Quantifying undetected tuberculosis in Ethiopia using a novel geospatial modelling approach.

Area of Science:

• Epidemiology
• Geospatial analysis
• Public health

Background:

• Tuberculosis (TB) remains a leading global infectious cause of death.
• Approximately three million TB cases go undetected, fueling community transmission.
• Understanding spatial distribution of undetected TB is crucial for targeted interventions in high-burden settings.

Purpose of the Study:

• To estimate the number of undetected TB cases in Ethiopia at national and local levels.
• To identify spatial clustering patterns of undetected TB cases.
• To determine factors associated with the clustering of undetected TB.

Main Methods:

• Employed a Bayesian geostatistical modeling framework.
• Integrated national TB prevalence and notification data with environmental variables.
• Utilized Moran's Index and Local Indicator of Spatial Autocorrelation (LISA) for spatial analysis.

Main Results:

• Estimated 51,041 undetected TB cases in Ethiopia (95% CI: 50,599-51,486).
• Major undetected TB burdens predicted in Oromia (20,440), Amhara (9614), and South Ethiopia (6061) regions.
• Spatial clustering observed in border districts and Southern Ethiopia; negatively associated with mass media exposure and wealth.

Conclusions:

• Ethiopia faces a substantial burden of undetected TB, with spatial clustering in border areas and regions with limited healthcare access.
• Targeted TB screening for low socioeconomic communities is recommended.
• Improving mass media exposure in high-burden areas can significantly reduce undetected TB.