The environmental drivers of bacterial meningitis epidemics in the Democratic Republic of Congo, central Africa.
Introduction Bacterial meningitis still constitutes an important threat in Africa. In the meningitis belt, a clear seasonal pattern in the incidence of meningococcal disease during the dry season has been previously correlated with several environmental parameters like dust and sand particles as wel...
| Published in: | PLOS Neglected Tropical Diseases |
|---|---|
| Main Authors: | , , , , |
| Format: | Article in Journal/Newspaper |
| Language: | English |
| Published: |
Public Library of Science (PLoS)
2020
|
| Subjects: | |
| Online Access: | https://doi.org/10.1371/journal.pntd.0008634 https://doaj.org/article/e8bad758769448c8821a48b779fc0e42 |
| Summary: | Introduction Bacterial meningitis still constitutes an important threat in Africa. In the meningitis belt, a clear seasonal pattern in the incidence of meningococcal disease during the dry season has been previously correlated with several environmental parameters like dust and sand particles as well as the Harmattan winds. In parallel, the evidence of seasonality in meningitis dynamics and its environmental variables remain poorly studied outside the meningitis belt. This study explores several environmental factors associated with meningitis cases in the Democratic Republic of Congo (DRC), central Africa, outside the meningitis belt area. Methods Non-parametric Kruskal-Wallis' tests were used to establish the difference between the different health zones, climate and vegetation types in relation to both the number of cases and attack rates for the period 2000-2018. The relationships between the number of meningitis cases for the different health zones and environmental and socio-economical parameters collected were modeled using different generalized linear (GLMs) and generalized linear mixed models (GLMMs), and different error structure in the different models, i.e., Poisson, binomial negative, zero-inflated binomial negative and more elaborated multi-hierarchical zero-inflated binomial negative models, with randomization of certain parameters or factors (health zones, vegetation and climate types). Comparing the different statistical models, the model with the smallest Akaike's information criterion (AIC) were selected as the best ones. 515 different health zones from 26 distinct provinces were considered for the construction of the different GLM and GLMM models. Results Non-parametric bivariate statistics showed that there were more meningitis cases in urban health zones than in rural conditions (χ2 = 6.910, p-value = 0.009), in areas dominated by savannah landscape than in areas with dense forest or forest in mountainous areas (χ2 = 15.185, p-value = 0.001), and with no significant difference between ... |
|---|