Journal Article Epublish
PubMed: 40575492 DOI: 10.1016/j.ijregi.2025.100673
Estimating childhood tuberculosis incidence and under-reporting in Gedeo Zone, Ethiopia: a Bayesian hidden Markov model.
IJID Reg 15:100673
Authors: Solomon Hailemariam Tesfaye, Tsion Mulat Tebeje, Daneil Sisay, Mektew Belete, Yohannes Kifle, Asresu Feleke, Abinet Meno
Affiliations
- School of Public Health, College of Health Sciences and Medicine, Dilla University, Dilla, Ethiopia.
- Gedeo Zone Health Department, Dilla, Ethiopia.
- Department of Nursing, College of Health Sciences and Medicine, Dilla University, Dilla, Ethiopia.
- School of Medicine, College of Health Sciences and Medicine, Dilla University, Dilla, Ethiopia.
Abstract
Objectives
To estimate the incidence and under-reported cases of childhood tuberculosis (TB) in rural Ethiopia.
Methods
We used a Bayesian hidden Markov model with Spatio-temporal random effects to analyze TB notification data from rural areas between 2018 and 2023. Spatial dependence and HIV infection were included as covariates for estimating TB incidence, while the availability of diagnostic services informed the case detection process. Sensitivity analysis were conducted to assess the robustness of the results with alternative prior distributions.
Results
Childhood TB incidence in the Gedeo Zone increased from 141 cases per 100,000 population (95% uncertainty interval: 96-193) in 2018 to 157 cases (95% uncertainty interval: 114-207) in 2023. Estimated case detection rates ranged from 56 cases per 100,000 in 2018 to 62 in 2023, indicating substantial under-reporting. Spatial lags of TB notifications predicted incidence in adjacent areas. Sensitivity analysis confirmed result robustness.
Conclusions
The gap between estimated TB incidence and reported cases highlights the urgent need to strengthen TB surveillance in the study area.
MeSH
Incidence, Uncertainty, Tuberculosis