Analysing Spatio-Temporal Clustering of Meningococcal Meningitis Outbreaks in Niger Reveals Opportunities for Improved Disease Control

Abstract

Background

Meningococcal meningitis is a major health problem in the “African Meningitis Belt” where recurrent epidemics occur during the hot, dry season. In Niger, a central country belonging to the Meningitis Belt, reported meningitis cases varied between 1,000 and 13,000 from 2003 to 2009, with a case-fatality rate of 5–15%.

Methodology/Principal Findings

In order to gain insight in the epidemiology of meningococcal meningitis in Niger and to improve control strategies, the emergence of the epidemics and their diffusion patterns at a fine spatial scale have been investigated. A statistical analysis of the spatio-temporal distribution of confirmed meningococcal meningitis cases was performed between 2002 and 2009, based on health centre catchment areas (HCCAs) as spatial units. Anselin's local Moran's I test for spatial autocorrelation and Kulldorff's spatial scan statistic were used to identify spatial and spatio-temporal clusters of cases. Spatial clusters were detected every year and most frequently occurred within nine southern districts. Clusters most often encompassed few HCCAs within a district, without expanding to the entire district. Besides, strong intra-district heterogeneity and inter-annual variability in the spatio-temporal epidemic patterns were observed. To further investigate the benefit of using a finer spatial scale for surveillance and disease control, we compared timeliness of epidemic detection at the HCCA level versus district level and showed that a decision based on threshold estimated at the HCCA level may lead to earlier detection of outbreaks.

Conclusions/Significance

Our findings provide an evidence-based approach to improve control of meningitis in sub-Saharan Africa. First, they can assist public health authorities in Niger to better adjust allocation of resources (antibiotics, rapid diagnostic tests and medical staff). Then, this spatio-temporal analysis showed that surveillance at a finer spatial scale (HCCA) would be more efficient for public health response: outbreaks would be detected earlier and reactive vaccination would be better targeted.

Author Summary

Meningococcal meningitis (MM) is an infection of the meninges caused by a bacterium, Neisseria meningitidis, transmitted through respiratory and throat secretions. It can cause brain damage and results in death in 5–15% of cases. Large epidemics of MM occur almost every year in sub-Saharan Africa during the hot, dry season. Understanding how epidemics emerge and spread in time and space would help public health authorities to develop more efficient strategies for the prevention and the control of meningitis. We studied the spatio-temporal distribution of MM cases in Niger from 2002 to 2009 at the scale of the health centre catchment areas (HCCAs). We found that spatial clusters of cases most frequently occurred within nine districts out of 42, which can assist public health authorities to better adjust allocation of resources such as antibiotics or rapid diagnostic tests. We also showed that the epidemics break out in different HCCAs from year to year and did not follow a systematic geographical direction. Finally, this analysis showed that surveillance at a finer spatial scale (health centre catchment area rather than district) would be more efficient for public health response: outbreaks would be detected earlier and reactive vaccination would be better targeted.