BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4//
BEGIN:VEVENT
UID:20251025T224022EDT-7378RaoF7A@132.216.98.100
DTSTAMP:20251026T024022Z
DESCRIPTION:Abstract:     The use of longitudinal data for predicting a sub
 sequent binary event is often the focus of diagnostic studies. This is par
 ticularly important in obstetrics\, where ultrasound measurements taken du
 ring fetal development may be useful for predicting various poor pregnancy
  outcomes. The focus of this paper is on developing a class of joint model
 s for the longitudinal measurements and binary events that can be used for
  prediction. A shared random parameter model is proposed for linking the t
 wo processes together. Under a Gaussian random effects assumption\, the ap
 proach is simple to implement with standard statistical software. Under th
 is assumption\, it is also easy to develop a predictor using multivariate 
 or high dimensional longitudinal data. Using asymptotic and simulation res
 ults\, we show that estimates of predictive accuracy under a Gaussian rand
 om effects distribution are robust to severe misspecification of this dist
 ribution. However\, under some circumstances\, estimates of individual ris
 k may be sensitive to severe random effects misspecification. We illustrat
 e the methodology with data from a longitudinal fetal growth study.  If ti
 me permits\, I will also discuss some statistical issues encountered in th
 e on-going NICHD fetal growth studies.   
DTSTART:20130219T210000Z
DTEND:20130219T220000Z
LOCATION:Room 25\, Purvis Hall\, CA\, QC\, Montreal\, H3A 1A2\, 1020 avenue
  des Pins Ouest
SUMMARY:Biostatistics Seminar - Dr Paul Albert - Joint models for multivari
 ate longitudinal measurements and a binary event: An application to predic
 ting poor pregnancy outcomes from longitudinal ultrasound measurements
URL:/epi-biostat-occh/channels/event/biostatistics-sem
 inar-dr-paul-albert-joint-models-multivariate-longitudinal-measureme
END:VEVENT
END:VCALENDAR