Spatial distribution and socioeconomic context of tuberculosis in Rio de Janeiro, Brazil

OBJECTIVE To analyze the spatial distribution of risk for tuberculosis and its socioeconomic determinants in the city of Rio de Janeiro, Brazil.METHODS An ecological study on the association between the mean incidence rate of tuberculosis from 2004 to 2006 and socioeconomic indicators of the Censo Demográfico (Demographic Census) of 2000. The unit of analysis was the home district registered in the Sistema de Informação de Agravos de Notificação (Notifiable Diseases Information System) of Rio de Janeiro, Southeastern Brazil. The rates were standardized by sex and age group, and smoothed by the empirical Bayes method. Spatial autocorrelation was evaluated by Moran’s I. Multiple linear regression models were studied and the appropriateness of incorporating the spatial component in modeling was evaluated.RESULTS We observed a higher risk of the disease in some neighborhoods of the port and north regions, as well as a high incidence in the slums of Rocinha and Vidigal, in the south region, and Cidade de Deus, in the west. The final model identified a positive association for the variables: percentage of permanent private households in which the head of the house earns three to five minimum wages; percentage of individual residents in the neighborhood; and percentage of people living in homes with more than two people per bedroom.CONCLUSIONS The spatial analysis identified areas of risk of tuberculosis incidence in the neighborhoods of the city of Rio de Janeiro and also found spatial dependence for the incidence of tuberculosis and some socioeconomic variables. However, the inclusion of the space component in the final model was not required during the modeling process.

研究目的：分析巴西里约热内卢市结核病发病风险的空间分布及其社会经济影响因素。
研究方法：本研究为一项生态学研究，探讨2004年至2006年结核病平均发病率与2000年人口普查（Censo Demográfico, Demographic Census）的社会经济指标之间的关联。分析单位为巴西东南部里约热内卢市法定传染病信息系统（Sistema de Informação de Agravos de Notificação, Notifiable Diseases Information System）登记的居民区。发病率按性别和年龄组进行标准化，并通过经验贝叶斯（empirical Bayes）方法进行平滑处理。采用莫兰I指数（Moran’s I）评估空间自相关性。构建多元线性回归模型，并评估在建模过程中纳入空间成分的适宜性。
研究结果：本研究观察到，港口区与北部区域的部分居民区结核病发病风险较高；南部区域的罗西尼亚（Rocinha）、维迪加尔（Vidigal）贫民窟以及西部的上帝之城（Cidade de Deus）同样存在较高发病率。最终模型识别出以下变量与结核病发病风险呈正相关：户主收入为3至5倍最低工资的常住私有住房占比、居民区常住人口占比，以及卧室居住人数超过2人的居民占比。
结论：空间分析明确了里约热内卢市部分居民区存在结核病发病风险区域，同时发现结核病发病率与部分社会经济变量存在空间依赖性。但在本次建模过程中，无需在最终模型中纳入空间成分。