Table 2.
Summary of key studies on spatial variation in snakebite incidence or mortality, ranging from simple descriptive studies to fine-scale predictions.
| Type | Measure | Area | Resolution | Method | Important Predictors | |
|---|---|---|---|---|---|---|
| Studies describing broad scale spatial patterns and hotspots in snakebite incidence | ||||||
| Swaroop (1954) | Spatial* | Mortality | Global | Source data: | NA | NA |
| Temporal* | Country | |||||
| Predictions: | ||||||
| NA | ||||||
| Chippaux (1998) | Spatial | incidence | Global | Source data: | NA | NA |
| Country | ||||||
| Predictions: | ||||||
| Snakebite Regions | ||||||
| Kasturiratne et al. (2008) | Spatial | Incidence | Global | Source data: | NA | NA |
| Mortality | Countries | |||||
| Predictions: global burden region | ||||||
| Studies using simple statistics, epidemiology, and coarse scale spatial predictors to describe spatial variation in snakebite incidence | ||||||
| Molesworth (2003) | Spatial | Incidence | West Africa (Ghana & Nigeria) | Source data: | LogR | NDVI↑ |
| Temporal | 29 health facilities Predictions: ~15 km grid | Season (Rainy season) | ||||
| Leynaud and Reati (2009) | Spatial | Incidence | Cordoba, Argentina | Source data: | Spatial smoothing model | Location in departments with high percentage of persistence farming |
| Department | Species identity | |||||
| Predictions: | ||||||
| department | ||||||
| Mohapatra et al. (2011) | Spatial | Mortality | India | Source data: | LogR | Male/Female |
| Temporal | ~7000 small areas | Religion (Hindu↑) | ||||
| Individual | Predictions: states | Occupation (Agricultural worker↑) | ||||
| Season (Monsoon↑) | ||||||
| State (high prevalence states↑) | ||||||
| Age (15–29↑) | ||||||
| Chippaux, 2017** | Spatial | Incidence | Americas | Source Data: | t-test, Pearson Correlation, Chi Squared, Mann-Whitney Test | Altitude↓ |
| Temporal | mortality | Province | Male/Female | |||
| Individual | Predictions: | Age (young to middle aged↑) | ||||
| Province | Climate Zone | |||||
| Season (Rainy or Summer↑) | ||||||
| Population density↑↓ | ||||||
| Year↑↓ | ||||||
| Angarita-Gerlein et al. (2017) | Spatial | incidence | Colombia | Source Data: | Cross-correlation analysis | Precipitation |
| Temporal | Municipality | Municipality Identity | ||||
| Predictions: | ||||||
| Municipality | ||||||
| Riascos et al., 2019 | Spatial | Incidence | Coffee Triangle Region, Colombia | Source data: Municipality | NA | Year |
| Temporal | Predictions: | Season | ||||
| NA | ||||||
| León-Núñez et al. (2020) | Spatial | Incidence | Colombia | Source data: | t-test, Pearson Correlation, Chi Squared, Mann-Whitney Test | Male/Female |
| Individual | Department | Urban/Rural | ||||
| Predictions: Department | Ethnicity (Afro-Colombian & Indigenous↑) | |||||
| Age (28–35↑) | ||||||
| Region (Amazonia & Orinoquia↑) | ||||||
| Species identity | ||||||
| Year↑ | ||||||
| Studies using relatively novel fine scale source data, advanced statistical models, and improved resolution | ||||||
| Hansson et al. (2010) | Spatial | Incidence | Nicaragua | Source data: municipality | Poisson regression | Season (Rainy Season↑) |
| Temporal | Predictions: | Environmental Region (altitude, precipitation, geographic clustering; Wet Lowlands↑) | ||||
| Individual | municipality | Rural population percentage↑ | ||||
| Male population percentage↑ | ||||||
| Young population percentage | ||||||
| Underreporting index↑ | ||||||
| Hansson et al. (2013) | Spatial | Incidence | Costa Rica | Source data: district | Bayesian Poisson regression | altitude↓ |
| Predictions: | precipitation↑ | |||||
| district | length of dry season↓ | |||||
| rural population percentage↑ | ||||||
| population percentage near large forests↑ | ||||||
| Snake habitat suitability↑ | ||||||
| Chaves et al. (2015) | Spatial | incidence | Costa Rica | Source data: County | geographically weighted regression | Weather & Climate Oscillations |
| Temporal | Predictions: | Temperature↑ | ||||
| County | Precipitation | |||||
| Poverty Indicators (Poverty gap | ||||||
| index and percentage of destitute housing)↑ | ||||||
| Altitude↓ | ||||||
| Yañez-Arenas et al. (2016) | Spatial | Incidence | Americas | Source data: | GLM | Cumulative MRS presence & abundance index (SRI_2) ↑ |
| Provinces Predictions: ~20 km grid | ||||||
| Yañez-Arenas et al., 2014 | spatial | Incidence | Veracruz, Mexico | Source data: Municipality | GAM | 2 MRS species' abundance estimate↑ |
| Predictions: Municipality | Index of marginalization↑ | |||||
| Suraweera et al. (2020) | Spatial | Mortality | India | Source data: | Spatial Poisson model | Age group (30–69↑) |
| Temporal | Incidence (inferred) | ~7000 small areas Predictions: ~50 km grid | Male/Female | |||
| Individual | Season (Monsoon↑) | |||||
| Elevation to 400m↓ | ||||||
| Urban/Rural | ||||||
| Poverty (rural female illiteracy)↑ | ||||||
| Monthly mean temperature to 20 °C↑ | ||||||
| Year↓ | ||||||
| Species identity | ||||||
| Schneider et al. (2021) | Spatial | Incidence | Brazil | Source data: | Negative binomial regression model | Major habitat type (Tropical↑) |
| Municipality | Temperature↑ | |||||
| Predictions: | Precipitation↑ | |||||
| Municipality | Elevation↑ | |||||
| Urbanization percentage↓ | ||||||
| Venomous snake richness | ||||||
| Forest loss↑ | ||||||
| GDP per capita↓ | ||||||
| Studies resulting in fine scale predictions of snakebite incidence | ||||||
| Ediriweera et al., (2016) | Spatial | Sri Lanka | Source Data: household clusters in smallest administrative divisions | GLM GAM Geostatistical binomial logistic Log-linear models |
Male/Female Age (middle aged↑) Time of day (evening↑) Occupation (farm labourer↑) Education↓ Monthly income↓ Population density↓ Elevation Occupation distribution Climatic zone Season humidity weather abnormalities↓ |
|
| Ediriweera et al. (2018) | Temporal | Predictions: | ||||
| Ediriweera et al., 2019 | Individual | 1 km | ||||
| Bravo-Vega et al. (2019) | Spatial | incidence | Costa Rica | Source Data: District Predictions: 1 km |
Linear regression | Encounter frequency of Bothrops asper Human population density |
| Goldstein et al. (2021) | Spatial temporal |
Incidence | Sri Lanka | Source data: 10m-2km Predictions:2 km study squares |
Bottom-up Agent based modelling | Snake-famer activity overlap patterns based on: Monthly precipitation Number of rainy days Farmer type Land type Daily farmer activity time↑ Population percentage farmers↑ Snake activity season↑ Circadian snake activity time↑ Snake aggressiveness↑ Snake land type association↑ Snake abundance estimate↑ |
GAM = generalized additive models; GLM = generalized linear models; LogR = Logistic regression; SRI = ‘snakebite risk index’; NDVI = normalized difference vegetation index.
↑ = positive correlation; ↓ = negative correlation; no arrow = complex correlation pattern; bold text = significant categorical predictor.
*Information given in written form such as tables but could be analysed spatially and/or temporally.
**small scale studies already summarized in this review are not listed again separately in the table.