Lindsey J K, Jones B, Lewis J A
Department of Biostatistics, Limburgs Universitair Centrum, Diepenbeek, Belgium.
Stat Med. 1996 Mar 15;15(5):527-35. doi: 10.1002/(SICI)1097-0258(19960315)15:5<527::AID-SIM179>3.0.CO;2-K.
Survival models and cross-over designs both have an established place in biomedical research. Surprisingly, there are a few examples of proper exploitation of two in combination. A number of advantages and disadvantages of such studies are discussed. Two examples are used to illustrate the application of semi-Markov models with time-varying covariates, as standard log-linear models, to such data.
生存模型和交叉设计在生物医学研究中都有既定的地位。令人惊讶的是，将两者结合起来合理运用的例子却很少。本文讨论了此类研究的一些优缺点。文中使用两个例子来说明具有时变协变量的半马尔可夫模型（作为标准对数线性模型）在此类数据中的应用。
