| Summary: | This research aims to estimate the transition probabilities of lives becoming disabled and the extent to which they are disabled using the approach of functional Markov model. The transition probability from one disabled state to a more or less severely disabled state is best estimated using longitudinal data in which the change in the health status of each respondent can be monitored over one or more years. Such data are limited in Malaysia, typically covering only a smaller area of the nation. The functional Markov model overcomes such data limitations, using cross-sectional data which measures the disability status of individuals only at one point in time and build a functional form for the transition probabilities in a multiple state model. Results suggested that multiple state model's prevalence rates replicated the Malaysian prevalence rates quite well, indicating that the parameters of the probability of deterioration had been estimated accurately with sum squared of errors less than 5% for almost all age groups and disability levels. Furthermore, severely disabled elderlies, especially among the oldest age group, have the highest probability to die compared to less severely disabled elderlies.
|
|---|