分析卒中发病率和死亡率空间模式的地理空间方法：一项系统综述。
完整摘要 · 中文译文
卒中是全球死亡和残疾的主要原因之一；然而，其负担在地理区域之间分布不均。本系统综述旨在综合考察卒中发病率、患病率和死亡率空间分布的研究，并回顾用于识别这些模式的地理空间分析方法。通过系统文献检索，共识别出29项应用空间分析或空间建模技术研究卒中相关结局的研究。纳入的研究在地理尺度、空间单元、结局指标和协变量选择方面表现出异质性。采用改良的Joanna Briggs Institute（JBI）批判性评价工具评估方法学质量。总体而言，卒中发病率和死亡率一致被发现在地理空间上呈非随机分布，在多个空间尺度上观察到显著的空间自相关和聚集现象。这些空间模式与区域层面因素密切相关，包括人口老龄化、社会经济状况和城乡特征。环境暴露以及医疗资源空间分布和医疗服务可及性方面的差异也被报道与卒中结局的地理差异有关。常用的地理空间方法包括基于GIS的制图、空间自相关分析、空间回归模型和贝叶斯疾病制图。尽管大多数研究的方法学质量为中等至高等，但在空间单元选择的合理性论证和控制 协变量。总之，卒中负担呈现出显著的空间异质性，这种异质性受结构性、环境及医疗保健可及性因素影响。未来的地理空间研究应采用循证的空间单元和综合分析方法来支持因地制宜的卒中预防策略，并为卫生政策制定提供依据。
完整摘要 · 英文原文
Stroke is a leading cause of mortality and disability worldwide; however, its burden is unevenly distributed across geographical areas. This systematic review aimed to synthesise studies examining the spatial distribution of stroke incidence, prevalence and mortality, with a review of geospatial analytical approaches used to identify these patterns. A systematic literature search identified 29 studies that applied spatial analysis or spatial modelling techniques to stroke-related outcomes. The included studies demonstrated heterogeneity in geographical scale, spatial units, outcome measures and covariate selection. Methodological quality was assessed using a modified Joanna Briggs Institute (JBI) critical appraisal tool. Overall, stroke incidence and mortality were consistently found to be non-randomly distributed in space, with significant spatial autocorrelation and clustering observed across multiple spatial scales. These spatial patterns were closely associated with area-level factors, including population ageing, socioeconomic conditions and urban-rural characteristics. Environmental exposures and disparities in the spatial distribution of healthcare resources and access to care were also reported to contribute to geographical differences in stroke outcomes. Geospatial methods commonly employed included GIS-based mapping, spatial autocorrelation analyses, spatial regression models and Bayesian disease mapping. Although most studies demonstrated moderate to high methodological quality, variation remained in the justification of spatial unit selection and the control of covariates. In conclusion, the burden of stroke exhibits pronounced spatial heterogeneity shaped by structural, environmental and healthcare accessibility factors. Future geospatial research should adopt evidence-based spatial units and integrated analytical approaches to support place-based stroke prevention strategies and inform health policy development.