定量偏倚分析能够有效估计潜在偏倚大小和方向,帮助研究人员评估结果的稳健性。为了提升学生专业能力,自2013年起,复旦大学公共卫生学院面向医学相关专业的硕士研究生和博士研究生开设流行病学偏倚分析技术课程,融合理论讲授、案例分析与计算机操作,旨在将理论知识与实际应用有机结合。2020至2024年75份学生汇报幻灯片和Take-home试卷分析结果显示,大部分学生能够准确识别偏倚、构建因果图并合理运用定量偏倚分析方法,考核成绩与学生的流行病学及卫生统计学专业背景密切相关。2024年匿名教学效果问卷调查结果(共回收26份有效问卷)显示,学生对课程的满意度评分为(9.0±1.3)分,理论掌握自评分为(8.3±1.4)分,技术运用自评分为(7.8±1.7)分,10名(38.4%)学生计划将偏倚分析纳入课题或论文。课程在提升学生偏倚分析能力方面取得积极效果,但技术运用仍需加强。未来将增加实践比重、优化教学内容,以持续提升教学质量。
Quantitative bias analysis can accurately estimate the magnitude and direction of potential biases, thereby helping researchers assess the robustness of findings against potential biases. To enhance students' professional competencies, the School of Public Health in Fudan University has developed a course on epidemiological bias analysis techniques since 2013 to master's and doctoral students in medical-related disciplines. The curriculum integrates theoretical lectures, case studies, and computer-based operations to combine theoretical knowledge with practical application. Analysis of 75 student presentations and take-home exams from 2020 to 2024 indicated that most students could accurately identify bias types, construct causal diagrams, and appropriately apply quantitative bias analysis methods. Examination performance was strongly associated with students' backgrounds in epidemiology and biostatistics. Results from an anonymous online teaching evaluation survey conducted in 2024 (26 valid responses) indicated overall satisfaction score of 9.0±1.3, self-rated theoretical mastery score of 8.3±1.4, self-rated technical application score of 7.8±1.7, and 10 students (38.4%) reported plans to incorporate bias analysis into their research projects or theses. This course has demonstrated positive outcomes in improving students' abilities in bias analysis, though practical application skills warrant further strengthening. Future course development will increase practical exercises and refine course content to continuously improve teaching quality.
[1] Fox MP, MacLehose RF,Lash TL. Applying quantitative bias analysis to epidemiologic data[M]. New York:Springer,2021:13-32.
[2] Brown JP, Hunnicutt JN, Ali MS, et al. Quantifying possible bias in clinical and epidemiological studies with quantitative bias analysis: common approaches and limitations[J]. BMJ, 2024,385:e076365. DOI: 10.1136/bmj-2023-076365.
[3] Petersen JM, Ranker LR, Barnard-Mayers R, et al. A systematic review of quantitative bias analysis applied to epidemiological research[J]. Int J Epidemiol, 2021,50(5):1708-1730. DOI: 10.1093/ije/dyab061.
[4] Davis LE, Banack HR, Calderon-Anyosa R, et al. Probabilistic bias analysis for exposure misclassification of household income by neighbourhood in a cohort of individuals with colorectal cancer[J]. Int J Epidemiol, 2024,53(6): dyae135.DOI: 10.1093/ije/dyae135.
[5] Shi L, Zhu Q, Wang Y, et al. Incident dementia and long-term exposure to constituents of fine particle air pollution: a national cohort study in the United States[J]. Proc Natl Acad Sci USA, 2023,120(1):e2211282119. DOI: 10.1073/pnas.2211282119.
[6] Pearl J. Probabilistic reasoning in intelligence systems[M].San Francisco, CA:Morgan Kaufmann,1988:1-466.
[7] Greenland S, Pearl J, Robins JM. Causal diagrams for epidemiologic research[J]. Epidemiology, 1999,10(1):37-48.
[8] University of Toronto. Advanced analytic methods for bias in epidemiologic studies[EB/OL]. https://www.dlsph.utoronto.ca/course/advanced-analytic-methods-for-bias-in-epidemiologic-studies/.
[9] Fox M. Quantitative bias analysis methods for epidemiologic research[EB/OL].[2025-07-11].https://www.bu.edu/academics/sph/courses/sph-ep-861/.
[10] Mora A, Kiang C, Lash T. Applied quantitative bias analysis to epidemiological research[EB/OL].[2025-07-11]. https://sph.emory.edu/degrees-programs/msph/epidemiology/key-courses.
[11] Hughes R,Kawabata E. Introduction to quantitative bias analysis[EB/OL].[2025-07-11].https://www.bristol.ac.uk/medical-school/study/short-courses/courses/quantitative-bias-analysis/.
[12] Blakely T, Aitken Z. Contemporary epidemiology[EB/OL].[2025-07-11]. https://ecommerce.unimelb.edu.au/aea-contemporary-epidemiology-course-1.
[13] Henan M.Causal diagrams[EB/OL].[2025-07-11]. https://pll.harvard.edu/course/causal-diagrams-draw-your-assumptions-your-conclusions.
[14] Shahar E, Shahar DJ. Causal diagrams and three pairs of biases[M]//Lunet N. Epidemiology-current perspectives on research and practice. Rijeka: IntechOpen, 2012:31-62.
