摘要
目的 :探讨 L ogistic回归模型在尘肺发病预测与控制中的应用。方法 :采用多元Logistic回归统计方法建立粉尘作业工人的接尘工龄 (ET)、工龄平均浓度 (AEC)、粉尘毒性 (T)三因素与尘肺发病概率的回归模型。结果 :尘肺发病预测与控制的回归模型为 :P=1/ { 1+ exp[- (- 5.4 70 7+ 0 .0 94 7ET+ 0 .0 0 2 4 AEC+ 1.9784 T) ]} ,接尘工龄等三因素对尘肺发病影响的比数比分别为 :1.0 994 (ET)、1.0 0 2 4 (AEC)和 7.2 310 (T)。结论 :所建立尘肺发病预测与控制的回归模型与所研究人群的符合率较高 ,对今后预防尘肺发生的科学化管理与决策有较好的实用性和应用价值。
Objective: To probe the application of logistic regression model to the prediction and control of pneumoconiosis. Methods: Multi-logistic regression was used to establish a regression model formed by three factors and pneumoconiosis prevalence probabilities. The three factors are exposure time (ET) of workers to the dust, the average exposure concentration (AEC) by exposure ages, and dust toxicity (T). Results: (1) The regression model for the prediction and control of pneumoconiosis is P=1/｛1+exp[ ( 5.4707+0.0947ET+0.0024AEC+1.9784T)]｝. (2) The odds rates of these three factors affecting the pneumoconiosis prevalence come respectively as follows: 1.0994(ET), 1.0024(AEC) and 7.2310(T). Conclusions: The high conformability of the regression model for the prediction and control of pneumoconiosis with the subject groups suggests its better practicality and application value in the scientific management and the decision-making in the field of pneumoconiosis prevention.
出处
《中国安全科学学报》
CAS
CSCD
2001年第1期40-43,共4页
China Safety Science Journal
关键词
LOGISTIC回归
接尘工
工龄平均浓度
粉尘毒性
预测
控制
尘肺病
Logistic regression Exposure time Average exposure concentration by exposure time Dust toxicity Prediction and control
