摘 要: 目的探讨颈痛现况调查中OR值的使用方法。方法采用整群随机抽样的方法,对某汽车厂3 800名工人进行颈痛及危险因素问卷调查。利用Logistic回归分析计算各变量的OR值,以log-binomial模型计算现患比(PR值),计算OR值与PR值的相对偏差值:(OR-PR)/PR×100%,以相对偏差的大小衡量危险度被高估的程度。结果OR值对各研究因素危险度的高估程度在13.3%~70.9%之间;且高估的程度随OR值的增加而上升,呈线性相关(r=0.853,P=0.002)。结论使用OR值时,影响因素的危险度在一定程度内被高估,OR值较高时应考虑使用PR值等替代指标。 |
关键词: 颈痛 现况调查 OR 值 现患比 |
中图分类号: R68
文献标识码:
|
基金项目: 国家“十二·五”科技支撑项目(编号:2014AI12B03); |
|
Odds Ratio in cross⁃sectional study of neck pain |
WU Jia⁃bing, QI Cheng, LING Rui⁃jie
|
WU Jia⁃bing, QI Cheng, LING Rui⁃jie
|
Abstract: Objective To explore the using of Odds Ratio (OR) in cross⁃sectional studies concerning neck pain. Methods A questionnaire survey was conducted among 3800 workers recruited by cluster random sampling in an automobile manufacturing corporation. OR and Prevalence Ratio (PR) of risk factors for neck pain were estimated using Logistic regression and log⁃binomial models, respectively. The relative deviation of OR and PR of every risk factor was calculated. Results Compared with PR, the OR overestimated the risk of influence factors by 13 3% ~70 9%; and the extent of overestimation rose with the ascending of OR (linear correlation, r =0.853, P =0.02). Conclusion The risk was overestimated when OR was used in crosssectional study on neck pain, and substitute indicators such as PR should be considered. |
Keywords: neck pain cross⁃sectional study Odds Ratio Prevalence Ratio |