What is it about?
Our study aimed to develop a Bayesian dynamic linear model for spatiotemporal analysis of perinatal and neonatal mortality, using Uganda as a case study. We built this model, estimated its parameters with Kalman filtering, and evaluated it through Monte Carlo methods. The model was diagnostically checked with residual tests for autocorrelation, normality, and homoscedasticity. Convergence was assessed visually through trace and density plots and numerically via the Gelman-Rubin diagnostic. To assess the model's predictive ability, we generated 10-month forecasts of perinatal and neonatal mortality rates.
Featured Image
Photo by Solen Feyissa on Unsplash
Why is it important?
Our greatest achievement is developing a Bayesian dynamic linear model that can monitor a country's health status—specifically neonatal and perinatal mortality—by examining the past, analyzing the present, and reliably forecasting the future, unlike other models that offer non-time-dependent estimates. This model is highly tailored to each country, especially as 2030 approaches, because most countries will want to evaluate their progress toward the Sustainable Development Goals policies they have implemented.
Perspectives
Writing this article was a great pleasure because it involved co-authors who have been my mentors as lecturers and work-based supervisors. As a result, I have been able to learn more from them, especially about academic writing. This article also led me to strengthen my relationships with Ministry of Health, Uganda staff, particularly those in the Department of Reproductive, Maternal, and Child Health, who shared their on-field experience with me.
GEORGE BAMWEBAZE
Kyambogo University
Read the Original
This page is a summary of: Formulation of a spatiotemporal model for the analysis of neonatal mortality amidst SDG interventions: The case of Uganda, PLOS One, March 2026, PLOS,
DOI: 10.1371/journal.pone.0323859.
You can read the full text:
Contributors
The following have contributed to this page
