Penalized Spline of Propensity Methods for Treatment Comparison

Valid causal inference from observational studies requires controlling for confounders. When time-dependent confounders are present that serve as mediators of treatment effects and affect future treatment assignment, standard regression methods for controlling for confounders fail. Similar issues also arise in trials with sequential randomization, when randomization at later time points is based on intermediate outcomes from earlier randomized assignments. We propose a robust multiple imputation-based approach to causal inference in this setting called penalized spline of propensity methods for treatment comparison (PENCOMP), which builds on the penalized spline of propensity prediction method for missing data problems. PENCOMP estimates causal effects by imputing missing potential outcomes with flexible spline models and draws inference based on imputed and observed outcomes. Under the SUTVA, positivity, and ignorability assumptions, PENCOMP has a double robustness property for causal effects. Simulations suggest that it tends to outperform doubly robust marginal structural modeling when the weights are variable. We apply our method to the multicenter AIDS cohort study to estimate the effect of antiretroviral treatment on CD4 counts in HIV-infected patients. Supplementary materials for this article are available online. Code submitted with this article was checked by an Associate Editor for Reproducibility and is available as an online supplement.

基于观察性研究开展有效的因果推断，需对混杂因素加以控制。当存在既作为治疗效应中介变量、又会影响后续治疗分配的时变混杂因素时，标准的混杂因素控制回归方法将失效。此类问题同样会出现在顺序随机化试验中：当后续时间点的随机化基于早期随机分组产生的中间结局时，便会出现该类困境。针对该场景，我们提出一种基于多重插补的稳健因果推断方法，名为治疗比较倾向得分惩罚样条法（PENCOMP），该方法依托针对缺失数据问题的倾向得分惩罚样条预测方法构建而来。PENCOMP通过灵活的样条模型对缺失的潜在结局进行插补以估计因果效应，并基于插补所得与实际观测到的结局开展统计推断。在满足稳定单位治疗值假设（SUTVA）、正性假设（positivity）与可忽略性假设（ignorability）的前提下，PENCOMP具备因果效应估计的双重稳健性。模拟实验结果表明，当权重存在变异时，该方法的表现往往优于双重稳健边际结构模型。我们将该方法应用至多中心艾滋病队列研究，以此评估抗逆转录病毒治疗对HIV感染者CD4细胞计数的影响。本文的补充材料可在线获取。随本文提交的代码已由可重复性副主编审核，同样可作为在线补充材料获取。