[Multiple comparisons, how to adjust the P value]

Zhonghua Nei Ke Za Zhi. 2024 Jan 1;63(1):24-27. doi: 10.3760/cma.j.cn112138-20230807-00038.

[Article in Chinese]

Authors

J Gao¹, X Q Wang²

Affiliations

¹ Nutritional Department, Zhongshan Hospital, Fudan University, Shanghai 200032, China Evidence-Based Medicine of Fudan University, Shanghai 200032, China.
² Department of Hematology,Huashan Hospital, Fudan University, Shanghai 200040, China Evidence-Based Medicine of Fudan University, Shanghai 200032, China.

PMID: 38186113
DOI: 10.3760/cma.j.cn112138-20230807-00038

Abstract

临床研究多重性问题的本质是对同一个问题的回答采用了多次假设检验，这会增加假阳性错误的概率。在不同的临床研究中需要采用不同的策略与方法将临床研究的总体Ⅰ类错误率控制在预设的水平，这一过程称为多重性调整。目前常用的方法主要有两种：Bonferroni法和预设的序贯流程。.