Bayesian spatio-temporal modeling of malaria risk in Rwanda

Area of Science:

• Epidemiology
• Biostatistics
• Public Health

Background:

• Malaria causes over 435,000 deaths annually, primarily in Africa and Asia, despite being preventable and curable.
• Rwanda aimed to reduce malaria incidence by 42% between 2012 and 2018 as part of global malaria elimination efforts.
• Traditional incidence rate approaches pose challenges for effective health policy assessment and decision-making.

Purpose of the Study:

• To propose and apply advanced statistical models for malaria control evaluation in Rwanda.
• To incorporate spatial structures and relative risk uncertainty into malaria incidence analysis.
• To inform the National Malaria Control Program with data-driven insights for targeted interventions.

Main Methods:

• Utilized a spatio-temporal modeling approach, including SIR (Susceptible-Infectious-Recovered) and BYM (Besag-York-Mollié) models.
• Employed routine health facility data from Rwanda spanning 2012 to 2018.
• Estimated excess probability and relative risk (RR) to evaluate incidence reduction and target attainment.

Main Results:

• Malaria incidence showed an overall increase from 2013 to 2018 compared to 2012 levels.
• Nearly half (47.36%) of Rwanda's sectors did not achieve the targeted malaria incidence reduction.
• The spatio-temporal model effectively estimated relative risk and assessed the probability of meeting control targets.

Conclusions:

• The proposed excess probability method offers a robust statistical framework for evaluating malaria control program effectiveness.
• This approach aids in identifying areas requiring intensified interventions and monitoring progress towards malaria elimination goals.
• The findings underscore the need for sustained and adaptive malaria control strategies in Rwanda and similar settings.