Spatial Incidence of Noma in Northern Nigeria, 1999-2004: A model-based study

BackgroundNoma is a neglected tropical disease that affects the mouth and face, with a case fatality rate of 80-90%. Active disease is common among two to six-year-olds in Sub-Saharan Africa. The geographical distribution and risk of the disease are ill-defined, and incidence estimates for small areas are unavailable. This study aims to estimate the spatial incidence of noma in 296 small areas across 12 Northern Nigerian states using hierarchical Bayesian models to identify areas with high incidence risks. MethodsThis study used data on patients residing in Northern Nigeria managed at a major noma specialist center in Sokoto between 1999 and 2024 to estimate and map the small-area incidence risk of the disease in the region. Incidence was calculated dynamically at the local government area (LGA) level using the WHO Oral Health Units method, while smoothed standardized incidence ratios (SIR) were modeled using a hierarchical Bayesian Poisson approach that accounts for spatial and non-spatial effects in risk estimation. Relative risks of noma incidence were used to draw choropleth maps followed by spatial autocorrelation analysis to determine clustering or variations in noma incidence among LGAs. ResultsAll states had at least one LGA with a significantly higher noma incidence risk than the regional average. Notably, 20 LGAs in Sokoto (87%), 12 LGAs in Zamfara (85.7%), and 13 LGAs in Kebbi (61.9%) had significantly higher median SIRs than the regional average. Based on the relative risk ranks, Illela (SIR: 17.33 [16.64-18.04]), Wamako (SIR: 13.92 [13.36-14.48]), Goronyo (SIR: 11.66 [11.15-12.17]), and Tangaza (SIR: 11.09 [10.47-11.75]), LGAs in Sokoto state and Bade LGA (SIR: 12.05 [11.47-12.64]) in Yobe state had higher incidence risks than other LGAs in the study period. Autocorrelation analysis also confirmed the clustering of high noma incidence in northwest Nigeria involving 24 LGAs in Kebbi, Sokoto, and Zamfara states. Differences in noma incidence risks were also observed according to different sexes, age groups, and time intervals in the LGAs. ConclusionsSpatial modeling was successful in highlighting small areas with high noma incidence risks in Northern Nigeria, which included at least one LGA in 12 northern Nigerian states. Maps of overall/stratified incidence risk estimates and trends should be considered to guide targeted individual and community actions that promote noma prevention/early detection to mitigate disease burden in Northern Nigeria.