Study design in relation to study aim.

Numbers are column percentages, with absolute numbers in parentheses, except for the column “Specifications”, which includes only absolute numbers.aSome cohort studies had a cross-sectional cohort design, which was possible because the predictor values did not change (gender, genes, etc.) or because the study involved a diagnostic prediction model study.bOf the 13 studies that used randomized trial data, 11 were predictor finding or model development studies. Of these 11 studies, five adjusted for the treatment effect, three did not adjust because there was no treatment effect, one did not adjust despite an effective treatment, and in two studies reporting and adjustment for treatment effects was entirely missing.cOne study used two designs: a cross-sectional case-cohort and a cross-sectional nested case-control (here both scored as nested case-control).

除"Specifications（规格说明）"列仅包含绝对数值外，其余列的数值均为列百分比，括号内标注绝对数。
a 部分队列研究采用了横断面队列设计，此类设计可行的原因在于预测变量值固定不变（如性别、基因等），或该研究属于诊断预测模型相关研究。
b 在13项采用随机试验数据的研究中，有11项为预测变量挖掘或模型开发研究。在这11项研究中，5项针对治疗效应进行了校正，3项因未观测到治疗效应而未进行校正，1项虽存在有效治疗效应但未进行校正，另有2项研究完全未报告治疗效应校正相关内容。
c 有1项研究同时采用了两种研究设计：横断面病例队列设计（cross-sectional case-cohort）与横断面嵌套病例对照设计（cross-sectional nested case-control），本次统计中二者均归类为嵌套病例对照研究。