论文部分内容阅读
我们针对肺癌的各种检查方法设计了以医院人群为基础的病例对照研究。应用Ｍａｎｔｅｌ－Ｈａｅｎｓｚｅｌ及非条件ｌｏｇｉｓｔｉｃ回归方程进行了单因素及多因素分析。结果表明：在研究对象中，吸烟史、家族史、肿瘤史、Ｘ线检查、ＣＴ检查与肺癌的关系在统计学上有显著意义。ＰＡＴ痰检、ＣＴ检查、Ｘ线检查、主要症状是影响肺癌临床诊断的主要因素。建立肺癌模型，回代结果显示：灵敏度为９３．９２％，特异度为７３．２８％，准确度９０．１９％。这对提高肺癌临床诊断水平，减少漏诊率有着积极的作用。
We designed case-control studies based on hospital populations for various lung cancer screening tests. Univariate and multivariate analyses were performed using Mantel-Haenszel and unconditional logistic regression equations. The results showed that among the subjects studied, the relationship between smoking history, family history, tumor history, X-ray examination, CT examination and lung cancer was statistically significant. PAT sputum examination, CT examination, X-ray examination, and main symptoms are the main factors affecting the clinical diagnosis of lung cancer. The lung cancer model was established. The results of the retrospective analysis showed that the sensitivity was 93.92%, the specificity was 73.28%, and the accuracy was 90.19%. This has a positive effect on improving the clinical diagnosis of lung cancer and reducing the rate of missed diagnosis.