[Construction of disease assessment index of silicosis patients].

Objective: To explore the establishment of disease assessment index model in silicosis patients. Methods: In October 2018, 171 silicosis patients who were hospitalized from November 2014 to November 2015 were selected as the study subjects. According to the standard of death risk, the subjects were divided into two groups, including the group without death risk (153 cases) and the group with death risk (18 cases) . Through literature analysis and clinical experience, the variables related to silicosis were preliminarily screened. Multifactorial logistic regression analysis variables were used to analyze the relationship between the variables and the risk of death. The variables associated with the risk of death were selected as the final variables to establish the disease assessment index model. And the receiver operating characteristic (ROC) curve was used to evaluate the clinical application of the disease assessment index. Results: Five variables of Modified British Medical Research Council Respiratory Questionnaire (mMRC) , pulmonary function injury, pneumoconiosis stage, aggravation of the disease and complications were selected as the variables of the disease assessment index, and the assessment index score ranged from 1 to 11 points. The area under the ROC curve of disease assessment index was 0.747 (95%CI: 0.590-0.904) , which could better identify the death risk of silicosis patients. With the increase of disease assessment index score, the death risk of silicosis patients increased. When the cutoff value was 7, the sensitivity and specificity were 0.667 and 0.876, respectively, for the risk of death of silicosis patients. The results of cross-validation showed that the correct discrimination rate of the disease assessment index to the risk of death was 66.7%. Conclusion: The disease assessment index can predict the death risk of patients with silicosis, and can evaluate the disease comprehensively.目的： 探讨矽肺患者病情评估指数模型的建立。 方法： 于2018年10月，选取2014年11月至2015年11月期间住院治疗的矽肺患者作为研究对象，共171例。根据死亡风险标准将研究对象分为两组：无死亡风险组（153例）和有死亡风险组（18例）。通过文献分析及临床资料的收集，初步筛选与矽肺患者病情相关的变量，观察初筛变量的变化。采用多因素logistic回归分析变量与死亡风险的关系，选择与死亡风险有关联的变量作为最终变量，建立病情评估指数模型。用受试者工作特征（ROC）曲线评价病情评估指数的临床实用性。 结果： 《改良版英国医学研究委员会呼吸问卷》（mMRC）评分、肺功能损伤分级、尘肺期别、病情加重和有无并发症5个变量作为矽肺病情评估指数的变量，病情评估指数评分范围1~11分。病情评估指数的ROC曲线下面积0.747（95%CI:0.590~0.904），能较好地判别矽肺患者有无死亡风险；随着病情评估指数评分升高，死亡风险增加。当截断值为7分时，对存在死亡风险的判别效果最好，灵敏度0.667，特异度0.876。病情评估指数对存在死亡风险的判别正确率66.7%。 结论： 病情评估指数可以预测矽肺患者的死亡风险，能够较全面评估病情。.