BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Dropouts in the AB/BA crossover design - John Matthews\, School of
  Mathematics and Statistics\, Newcastle University
DTSTART:20110215T143000Z
DTEND:20110215T153000Z
UID:TALK28337@talks.cam.ac.uk
CONTACT:Michael Sweeting
DESCRIPTION:Missing data are a problem in all forms of clinical trial and 
 many solutions to the problem have been proposed for parallel group trials
 . Moreover\,  the last two decades has seen much research into the issues 
 surrounding missing data not just in clinical trials but for many forms of
  statistical analysis and some of these methods have been particularly sui
 ted to longitudinal data.  However\, surprisingly little work has been don
 e on the specific problems of dropouts in crossover studies.  In this semi
 nar inverse probability weighting methods will be applied to the AB/BA des
 ign when some patients dropout in period two.  The use of common assumptio
 ns for this design to enable the assessment of missing not at random analy
 ses is explored.  The necessity of sophisticated methodology at all is als
 o questioned\, as complete case analyses may have more merit in this appli
 cation than is commonly supposed.\n\nThe seminar reports joint work with W
 eang Kee Ho\, Rob Henderson\, Lauren Rodgers and Daniel Farewell.\n
LOCATION:Large Seminar Room\, 1st Floor\, Institute of Public Health\, Uni
 versity Forvie Site\, Robinson Way\, Cambridge
END:VEVENT
END:VCALENDAR