BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Understanding longitudinal risk factor analysis in asymptomatic di
 sease - a role for artificial datasets? - Nomdo Jansonius - University Med
 ical Centre Groningen/Rotterdam Study
DTSTART:20130419T110000Z
DTEND:20130419T120000Z
UID:TALK44611@talks.cam.ac.uk
CONTACT:Steph Moore
DESCRIPTION:In population-based epidemiology\, large cohorts are followed 
 for long periods. Regular assessments are performed but they are often wid
 ely spaced in time\, making the exact timing of the onset of asymptomatic 
 disease troublesome. Also\, the follow-up duration may differ considerably
  between participants. Hence\, neither Cox regression nor logistic regress
 ion is optimal for risk factor analysis. Despite large sample sizes\, conf
 idence intervals of risk estimates resulting from either Cox or logistic r
 egression are often very wide\, hampering subgroup analyses and the assess
 ment of interactions. Use of an artificially generated dataset instead of 
 data from a real cohort gives full control of all relevant aspects\, enabl
 ing a better understanding of observations done in real data. The developm
 ent of such an artificial dataset for glaucoma and some applications will 
 be presented.\n\nNomdo Jansonius is professor of ophthalmology at the Univ
 ersity Medical Centre Groningen in the Netherlands and is principal invest
 igator for the glaucoma arm of the Rotterdam Study.  His background includ
 es masters degrees in physics and medicine\, a PhD undertaken jointly betw
 een the University of Groningen and University of Cambridge\, and post-doc
 toral training in epidemiology and statistics.  His current research inter
 ests include glaucoma\, perimetry and physiological optics. \n\nLunch will
  be provided\n
LOCATION:New Seminar Rooms\, Strangeways Research Laboratory
END:VEVENT
END:VCALENDAR