论文部分内容阅读
目的：对一般的临床随访资料构建Ｍａｒｋｏｖ模型，预测各观察点的生存人数和死亡人数。方法：将患者在各观察点的状态表示为生存、死亡、失访三类，作为一类特殊的纵向观测分类数据，采用加权最小二乘法估计未知参数。结果：提出了随访资料的Ｍａｒｋｏｖ模型，并结合实例，给出状态间的转移概率，以及各观察点生存人数、死亡人数的预测值。结论：采用Ｍａｒｋｏｖ模型对临床随访资料进行预测分析可作为寿命表等分析方法的有效补充。
OBJECTIVE: To construct a Markov model for general clinical follow-up data and predict the number of survivors and deaths at each observation point. Methods: The status of patients at each observation point was expressed as three categories of survival, death and loss of follow-up. As a kind of special longitudinal observation classification data, the weighted least square method was used to estimate unknown parameters. Results: The Markov model of follow-up data was proposed. The probability of transition between states and the predicted number of survivors and deaths at each observation point were given with examples. Conclusion: The predictive analysis of clinical follow-up data using Markov model can be used as an effective complement to the analytical methods such as life table.