Construction and application of Markov model for follow-up data

摘要

Objective: to construct Markov model for clinical follow-up data and predict the number of survivals and deaths. Methods: the state of patients was classified into survival, death and censorship. The weighted least squares method was used for estimating parameters. Results: Markov mole for surival analysis of follow-up data was presented. By using an example, the transition probability matrices were obtained and the number of surivals and deaths at each observation point was predicted respectively. Conclusion: Markov model constructed in the pre- sent study to analyze clinical follow-up data could be used as effective supplemention for life table analysis.