99 / 2019-10-20 10:17:13

Mapping scrub typhus transmission risk and identifying the associated environmental factors in Nepal using MaxEnt, Random Forest and BRT models

Scrub typhus,risk mapping,,machine learning,Nepal

Being a globally emerging mite-borne zoonotic disease, scrub typhus has been a serious public health concern in Nepal. Mapping the disease transmission risk and quantifying geographical variations in the occurrence of scrub typhus and the surrounding environment is important for prevention and control efforts. In this study, we modeled and mapped scrub typhus transmission risk using the ecological niche approach, machine learning modeling techniques and reported case of scrub typhus along with several climatic, topographic, NDVI and proximity explanatory variables. We modelled the ecological niche of scrub typhus with MaxEnt, Random Forest (RF) and Boosted Regression Tree (BRT) algorithms. The results showed that all three techniques have robust predictive power with test AUC and TSS of above 0.8 and 0.6, respectively. Spatial prediction revealed that environmentally suitable areas of scrub typhus is widely distributed across the country particularly in the low land Tarai and less elevated river valleys. We found areas close to agricultural land with gentle slope have higher risk of scrub typhus occurrence. Despite several speculation on the association between proximity to earthquake epicenter and scrub typhus, we did not find a significant role of proximity to earthquake epicenter in the distribution of scrub typhus transmission risk in Nepal. These findings could be useful to select priority areas for surveillance and control strategies.