临床依从性测量的离散型Markov模型

Discrete Markov Models for Measurement of Clinical Compliance

从单个患者的依从性内在关系出发,用离散型Markov链来模拟依从性关系,并通过两个常见的依从性测量指——依从天数百分比和药物服用剂量百分比,对患者的依从性给出定量的估计。同时,给出了估计量的渐近分布和估计值的标准误差。

To measure compliance of unique subject, two compliance indices： the percentage of compliant days and the percentage of doses taken, are introduced. A stationary Markov chain is applied to model the intrinsic dependence structure of clinical compliance, and the estimating of compliance is obtained by this model. Moreover, the asymptotic distribution and standard error of the estimate are presented in this paper.