Yuan William, Beaulieu-Jones Brett K, Yu Kun-Hsing, Lipnick Scott L, Palmer Nathan, Loscalzo Joseph, Cai Tianxi, Kohane Isaac S
Department of Biomedical Informatics, Harvard Medical School, Boston, MA, USA.
Department of Stem Cell and Regenerative Biology, Harvard University, Cambridge, MA, USA.
Nat Commun. 2021 Feb 17;12(1):1107. doi: 10.1038/s41467-021-21390-2.
One of the primary tools that researchers use to predict risk is the case-control study. We identify a flaw, temporal bias, that is specific to and uniquely associated with these studies that occurs when the study period is not representative of the data that clinicians have during the diagnostic process. Temporal bias acts to undermine the validity of predictions by over-emphasizing features close to the outcome of interest. We examine the impact of temporal bias across the medical literature, and highlight examples of exaggerated effect sizes, false-negative predictions, and replication failure. Given the ubiquity and practical advantages of case-control studies, we discuss strategies for estimating the influence of and preventing temporal bias where it exists.
研究人员用于预测风险的主要工具之一是病例对照研究。我们发现了一个缺陷，即时间偏差，它特定于这些研究并与之独特相关，当研究期间不能代表临床医生在诊断过程中所拥有的数据时就会出现。时间偏差会通过过度强调接近感兴趣结果的特征来破坏预测的有效性。我们研究了时间偏差在医学文献中的影响，并突出了效应大小夸大、假阴性预测和重复失败的例子。鉴于病例对照研究的普遍性和实际优势，我们讨论了估计时间偏差影响以及在存在时间偏差的情况下预防它的策略。
