Two advanced methods for adjusting the main coefficient in logistic regression
Two advanced methods for adjusting the main coefficient in logistic regression
复制标题
逻辑回归中调整主系数的两种先进方法
DOI：
10.1007/s00180-011-0294-9
复制
发表时间：
2013
影响因子：
1.3
通讯作者：
C. Fu
中科院分区：
文献类型：
--
作者：
Ya‐Wen Yang;C. Fu
在医学研究中，二元疾病结果通常用连续协变量（例如生化浓度）建模，相应的探索可以采用正常的判别方法。协变量关系影响二元结果和感兴趣的协变量之间的估计关联。偏离拟合值（分数多项式）的值的方法（缩写为 VDFV）可以减少估计偏差，特别是当协变量之间的关系是非线性时。然而，当无关变量与结果相关时，仅出于拟合值的目的，将合并数据（病例和对照）替换为对照数据。基于两种关联模式，即与结果无关的无关变量（I）和与结果相关的无关变量（II），模拟研究表明，VDFV-p（使用汇总数据）是可靠的，模式（I）中偏差较小，均方误差（MSE）较小，而VDFV-c（使用对照数据）在模式（II）中显示偏差较小。传统的协变量调整在 (I) 中表现较差，但在 (II) 中表现相当好。请注意，在 VDFV-p 或 VDFV-c 中从未观察到巨大的 MSE，而这是逻辑回归中与小样本量或稀疏数据相关的常见问题。图示了两项胎儿研究——一项针对模式 (I)，一项针对模式 (II)。
A binary disease outcome is commonly modeled with continuous covariates (e.g., biochemical concentration) in medical research, and the corresponding exploration may employ a normal discrimination approach. The covariate relationship affects the estimated association between binary outcome and the interesting covariate. The method of value deviated from a fitted value (fractional polynomial), which is abbreviated as VDFV, may reduce the estimation bias especially when the relationship between the covariates is nonlinear. However, when the extraneous variable relates to the outcome, the pooled data (cases and controls) are replaced by the control data only for the purpose of fitting values. Based on two association patterns, the extraneous variable unrelated to the outcome (I) and that related to the outcome (II), the simulation study reveals that VDFV-p (using pooled data) is reliable, with less bias and a smaller mean square error (MSE) in pattern (I) and that VDFV-c (using control data) shows less bias in pattern (II). The conventional covariate adjustment performs worse in (I) but fairly well in (II). Note that a huge MSE is never observed in VDFV-p or VDFV-c, while this is a common issue related to small sample size or sparse data in logistic regression. Two fetal studies are illustrated—one for pattern (I) and one for pattern (II).