Bayesian quantitative bias analysis of misclassification adjustment for prevalence
10.3760/cma.j.cn112338-20240924-00594
- VernacularTitle:流行率错误分类校正的贝叶斯定量偏倚分析
- Author:
Jin LIU
1
;
Shaowen TANG
;
Hui ZHANG
Author Information
1. 南京医科大学第一附属医院临床医学研究院，南京 210029
- Publication Type:Journal Article
- Keywords:
Quantitative bias analysis;
Prevalence;
Misclassification;
Bias adjustment;
Bayesian approach
- From:
Chinese Journal of Epidemiology
2025;46(6):1073-1078
- CountryChina
- Language:Chinese
-
Abstract:
In epidemiological research, accurate estimation of prevalence is important for understanding disease distribution, evaluating the effectiveness of interventions, and allocating health resources. However, the prevalence estimation is often influenced by misclassification bias. Quantitative bias analysis (QBA) can comprehensively evaluate the potential impact of bias on outcomes from three dimensions: bias type, level, and uncertainty. Although QBA research has been developed rapidly in the world in recent years, the introduction of QBA design principles, evaluation methods, and application cases is still insufficient in China. In our previous study, we introduced a new method for adjusting misclassification bias of prevalence and suggested the corresponding analytical tools. Based on the results of previous studies, this paper introduces the principles of QBA design, evaluation indexes, and the application of Bayesian methods in bias adjustment, which provide methodological support for epidemiologists conducting research in this field.
