Department of Vector Biology and Control, Rajendra Memorial Research Institute of Medical Sciences (ICMR), Agamkuan, Patna 800007, Bihar, India.
BMC Infect Dis. 2013 Feb 2;13:64. doi: 10.1186/1471-2334-13-64.
An improved understanding in transmission variation of kala-azar is fundamental to conduct surveillance and implementing disease prevention strategies. This study investigated the spatio-temporal patterns and hotspot detection for reporting kala-azar cases in Vaishali district based on spatial statistical analysis.
Epidemiological data from the study area during 2007-2011 was used to examine the dynamic space-time pattern of kala-azar outbreaks, and all cases were geocoded at a village level. Spatial smoothing was applied to reduce random noise in the data. Inverse distance weighting (IDW) is used to interpolate and predict the pattern of VL cases distribution across the district. Moran's I Index (Moran's I) statistics was used to evaluate autocorrelation in kala-azar spatial distribution and test how villages were clustered or dispersed in space. Getis-Ord Gi*(d) was used to identify the hotspot and cold spot areas within the study site.
Mapping kala-azar cases or incidences reflects the spatial heterogeneity in the incidence rate of kala-azar affected villages in Vaishali district. Kala-azar incidence rate map showed most of the highest endemic villages were located in southern, eastern and northwestern part of the district; in the middle part of the district generally show the medium occurrence of VL. There was a significant positive spatial autocorrelation of kala-azar incidences for five consecutive years, with Moran's I statistic ranging from 0.04-0.17 (P <0.01). The results revealed spatially clustered patterns with significant differences by village. The hotspots showed the spatial trend of kala-azar diffusion (P < 0.01).
The results pointed to the usefulness of spatial statistical approach to improve our understanding the spatio-temporal dynamics and control of kala-azar. The study also showed the north-western and southern part of Vaishali district is most likely endemic cluster region. To employ exact and geographically suitable risk-reduction programmes, apply of such spatial analysis tools should suit a vital constituent in epidemiology research and risk evaluation of kala-azar.
深入了解黑热病的传播变化对于开展监测和实施疾病预防策略至关重要。本研究通过空间统计分析，调查了瓦伊沙利地区报告黑热病病例的时空模式和热点检测。
利用研究区域 2007-2011 年的流行病学数据，研究了黑热病疫情的动态时空模式，所有病例均按村庄进行了地理编码。应用空间平滑法减少数据中的随机噪声。采用反距离权重法（IDW）对 VL 病例分布模式进行插值和预测。应用 Moran's I 指数（Moran's I）统计量评估黑热病空间分布的自相关性，并检验村庄在空间上的聚集或离散程度。应用 Getis-Ord Gi*（d）识别研究区域内的热点和冷点区域。
黑热病病例或发病率的映射反映了瓦伊沙利地区受影响村庄黑热病发病率的空间异质性。黑热病发病率图显示，大部分高流行村庄位于该地区的南部、东部和西北部；该地区中部通常显示 VL 中度发生。连续五年黑热病发病率均存在显著正空间自相关，Moran's I 统计量范围为 0.04-0.17（P <0.01）。结果显示出具有显著差异的空间聚类模式。热点显示了黑热病扩散的空间趋势（P < 0.01）。
结果表明空间统计方法有助于提高我们对黑热病时空动态和控制的理解。研究还表明，瓦伊沙利地区的西北部和南部很可能是地方性流行区域。为了实施精确和具有地理针对性的减少风险方案，这种空间分析工具的应用应成为黑热病流行病学研究和风险评估的重要组成部分。
