%0 Generic %A Paireau, Juliette %A B. Maïnassara, Halima %A Jusot, Jean-François %A Collard, Jean-Marc %A Idi, Issa %A Moulia-Pelat, Jean-Paul %A E. Mueller, Judith %A Fontanet, Arnaud %D 2014 %T Spatio-Temporal Factors Associated with Meningococcal Meningitis Annual Incidence at the Health Centre Level in Niger, 2004–2010 %U https://plos.figshare.com/articles/dataset/_Spatio_Temporal_Factors_Associated_with_Meningococcal_Meningitis_Annual_Incidence_at_the_Health_Centre_Level_in_Niger_2004_8211_2010_/1034371 %R 10.1371/journal.pntd.0002899 %2 https://ndownloader.figshare.com/files/1507900 %2 https://ndownloader.figshare.com/files/1507901 %2 https://ndownloader.figshare.com/files/1507903 %2 https://ndownloader.figshare.com/files/1507904 %2 https://ndownloader.figshare.com/files/1507905 %K Computational biology %K Population modeling %K Infectious disease modeling %K geoinformatics %K Geographic Information Systems %K Spatial analysis %K Spatial autocorrelation %K Atmospheric science %K Climatology %K geography %K Cartography %K Human geography %K epidemiology %K Spatial epidemiology %K Infectious diseases %K Bacterial diseases %K Meningococcal disease %K Infectious diseases of the nervous system %K meningitis %K Public and occupational health %K mathematics %K Statistics (mathematics) %K meningococcal %K incidence %K centre %X

Background

Epidemics of meningococcal meningitis (MM) recurrently strike the African Meningitis Belt. This study aimed at investigating factors, still poorly understood, that influence annual incidence of MM serogroup A, the main etiologic agent over 2004–2010, at a fine spatial scale in Niger.

Methodology/Principal Findings

To take into account data dependencies over space and time and control for unobserved confounding factors, we developed an explanatory Bayesian hierarchical model over 2004–2010 at the health centre catchment area (HCCA) level. The multivariate model revealed that both climatic and non-climatic factors were important for explaining spatio-temporal variations in incidence: mean relative humidity during November–June over the study region (posterior mean Incidence Rate Ratio (IRR) = 0.656, 95% Credible Interval (CI) 0.405–0.949) and occurrence of early rains in March in a HCCA (IRR = 0.353, 95% CI 0.239–0.502) were protective factors; a higher risk was associated with the percentage of neighbouring HCCAs having at least one MM A case during the same year (IRR = 2.365, 95% CI 2.078–2.695), the presence of a road crossing the HCCA (IRR = 1.743, 95% CI 1.173–2.474) and the occurrence of cases before 31 December in a HCCA (IRR = 6.801, 95% CI 4.004–10.910). At the study region level, higher annual incidence correlated with greater geographic spread and, to a lesser extent, with higher intensity of localized outbreaks.

Conclusions

Based on these findings, we hypothesize that spatio-temporal variability of MM A incidence between years and HCCAs result from variations in the intensity or duration of the dry season climatic effects on disease risk, and is further impacted by factors of spatial contacts, representing facilitated pathogen transmission. Additional unexplained factors may contribute to the observed incidence patterns and should be further investigated.

%I PLOS Neglected Tropical Diseases