Predicting geographic variation in cutaneous leishmaniasis, Colombia

Area of Science:

• Environmental epidemiology
• Geographic information systems (GIS) in public health
• Parasitic disease surveillance

Background:

• Colombia reports over 6,000 annual cutaneous leishmaniasis (CL) cases, a significant increase since the 1980s.
• Reported CL incidence likely underestimates true prevalence and is geographically biased by health service access.
• Understanding CL distribution is crucial for effective public health interventions.

Purpose of the Study:

• To assess the predictive power of freely available environmental data for CL case distribution in Colombian municipalities.
• To identify areas with potential underreporting of CL cases.
• To inform targeted disease control strategies.

Main Methods:

• Utilized logistic regression models incorporating elevation and land cover data derived from satellite imagery.
• Developed unique predictive models for each of the 1,079 municipalities.
• Compared the predictive accuracy of elevation, land cover, and combined datasets, and analyzed model performance across different ecological zones.

Main Results:

• Land cover data demonstrated higher predictive power for CL case distribution than elevation.
• Combining elevation and land cover data significantly improved prediction accuracy.
• Fitting separate models for distinct ecological zones enhanced predictive performance, reflecting diverse transmission cycles.

Conclusions:

• Environmental data, particularly land cover, effectively predict the geographic distribution of reported CL cases in Colombia.
• The study identifies municipalities at higher risk of underreporting CL, guiding resource allocation for control.
• This approach offers a scalable method for disease surveillance and public health planning in resource-limited settings.