Nzabakiriraho Jean Damascene, Gayawan Ezra
African Institute for Mathematical Sciences, Kigali, Rwanda.
Department of Statistics, Federal University of Technology, Akure, Nigeria.
BMC Public Health. 2021 Feb 17;21(1):369. doi: 10.1186/s12889-021-10305-x.
Malaria has continued to be a life-threatening disease among under-five children in sub-Saharan Africa. Recent data indicate rising cases in Rwanda after some years of decline. We aimed at estimating the spatial variations in malaria prevalence at a continuous spatial scale and to quantify locations where the prevalence exceeds the thresholds of 5% and 10% across the country. We also consider the effects of some socioeconomic and climate variables.
Using data from the 2014-2015 Rwanda Demographic and Health Survey, a geostatistical modeling technique based on stochastic partial differential equation approach was used to analyze the geospatial prevalence of malaria among under-five children in Rwanda. Bayesian inference was based on integrated nested Laplace approximation.
The results demonstrate the uneven spatial variation of malaria prevalence with some districts including Kayonza and Kirehe from Eastern province; Huye and Nyanza from Southern province; and Nyamasheke and Rusizi from Western province having higher chances of recording prevalence exceeding 5%. Malaria prevalence was found to increase with rising temperature but decreases with increasing volume for rainfall. The findings also revealed a significant association between malaria and demographic factors including place of residence, mother's educational level, and child's age and sex.
Potential intervention programs that focus on individuals living in rural areas, lowest wealth quintile, and the locations with high risks should be reinforced. Variations in climatic factors particularly temperature and rainfall should be taken into account when formulating malaria intervention programs in Rwanda.
疟疾在撒哈拉以南非洲地区仍是五岁以下儿童的一种危及生命的疾病。最近的数据表明，经过数年下降后，卢旺达的疟疾病例有所上升。我们旨在估计疟疾患病率在连续空间尺度上的空间变化，并量化全国范围内患病率超过5%和10%阈值的地点。我们还考虑了一些社会经济和气候变量的影响。
利用2014 - 2015年卢旺达人口与健康调查的数据，采用基于随机偏微分方程方法的地质统计学建模技术，分析卢旺达五岁以下儿童疟疾的地理空间患病率。贝叶斯推断基于集成嵌套拉普拉斯近似。
结果表明疟疾患病率存在空间不均匀变化，一些地区包括东部省的卡扬扎和基雷赫；南部省的胡耶恩扎；以及西部省的尼亚马舍克和鲁济济，记录到患病率超过5%的可能性更高。发现疟疾患病率随温度升高而增加，但随降雨量增加而降低。研究结果还揭示了疟疾与包括居住地点、母亲教育水平以及儿童年龄和性别在内的人口因素之间存在显著关联。
应加强针对农村地区居民、最低财富五分位数人群以及高风险地区人群的潜在干预项目。在卢旺达制定疟疾干预项目时，应考虑气候因素特别是温度和降雨的变化。
