首页|基于地理加权回归模型的山东省临沂市结核病发病情况及其影响因素分析

基于地理加权回归模型的山东省临沂市结核病发病情况及其影响因素分析

扫码查看
目的 探讨山东省临沂市各县区结核病发病登记率现况及其与相关影响因素的空间局域关系。方法 收集2019年1-12月临沂市12个县区的结核病发病及相关影响因素资料;采用Geoda 4。02软件分析空间自相关性;采用GWR 4。09软件构建最小二乘法(OLS)和地理加权回归(GWR)模型;采用ArcGIS 10。2软件绘制GWR模型影响因素的空间分布地图。结果 2019年临沂市结核病发病总计3 261例,总发病率为32。39/10万。不同县区、不同性别及年龄组之间结核病发病率差异均有统计学意义(x2=71。257~1 532。464,P<0。001)。空间自相关分析结果显示为空间正相关(Moran I=0。234,P=0。021)。LISA聚类图结果显示,罗庄区和兰陵县为低-低聚集,莒南县为高-高聚集。OLS和GWR结果显示GWR模型各种诊断标准均优于OLS模型。GWR模型结果显示人口密度(β=-0。019)、人均受教育年限(β=-14。509)、每万人平均卫生技术人员数(β=-0。293)、空气质量指数(β=-15。438)对结核病发病率影响具有统计学意义(P<0。05)。结论 山东省临沂市2019年结核病发病率存在性别、年龄、地域差异。各县区结核病发病率存在空间聚集性。社会、环境、经济等因素对结核病发病率影响存在空间异质性。GWR模型较OLS模型对发病率影响因素的分析上具有独特优势性。
Incidence rate of tuberculosis and related influencing factors in Linyi City,Shandong Province,China:An analysis based on the geographical weighted regression model
Objective To investigate the current registration rate of tuberculosis in each county and district of Linyi City,Shandong Province,China and its spatial local relationship with related influencing factors.Methods Related data were collected for the incidence rate of tuberculosis and related influencing factors in 12 counties and districts of Linyi City from January to Decem-ber 2019.Geoda 4.02 software was used to analyze spatial autocorrelation;GWR 4.09 software was used to construct an ordinary least squares regression(OLS)model and a geographic weighted regression(GWR)model;ArcGIS 10.2 software was used to plot the spatial distribution map of influencing factors based on the GWR model.Results In 2019,there were 3 261 cases of tubercu-losis in Linyi City,with an overall incidence rate of 32.39/100 000.There was a significant difference in the incidence rate of tuber-culosis between different counties,sexes,and age groups(x2=71.257-1532.464,P<0.001).The results of spatial autocorrela-tion analysis showed a positive spatial correlation(Moran I=0.234,P=0.021).The results of LISA clustering analysis showed low-low clustering in Luozhuang District and Lanling County and high-high clustering in Junan County.OLS and GWR model ana-lyses showed that the GWR model had better diagnostic criteria than the OLS model.The GWR model showed that population den-sity(β=-0.019),per capita education years(β=-14.509),the number of health technicians per ten thousand people(β=-0.293),and air quality index(β=-15.438)had significant effects on the incidence rate of tuberculosis(P<0.05).Conclusion There are sex,age,and regional differences in the incidence rate of tuberculosis in Linyi City,Shandong Province in 2019.Spatial aggregation is observed for the incidence rate of tuberculosis in each county,and spatial heterogeneity is observed for the influence of social,environmental,economic factors on the incidence rate of tuberculosis.Compared with the OLS model,the GWR model has unique advantages in analyzing the influencing factors for incidence rate.

TuberculosisPrevalenceSpatial regressionEpidemiologyLeast-squares analysisRoot cause analysisGeography,medical

董振、王萍萍、蔺跃付、姜秀波

展开 >

青岛大学公共卫生学院,山东青岛 266071

临沂市人民医院预防科

山东省第一康复医院心内科

结核 患病率 空间回归 流行病学 最小二乘法 影响因素分析 地理学,医学

青岛市科技局项目

18-6-1-79-nsh

2024

精准医学杂志
青岛大学

精准医学杂志

ISSN:2096-529X
年,卷(期):2024.39(2)
  • 28