Modeling Optimal Laboratory Testing Strategies for Bacterial Meningitis Surveillance in Africa

Abstract

Since 2010, the introduction of an effective serogroup A meningococcal conjugate vaccine has led to the near-elimination of invasive Neisseria meningitidis serogroup A disease in Africa's meningitis belt. However, a significant burden of disease and epidemics due to other bacterial meningitis pathogens remain in the region. High-quality surveillance data with laboratory confirmation is important to monitor circulating bacterial meningitis pathogens and design appropriate interventions, but complete testing of all reported cases is often infeasible. Here, we use case-based surveillance data from 5 countries in the meningitis belt to determine how accurately estimates of the distribution of causative pathogens would represent the true distribution under different laboratory testing strategies. Detailed case-based surveillance data was collected by the MenAfriNet surveillance consortium in up to 3 seasons from participating districts in 5 countries. For each unique country-season pair, we simulated the accuracy of laboratory surveillance by repeatedly drawing subsets of tested cases and calculating the margin of error of the estimated proportion of cases caused by each pathogen (the greatest pathogen-specific absolute error in proportions between the subset and the full set of cases). Across the 12 country-season pairs analyzed, the 95% credible intervals around estimates of the proportion of cases caused by each pathogen had median widths of ±0.13, ±0.07, and ±0.05, respectively, when random samples of 25%, 50%, and 75% of cases were selected for testing. The level of geographic stratification in the sampling process did not meaningfully affect accuracy estimates. These findings can inform testing thresholds for laboratory surveillance programs in the meningitis belt.

Keywords: Bacterial Meningitis; Burkina Faso; Chad; Laboratory Surveillance; Mali; Modeling; Niger; Togo.

Figure 1.

Map of health districts by first season of inclusion in analysis.

Figure 2.

Confirmed bacterial meningitis cases by pathogen. Bar heights represent numbers of confirmed meningitis cases associated with each bacterial pathogen, pooled across all 12 analyzed country-seasons. Abbreviations: Hi non-b, Haemophilus influenzae non–type b; Hib, Haemophilus influenzae type b; NmA, NmC, NmW, NmX, and NmY, Neisseria meningitidis serogroups A, C, W, X, and Y, respectively; NmInd, indeterminate/unknown serogroup of N. meningitidis; Spn, Streptococcus pneumoniae.

Figure 3.

Margin of error of pathogen proportion estimates by testing coverage. When a sample of tested cases is used to estimate the true share of bacterial meningitis cases caused by each pathogen (the pathogen proportion), the margin of error for the sample is defined as the greatest pathogen-specific absolute error, such that the estimates for all pathogens are within the margin of error of the true proportion. Thus, lower margins of error indicate more accurate and precise sets of estimates. In this figure, we show how different levels of geographic stratification (columns) and testing coverage (x-axis) affect the average (A) and 95% credible (B) margin of error under random sampling, and the observed margin of error (C) under sequential sampling. These margin of error values (y-axis) are calculated individually for each of the 12 country-seasons, and summarized in terms of the median (black lines) and interquartile range (IQR; shaded regions); distinct curves for each country-season can be found in Supplementary Figure 1). Green, orange, and purple text denote specific median (IQR) margin of error values when 25%, 50%, and 75% of cases are tested, respectively.

Figure 4.

Margin of error of test-positive proportion estimates by testing coverage. We define the test-positive proportion as the proportion of tested suspect meningitis cases which were confirmed as bacterial meningitis. Figure shows how different levels of geographic stratification (columns) and testing coverage (x-axis) affect the mean (A) and 95th percentile (B) absolute error under random sampling, and the observed absolute error (C) under sequential sampling, of the test-positive proportion. These absolute error values (y-axis) are calculated individually for each of the 12 country-seasons, and summarized in terms of the median (black lines) and interquartile range (IQR; shaded regions). Green, orange, and purple text denote specific median (IQR) absolute error values when 25%, 50%, and 75% of cases are tested, respectively.

Figure 5.

Probability of detecting a rare Neisseria meningitidis serogroup A (NmA) case by testing coverage, Burkina Faso, 2014–2015 season.