BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Sensitivity of parameter estimates of marginal and random-effects 
 models to missing data - Rumana Omar\, Department of Statistical Science\,
  UCL
DTSTART:20110315T143000Z
DTEND:20110315T153000Z
UID:TALK28338@talks.cam.ac.uk
CONTACT:Michael Sweeting
DESCRIPTION:Random effects (RE) models and marginal models based on genera
 lised estimating equations (GEE) are frequently used to analyse longitudin
 al repeated measurements health studies where subject dropout is common. R
 E models require MAR assumption. Because marginal models are not based on 
 likelihoods\, they require data to be MCAR. When the data are Gaussian\, t
 he GEEs reduce to score equations and provided the correct correlation str
 ucture is applied the two types of models are equivalent and marginal mode
 ls should then be robust to MAR. However\, equivalent marginal and RE mode
 ls for Gaussian data may not necessarily produce identical parameter estim
 ates due to missing data as GEEs may not reduce to score equations in that
  situation\, even in presence of MCAR. For binary data the marginal models
  require MCAR assumption. By definition neither RE or marginal models are 
 robust to MNAR.\n    \nIn practice\, the extent to which missing observati
 ons cause bias to the parameter estimates of these models and affect their
  clinical and statistical significance is not clear. Limited simulation st
 udies have been conducted. However\, it is not clear what proportion of mi
 ssingness leads to substantial bias or how sensitivity to missing data com
 pares between cluster-level and cluster-varying covariates. It is not know
 n to what extent the marginal model is robust to misspecification of the w
 orking correlation matrix in presence of missing data and whether the stre
 ngth of the intracluster correlation coefficient affects the bias in param
 eter estimates caused by MNAR. The aim here is to explore the effects of d
 ropout on parameter estimates of RE and marginal models for repeated measu
 rements data for both Gaussian and binary outcome using simulation studies
  in order  to make some practical recommendations regarding analyses.    \
 n
LOCATION:Large Seminar Room\, 1st Floor\, Institute of Public Health\, Uni
 versity Forvie Site\, Robinson Way\, Cambridge
END:VEVENT
END:VCALENDAR