摘要
目的 为有缺失数据的生物等效性评价资料的处理提供一种实用的统计方法。方法 构建两阶段交叉设计的随机效应模型 ,使用WinBUGS软件由MCMC方法给出两种药物的基于贝叶斯理论的生物等效性评价。结果 对完整数据和有缺失数据的情形 ,MCMC方法分别计算出了待评价指标落于等效区间的后验概率从而做出了生物等效性评价 ,结论与文献报道一致。结论 MCMC方法可以方便地给出生物等效性评价的贝叶斯推断结果 ,能有效处理有缺失数据的试验资料并提供缺失数据的估计值 (或分布 )。
Objective To provide a useful statistical method for bioequivalence assessment with missing data.Methods Constructing random effect model for the data of crossover design and using the software WinBUGS, bioequivalence assessment based on bayesian theory was made through MCMC.Results For complete data and missing data, MCMC calculated the posterior probabilities for each situations. Bioequivalence assessment was done and the result was the same as the related report.Conclusion MCMC is a effective solution for bayesian method in bioequivalence assessment with missing data.
出处
《中国卫生统计》
CSCD
北大核心
2004年第4期207-208,211,共3页
Chinese Journal of Health Statistics
