Additional file 2: of Multimorbidity states associated with higher mortality rates in organ dysfunction and sepsis: a data-driven analysis in critical care

Supplementary methods. Figure S1. Graph summary of ROC curves depicting predictive performance of LCA input variables at differentiating a given subgroup from the remaining subgroups. Table S2. Table summary of classifier performance (assessed using area under the receiver operating curve, AUC) at predicting subgroup membership based on the input variables used in the LCA (age, sex, type of admission, and morbidities). Figure S2. Violin plot summary of organ systems impairment in the multimorbidity subgroups. Figure S3. Network summary of the multimorbidity subgroups with lower rates of sepsis and death. (ZIP 524 kb)

补充方法。
补充图S1：展示受试者工作特征曲线（Receiver Operating Characteristic, ROC）的图形汇总，用于描述潜在类别分析（Latent Class Analysis, LCA）的输入变量在区分某一亚组与其余亚组时的预测性能。
补充表S2：以表格形式汇总基于LCA所用输入变量（年龄、性别、入院类型与共病情况）预测亚组归属的分类器性能，该性能以受试者工作特征曲线下面积（Area Under the Receiver Operating Characteristic Curve, AUC）作为评估指标。
补充图S2：展示多共病亚组器官系统损害情况的小提琴图汇总。
补充图S3：展示败血症与死亡率更低的多共病亚组的网络汇总图。
（ZIP 524 KB）