full-text ►
Evaluation of Meningitis Surveillance Before Introduction of Serogroup A Meningococcal Conjugate Vaccine — Burkina Faso and Mali

	MMWR Weekly Volume 61, No. 50

Evaluation of Meningitis Surveillance Before Introduction of Serogroup A Meningococcal Conjugate Vaccine — Burkina Faso and Mali

Weekly

December 21, 2012 / 61(50);1025-1028

Each year, 450 million persons in a region of sub-Saharan Africa known as the "meningitis belt" are at risk for death and disability from epidemic meningitis caused by serogroup A Neisseria meningitidis (1). In 2009, the first serogroup A meningococcal conjugate vaccine (PsA-TT) developed solely for Africa (MenAfriVac, Serum Institute of India, Ltd.), was licensed for persons aged 1–29 years. During 2010–2011, the vaccine was introduced in the hyperendemic countries of Burkina Faso, Mali, and Niger through mass campaigns. Strong meningitis surveillance is critical for evaluating the impact of PsA-TT because it was licensed based on safety and immunogenicity data without field effectiveness trials. Case-based surveillance, which includes the collection of epidemiologic and laboratory data on individual cases year-round, is recommended for countries that aim to evaluate the vaccine's impact. A key component of case-based surveillance is expansion of laboratory confirmation to include every case of bacterial meningitis because multiple meningococcal serogroups and different pathogens such as Haemophilus influenzae type b and Streptococcus pneumoniae cause meningitis that is clinically indistinguishable from that caused by serogroup A Neisseria meningitidis. Before the introduction of PsA-TT, evaluations of the existing meningitis surveillance in Burkina Faso and Mali were conducted to assess the capacity for case-based surveillance. This report describes the results of those evaluations, which found that surveillance infrastructures were strong but opportunities existed for improving data management, handling of specimens shipped to reference laboratories, and laboratory capacity for confirming cases. These findings underscore the need to evaluate surveillance before vaccine introduction so that activities to strengthen surveillance are tailored to a country's needs and capacities.
Before introduction of the meningococcal conjugate vaccine, meningitis surveillance in Burkina Faso and Mali included aggregate case counts only, enhanced by cerebrospinal fluid (CSF) collection from a subset of cases during the epidemic season to guide epidemic preparedness and choice of polysaccharide vaccine. In collaboration with the West Africa Inter-Country Support Team of the World Health Organization's Africa Regional Office, CDC evaluated 2007 meningitis surveillance data from Burkina Faso during 2007–2008 and from Mali in 2010. Surveillance was evaluated according to CDC guidelines (2). Each country's surveillance system was evaluated for compliance with standard operating procedures for enhanced meningitis surveillance and case-based surveillance in Africa developed by the World Health Organization (3–5). Meningitis surveillance data were analyzed, stakeholders were consulted, and surveillance databases, reports, and registers were examined. Data management was evaluated, along with data completeness, reporting completeness, and representativeness; specimen collection and transport; and laboratory confirmation.
Burkina Faso
In Burkina Faso in 2007, all 55 districts reported a total of 25,695 meningitis cases to the national surveillance office. Cases were reported weekly in aggregate, and reporting was supplemented with line lists of case-level data during the epidemic season. Multiple databases rather than a single database were used, and unique identifiers were not used to link epidemiologic and laboratory data; instead, hand-matching (i.e., by name, age, and residence) was attempted.
Completeness of case-level data was greater for demographic information (98%) than for vaccination status (81%). Reporting completeness of the surveillance system, defined as the 10,614 line-listed cases divided by the 25,695 total cases reported in aggregate, was 41%. Of the line-listed cases, 9,824 (93%) had CSF specimens collected. Population representativeness of surveillance data based on the proportion of districts submitting line lists and CSF specimens was 91% (50/55) and 85% (47/55), respectively; 4% (443/10,614) of line-listed cases and 4% (423/9,824) of specimens were from the Burkina Faso capital, Ouagadougou.
The proportion of all reported cases with a specimen reaching a national reference laboratory was 11% (2,898/25,695) for cases reported in aggregate and 27% (2,898/10,614) for line-listed cases. CSF macroscopic examination, Gram stain, and white blood cell count were performed routinely at district laboratories; results of these tests were suggestive of bacterial meningitis* in 35% (3,428/9,824) of specimens. Five reference laboratories in Burkina Faso performed culture or latex agglutination, and one of these performed conventional polymerase chain reaction (PCR) for pathogen confirmation. The proportion of specimens reaching a national reference laboratory that were confirmed as bacterial meningitis† was 24% (685/2,898).
Mali
In Mali in 2007, all 59 districts reported a total of 978 meningitis cases to the national surveillance office. Cases were reported weekly in aggregate, but reporting was not supplemented with line-listed cases during the epidemic season. Multiple databases rather than a single database were used, and unique identifiers were not used to link epidemiologic and laboratory data. Case-level data were recorded for the 514 specimens that reached the national reference laboratory, but these data were not systematically entered into any database. Completeness of these case-level data was greater for demographic information and confirmatory laboratory results than for vaccination status and outcome (95% and 100% versus 11% and 30%).
In Mali, the total number of specimens collected was unknown and line lists were not available; therefore, measures of reporting completeness could not be evaluated. Population representativeness of surveillance data based on proportion of districts submitting CSF specimens was 61% (36/59); 63% (324/514) of specimens received at the reference laboratory were from the Mali capital, Bamako. The proportion of reported cases with a specimen reaching the national reference laboratory was 53% (514/978). The median interval between specimen collection and receipt at a reference laboratory was 2 days (range: <1 21="21" 39="39" 57="57" agglutination="agglutination" although="although" and="and" as="as" at="at" bacterial="bacterial" blood="blood" but="but" cell="cell" collected.="collected." collected="collected" confirmed="confirmed" count="count" csf="csf" culture="culture" days="days" district="district" examination="examination" findings="findings" from="from" gram="gram" in="in" laboratories="laboratories" laboratory="laboratory" latex="latex" macroscopic="macroscopic" meningitis="meningitis" national="national" nationally="nationally" not="not" of="of" one="one" p="p" performed="performed" proportion="proportion" reference="reference" results="results" retesting="retesting" routinely="routinely" specimens.="specimens." specimens="specimens" stain="stain" suggested="suggested" tests="tests" that="that" the="the" these="these" to="to" was="was" were="were" white="white">

Reported by

Mamoudou Djingarey, MD, Denis Kandolo, MD, Clement Lingani, MSc, Fabien Diomandé, MD, World Health Organization West Africa Inter-Country Support Team, Burkina Faso. Isaïe Medah, MD, Ludovic Kambou, MD, Felix Tarbangdo, Ministère de la Santé, Burkina Faso. Seydou Diarra, MD, Kandioura Touré, MD, Flabou Bougoudogo, PhD, Ministère de la Santé, Mali. Sema Mandal, MD, Ryan T. Novak, PhD, Amanda C. Cohn, MD, Thomas A. Clark, MD, Nancy E. Messonnier, MD, Div of Bacterial Diseases, National Center for Immunizations and Respiratory Diseases, CDC. Corresponding contributor: Sema Mandal, smandal1@cdc.gov, 404-639-3158.

Editorial Note

High-quality surveillance with laboratory confirmation is necessary to evaluate vaccine effectiveness, inform vaccination strategies to maintain population immunity, and monitor for changes in disease epidemiology. In this evaluation of meningitis surveillance in Burkina Faso and Mali, good organizational structures, capable staff, and clear protocols for collecting both aggregate and case-level data and collecting CSF specimens were found. However, a major gap was that case-level data and specimens often were not sent to the national level for analysis. Harmonized data management tools and linking case identifiers were lacking. Moreover, the ability of the reference laboratories to confirm cases was limited by the low number of submitted specimens, along with delayed specimen transport, and inadequate capacity for testing.
Based on the findings from the evaluation, recommendations were made to Burkina Faso and Mali to improve data management, epidemiology, and laboratory capacity. Since March 2008 in Burkina Faso and December 2010 in Mali, these surveillance domains have been strengthened through baseline assessments, technology transfer, training, and mentorship. This is the model for meningitis surveillance and capacity-building in the meningitis belt (Figure). Surveillance needs assessments were conducted and pilot projects for case-based surveillance were implemented in selected districts, which were subsequently scaled up to the appropriate level in each country. To improve case-level data reporting to the national level, district visits by supervision teams focused on introducing data management tools that included deploying a standardized surveillance database, introducing systemwide linking using unique case identifiers, and conducting training for surveillance officers. Additionally, national level surveillance epidemiologists and data managers were mentored in collating, analyzing, and interpreting data. To improve specimen transport, district visits focused on reconnecting the network and conducted training on appropriate transport conditions. To improve laboratory capacity for case confirmation, real-time PCR§ and external quality-control programs were established at reference laboratories.
Preliminary data from Burkina Faso for 2011 show improvements in surveillance. Compared with 2007, in 2011 the proportion of line-listed cases doubled from 41% to 88%, and the proportion of all reported cases with a specimen reaching a reference laboratory increased from 11% to 85%. With implementation of real-time PCR in four national reference laboratories, causative pathogen confirmation increased from 24% to 41%. In Mali, most surveillance-strengthening activities are still in progress, but compared with 2007, early 2012 indicators are encouraging. Two of the first districts to introduce PsA-TT now send electronic line-list data to the national level, the proportion of districts submitting specimens has increased from 61% to 80%, and PCR has been introduced at the national reference laboratory (conventional PCR in 2009, real-time PCR in 2011). In Burkina Faso, high-quality surveillance data revealed the impact of PsA-TT 1 year after it was introduced, with significant decreases in the incidence of all bacterial meningitis, serogroup A–specific meningococcal disease, and bacterial meningitis mortality, with no outbreaks identified (6). In Mali, no meningitis outbreaks have occurred in 2012, and preliminary surveillance data have not identified serogroup A disease (7).
Burkina Faso and Mali differed in how they built on existing infrastructure to establish case-based surveillance. Depending on local capacity, populations at risk, disease incidence, and geographic distribution, subnational rather than nationwide population-based case-based surveillance might be appropriate. For example, although Burkina Faso and Mali are neighbors with similar sized populations (15–16 million) and a history of meningitis epidemics, disease epidemiology over the past decade has differed substantially. The incidence of meningitis disease in Burkina Faso is one of the highest in Africa, with a mean annual incidence of 90 per 100,000 during 2005–2009. The last major epidemic was in 2007, with 25,695 cases. Mali has a much lower mean annual incidence, seven per 100,000 during 2005–2009, and the last major epidemic was in 1997, with 11,228 cases.
Unlike Burkina Faso, which lies entirely within the meningitis belt, Mali's northern, sparsely populated desert regions do not. Therefore, Mali concentrated its surveillance-strengthening efforts on the most populous districts in the meningitis belt to achieve a high proportion of laboratory-confirmed cases. The experience of case-based surveillance in Burkina Faso and Mali has shown that one size might not fit all, but key factors for achieving surveillance objectives are conducting baseline surveillance evaluations, placing a high priority on developing surveillance expertise (e.g., through staff training and development), and building on existing infrastructure.
The public health goal of introducing a serogroup A meningococcal conjugate vaccine is to eliminate meningitis epidemics in sub-Saharan Africa.¶ Strong case-based surveillance with pathogen-specific laboratory confirmation is essential to enable accurate assessments of vaccine effectiveness, vaccine failures, duration of protection, and herd immunity. Assessment of all of these factors will help define a national vaccination strategy to maintain population immunity so that epidemics do not recur. Such surveillance also enables identification of susceptible populations that might emerge as a result of low vaccine coverage or loss of vaccine potency during vaccine storage and handling. Additionally, case-based surveillance is essential to detect other meningococcal serogroups and other meningitis pathogens with epidemic potential. Finally, case-based meningitis surveillance can be of even greater value in the many countries that have introduced Haemophilus influenzae type b vaccines and in those that plan to introduce pneumococcal conjugate vaccines, providing necessary information on vaccine effectiveness and changes in the epidemiology of meningitis following implementation of the vaccination programs.

References

1. Lapeyssonnie L. Cerebrospinal meningitis in Africa. Bull World Health Organ 1963;28(Suppl).
2. CDC. Updated guidelines for evaluating public health surveillance systems: recommendations from the Guidelines Working Group. MMWR 2001;50(No. RR-13).
3. World Health Organization. Control of epidemic meningococcal disease. WHO practical guidelines. 2nd ed. Geneva, Switzerland: World Health Organization; 1998.
4. World Health Organization Regional Office for Africa. Standard operating procedures for enhanced meningitis surveillance in Africa. Geneva, Switzerland: World Health Organization; 2005.
5. World Health Organization Regional Office for Africa. Guide générique pour la surveillance cas par cas des méningites bactériennes dans la région Africaine de l'OMS. Geneva, Switzerland: World Health Organization; 2009.
6. Novak RT, Kambou JL, Diomande FV, et al. Serogroup A meningococcal conjugate vaccination in Burkina Faso: analysis of national surveillance data. Lancet Infect Dis 2012;12:757–64.
7. Mandal S, Diarra S, Touré KT, et al. Meningitis surveillance in Mali: monitoring the elimination of epidemic meningitis. Presented at the 2012 International Conference on Emerging Infectious Diseases, March 13, 2012, Atlanta, GA.

* Suggestive of bacterial meningitis: any suspected case with gram-negative cocci; gram-negative rods or gram-positive cocci in cerebrospinal fluid (CSF) by direct microscopic examination; or a leukocyte count of >10 per µL; or turbid or purulent macroscopic appearance.

† Confirmed bacterial meningitis: isolation or detection in CSF by latex agglutination or polymerase chain reaction of Neisseria meningitidis, Streptococcus pneumoniae, Haemophilus influenzae, or other bacterial pathogens known to cause meningitis.

§ Advantages of real-time over conventional PCR include the following: 1) in real-time PCR, amplification products are measured quantitatively each amplification cycle by measuring the fluorescence of a dye, whereas in conventional PCR, amplification products are detected only after the last amplification cycle when the products are separated by gel electrophoresis and stained; 2) real-time PCR is more sensitive than conventional PCR; and 3) real-time PCR amplification is performed in a closed system, whereas amplification in conventional PCR is performed in an open system, allowing a greater chance of contamination.

¶ Additional information available at http://www.meningvax.org/mission.php.