A LONGITUDINAL APPROACH TO MODELING OUTCOMES IN A NURSING-HOME STUDY

Recent developments in longitudinal statistical methodology have improved our ability to model dynamic processes such as adaptation to nursing homes. Longitudinal observations provide information on individual patterns of change and factors affecting them. However, longitudinal analyses are often complicated by unequal periods of observation and individual variability in patterns of change. This paper demonstrates the use of a linear mixed-effects model to study adaptation in a longitudinal nursing home study with different numbers of repeated measurements for each individual becase of discharges, transfers, and mortality.