基于INLA算法的艾滋病和结核病联合时空分布特征分析

目的 由于艾滋病与结核病在流行病学特征上存在着重叠性，为探讨分析中国2012—2017年31个省、市、自治区的艾滋病与结核病两者之间的联合和特定疾病的时空分布特征，为疾病防控工作提供理论依据。方法 收集并整理2012—2017年中国31个省份的艾滋病与结核病发病数据，建立贝叶斯分层时空分布模型，利用INLA算法探讨两者联合和特定疾病的时空风险模式及其成因。结果 空间分布上，艾滋病与结核病在特定与共享模式下的风险分布具有相似性，其相似的高发地区有四川、重庆、云南、贵州、广西、广东、新疆等地，且成聚集分布；时间分布上，艾滋病时间风险逐年增加，而结核病时间风险基本保持恒定，且艾滋病发病风险已超过结核病；时空分布上具有相似性，高发区域分布在云南、贵州、广西、广东、新疆等地，该共性时空模式的形成与经济、交通、医疗水平等宏观因素密切相关。结论 艾滋病与结核病在空间、时空分布上都存在着相似性，今后要重点防控两种疾病联合高发区，并依据社会宏观因素的差异性制定防控策略。

Characterization of joint spatial and temporal distribution of AIDS and tuberculosis based on integrated nested laplace approximation algorithm

Objective:Due to the overlap in epidemiological characteristics of AIDS and tuberculosis, to explore and analyze the joint and disease-specific spatio -temporal distribution characteristics between AIDS and tuberculosis in 31 provinces, cities and autonomous regions of China from 2012 to 2017, and to provide a theoretical basis for disease prevention and control. Methods The incidence data of AIDS and tuberculosis in China from 2012 to 2017 were collected and collated, and a Bayesian hierarchical spatio -temporal distribution model was developed. Integrated nested laplace approximation (INLA) algorithm was used to estimate the model to explore the spatio-temporal risk patterns and its causes of the joint and specific diseases. Results In terms of spatial distribution, the distribution of incidence under the specific and shared model of AIDS and tuberculosis shared similarity, and their similar high-risk incidence areas were Sichuan, Chongqing, Yunnan, Guizhou, Guangxi, Guangdong, and Xinjiang, and they became aggregated distribution. The temporal distribution showed that the temporal risk of AIDS increased year by year, while the temporal risk of tuberculosis remained essentially a constant trend, and the risk of AIDS incidence exceeded the risk of tuberculosis incidence. The spatio-temporal distribution had similarity, and the high prevalence areas were distributed in Yunnan, Guizhou, Guangxi, Guangdong, Xinjiang and other regions. This common spatio -temporal pattern formation was closely related to macroscopic factors such as economy, transportation and medical level. Conclusion Similarities exist in the spatial and spatio-temporal distribution of both AIDS and tuberculosis, and we should focus on the prevention and control of the joint high prevalence areas of the two diseases in the future, and develop prevention and control strategies based on the differences in social macro factors.