利用空间自回归模型分析四川省肺结核登记率影响因素

Analysisof Influencing Factors for the Registration Rate of Pulmonary Tuberculosis in Sichuan Province Using the Spatial Autoregressive Model

目的建立空间自回归模型,分析社会经济、资源分配等因素与肺结核登记率之间的关系,为制定科学的疾病防控措施提供参考依据。方法收集四川省2014-2016年各县区分年报告登记的活动性肺结核、涂阳肺结核、复治涂阳肺结核患者数、当地当年的人口数,选取城镇化率、人均地区生产总值、人均公共财政投入3个因素做影响因素分析,采用Geo Da软件建立一阶空间自回归模型、空间误差回归模型、空间滞后回归模型,比较拟合优度选择最优模型,获得具有统计意义的变量,即有参考价值的空间分布影响因素。结果 2014年活动性肺结核,城镇化率对登记率影响是负向的,人均地区生产总值影响是正向的,涂阳肺结核,城镇化率对登记率影响是负向的,复治涂阳肺结核,人均公共财政投入对登记率的影响是负向的;2015年活动性和涂阳肺结核,城镇化率对登记率的影响均是负向的,复治涂阳肺结核,所有分析因素均对登记率的影响无统计学意义;2016年活动性和复治涂阳肺结核,所有分析因素均对登记率的影响无统计学意义,涂阳肺结核,城镇化率对登记率的影响是负向的。结论空间自回归模型分析的主要结果表明城镇化率与登记率呈负相关,提示须积极推进城乡一体化建设,改善民生,增进健康。

Objective To analyze the relationship between social and economic factors and resource allocation and the registration rate of pulmonary tuberculosis（ TB） cases using the spatial autoregressive model so as to provide a reference basis for developing scientific measures for disease prevention and control. Methods Numbers of patients with active TB,smear-positive TB and retreated smear-positive TB reported and registered each year in each county of Sichuan Province during 2014 to 2016 and the local population in the corresponding year were collected. Influencing factors were analyzed with the involvement of the per capita GDP,urbanization rate and per capita public financial investment. First order spatial autoregressive model,spatial error model and spatial lag model were established using Geo Da software,the optimal model was selected by comparing the goodness of fit,and the statistically significant variables,namely the influencing factors for space distribution with a reference value,were obtained. Results In 2014,for active TB,the urbanization rate had a negative impact on the registration rate,while the per capita GDP had a positive impact; for smear-positive TB,the urbanization rate had a negative impact on the registration rate; for retreated smear-positive TB,the per capita public financial investment had a negative impact on the registration rate. In 2015,for both active TB and smear-positive TB,the urbanization rate had a negative impact on the registration rate; for retreated smear-positive TB,none analyzed factor had a statistically significant impact on the registration rate. In 2016,for active TB and retreated smear-positive TB,none analyzed factor had a statistically significant impact on the registration rate; for smear-positive TB,the urbanization rate had a negative impact on the registration rate.Conclusion The main results of analysis using the spatial autoregressive model show that the urbanization rate is negatively related to the registration rate,suggesting it is necessary to actively promote the integration of urban and rural areas,improve people＇s livelihood and their health.