江苏省脑卒中地理分布特征及其与区域危险因素的关系

目的利用地理信息系统(GIS)技术建立江苏省脑卒中空间预测模型。方法收集1998—2002年江苏省所有开展“35岁以上居民慢性非传染性疾病基线调查”地区的数据，以江苏省数字地图为背景，在ArcView软件的支持下，与建立的数据库关联并对各地区的脑卒中标化患病率进行反距离权重插值分析，并进一步利用区域危险因素对脑卒中分布差异进行多因素分析。结果脑卒中患病资料的空间分析预测模型显示，脑卒中和缺血性脑卒中主要分布于南京、徐州和赣榆，呈现从西北向东南逐渐减低的趋势；而出血性脑卒中主要分布于苏南的南京、苏州、无锡、镇江和徐州等城区，以南京地区最为严重，呈现为南高北低的趋势；饮酒率、1998年人均GDP、超重率、肥胖率、糖尿病患病率进入脑卒中线性回归模型，饮酒率和冠心病患病率进入出血性脑卒中空间线性回归模型，吸烟率、1998年人均GDP、超重率、肥胖率、糖尿病患病率进入缺血性脑卒中线性回归模型。结论GIS技术建立的脑卒中空间预测模型对江苏省脑卒中防治重点区域的确定以及进一步预测有一定意义。

Study on the Application of Geographic Information System in Spatial Distribution of Stroke in Jingsu Province

Objective Using the techniques of the Geographic Information System (GIS) to set up the spatial predication model for Stroke in Jiangsu Province. Methods A database was set up based on the information collected and linked to electronic maps of Jiangsu Province with software ArcView 3.3 in 39 counties or districts during 1998-2002. The spatial distribution models of stroke were developed using inverse distance weighted interpolation with the Spatial Analysis. The linear regression model based on ordinary least squares and the spatial linear regression model based on the spatial pattern were performed through S-plus and S+spatial Stats. Results It was found that high risk areas of stroke and ischaemic stroke were located in Nan Jing, Xu Zhou, and Gan Yu, with the continuous decrease from the northwest to southeast of Jiangsu Province. The high risk areas of hemorrhagic stroke were located in the southern parts of Jiangsu Province such as Nan Jing, Su Zhou, Wu Xi, Zhen Jiang etc. with the continuous decrease from the south to north in Jiangsu Province. Some risk factors entered the corresponding multivariate regression models with drinking, GDP, overweight, obesity and diabetes to stroke model, drinking and CHD to hemorrhagic stroke model, and smoke, GDP, overweight, obesity and diabetes to ischaemic stroke model. Conclusion This study highlighted the importance in the predication of some high-risk areas in the spatial model of stroke by using GIS technique.