-like statistical techniques and I will present some of our recent developments of the method. First, to improve its statistical efficiency and, second, to integrate it with models for exposure and outcome in pharmacoepidemiologic case-control studies.