Prediction of hookworm prevalence in southern India using environmental parameters derived from Landsat 8 remotely sensed data

Abstract

Soil-transmitted helminth infections propagate poverty and slow economic growth in low-income countries. As with many other neglected tropical diseases, environmental conditions are important determinants of soil-transmitted helminth transmission. Hence, remotely sensed data are commonly utilised in spatial risk models intended to inform control strategies. In the present study, we build upon the existing modelling approaches by utilising fine spatial resolution Landsat 8 remotely sensed data in combination with topographic variables to predict hookworm prevalence in a hilly tribal area in southern India. Hookworm prevalence data collected from two field surveys were used in a random forest model to investigate the predictive capacity of 15 environmental variables derived from two remotely sensed images acquired during dry and rainy seasons. A variable buffer radius (100-1000 m) was applied to the point-prevalence locations in order to integrate environmental conditions around the village centroids into the modelling approach and understand where transmission is more likely. Elevation and slope were the most important variables in the models, with lower elevation and higher slope correlating with higher transmission risk. A modified normalised difference water index was among other recurring important variables, likely responsible for some seasonal differences in model performance. The 300 m buffer distance produced the best model performance in this setting, with another spike at 700 m, and a marked drop-off in R 2 values at 1000 m. In addition to assessing a large number of environmental correlates with hookworm transmission, the study contributes to the development of standardised methods of spatial linkage of continuous environmental data with point-based disease prevalence measures for the purpose of spatially explicit risk profiling.

Epidemiology of soil transmitted helminths and risk analysis of hookworm infections in the community: Results from the DeWorm3 Trial in southern India

“…In this study, both increased vegetation (NDVI) and elevation were associated with increased odds of infection as well as an increase in intensity of infection. In another study using remotely sensed data at a fine resolution in Jawadhu Hills, topographical parameters of elevation and slope were negatively and positively associated with hookworm infections at the village level [ 58 ]. Riess et al .…”

“…In this study, both increased vegetation (NDVI) and elevation were associated with increased odds of infection as well as an increase in intensity of infection. In another study using remotely sensed data at a fine resolution in Jawadhu Hills, topographical parameters of elevation and slope were negatively and positively associated with hookworm infections at the village level [58]. Riess et al have shown that ecological variables are associated with hookworm infection but have differing effects within a geographical region, are scale-dependent and urge caution against prediction at smaller scales using large-scale data [52].…”

How do disease control measures impact spatial predictions of schistosomiasis and hookworm? The example of predicting school-based prevalence before and after preventive chemotherapy in Ghana

“…The random forest approach was chosen because it can deal with continuous outcome data, multicollinear predictor variables, and low numbers of training samples, and hence, it is the recommended machine learning method for generating predictions [ 26 ]. It has been successfully applied in similar studies [ 10 , 11 , 23 ].…”

“…Several improvements have recently been suggested to common modeling methods applied to smaller sub-national spatial extents. These include utilizing fine resolution RS data (e.g., Landsat 8), employing a larger number of relevant environmental indicators derived from the spectral bands (e.g., modified normalized difference water index [MNDWI]), and using a variable distance radius to extract and aggregate environmental indicator variables around point-prevalence locations [ 10 , 11 ].…”

Canine parasitoses in north Patagonia (Argentina): comparison between different social and environmental factors

“…Although up to now the knowledge about the factors that influence canine parasite infections in cities has been scarce (Tull et al 2019;Otero et al 2018), it is known that the presence and spread of parasites in urban populations is related mainly to socioeconomic conditions (Amundson Romich 2008; Kulinkina et al 2020). The presence of parasites is associated with poverty, inefficient health systems, illiteracy, overcrowding (Pinto et al 2016), poor hygiene, poor housing, limited access to safe water and inadequate rubbish disposal (Álvarez Di Fino et al 2020).…”