Synergistic self-learning approach to establishing individualized treatment rules from multiple benefit outcomes in a calcium supplementation trial
本文提出协同自学习方法，解决钙补充试验中多重结局带来的异质性和缺失数据问题，从而制定个体化治疗规则，指导孕妇补钙以降低铅暴露风险。
In utero lead exposure poses risks to children's neurobehavioral development. The Early Life Exposure in Mexico to ENvironmental Toxicants' calcium supplementation trial studies the effect of calcium supplement in reducing maternal lead exposure to infants during pregnancy. An individualized treatment rule (ITR) is needed to guide pregnant women on taking calcium supplement. This article introduces a statistical learning method, synergistic self-learning (SS-learning), to tackle two challenges in deriving ITR with multiple outcomes, including heterogeneous multidimensional outcomes and complex missing data patterns. Applying SS-learning to the trial, important covariates were identified to form an ITR, expected to lead to higher lead reduction if implemented across the study population.