Cutaneous leishmaniasis in Casablanca-Settat region (Morocco): spatio-temporal analysis of disease dynamic and machine learning based case prediction

Cutaneous leishmaniasis (CL) caused by Leishmania protozoa and transmitted through infected sandfly bites, poses a significant public health burden in Morocco. Our research aims to retrospectively assess spatio-temporal patterns of CL in the most densely populated region of the country, Casablanca-Settat, during a period of 14 years (2009–2022). We investigate epidemiological trends, seasonal fluctuations and climatic factors influence on CL prevalence and predict cases in the region’s most important CL focus. Using Geographic Information Systems (GIS) and statistical methods, we analyzed CL data obtained from the Moroccan Ministry of Health and Social Protection with demographic information and climate-related parameters. Four machine learning prediction models for CL cases in the El Brouj focus were compared and their performance was evaluated. Our results revealed a steady increase in CL cases until 2019, followed by a significant surge, with Leishmania tropica identified as the causative agent for all documented cases. The incidence rates varied among demographic groups, with higher rates observed in females and children aged 5-14. Seasonal analysis indicated a notable increase in cases during spring. Spatial distribution analysis identified El Brouj as the sole endemic CL focus_and the highest hotspot in the area, underscoring the necessity of targeted interventions in this location. Furthermore, Spearman correlation analysis revealed a significant association between CL cases and low temperature and precipitation, emphasizing climate variables role in disease incidence. Prediction modeling for future CL cases in El Brouj focus was assessed with six different models and Linear Regression was the most suitable. This retrospective study provides crucial insights into CL epidemiology in the Casablanca-Settat region, highlighting the need for implementing a coordinated approach tailored to the region that could help curb the spread of CL in Settat province, given the predictive model’s outlook on future CL trends in this area.