The rationale for integrated childhood meningoencephalitis surveillance: a case study from Cambodia
Sok Touch a, John Grundy b, Susan Hills c, Manju Rani d, Chham Samnang e, Asheena Khalakdina f & Julie Jacobson g
a. Communicable Disease Control Department, Ministry of Health, Phnom Penh, Cambodia.
b. Nossal Institute for Global Health, University of Melbourne, 187 Grattan Street, Melbourne, Vic., Australia.
c. PATH, Seattle, WA, United States of America (USA).
d. Expanded Programme on Immunization, World Health Organization Regional Office for the Western Pacific, Manila, Philippines.
e. PATH, Phnom Penh, Cambodia.
f. PATH, Bangkok, Thailand.
g. Bill & Melinda Gates Foundation, Seattle, WA, USA.
Correspondence to John Grundy (e-mail: firstname.lastname@example.org).
(Submitted: 11 March 2008 – Revised version received: 07 August 2008 – Accepted: 03 September 2008 – Published online: 04 March 2009.)
Bulletin of the World Health Organization 2009;87:320-324. doi: 10.2471/BLT.08.052951
Neurological infection is an important cause of death and disability in children in Asia.1–4 Major vaccine-preventable etiologies of meningoencephalitis (ME) in Asia include Japanese encephalitis (JE) virus and bacteria such as Haemophilus influenzae type b (Hib), Neisseria meningitidis and Streptococcus pneumoniae. Public health initiatives to control these diseases are becoming more feasible with improved vaccine availability and affordability.5
However, in many Asian countries, the epidemiology and public health burden of JE and bacterial meningitis are poorly understood. Generation of disease-specific data for ME was spearheaded in some countries by the establishment of vertical disease-specific initiatives for control of JE, Hib and pneumococcal diseases (e.g. JE project at PATH, and the GAVI Alliance’s Hib Initiative and pneumoADIP). However, the benefits of combining surveillance for meningitis and encephalitis are evident in terms of case identification, simplified logistics and systems, and the potential for more coordinated data analysis and consistent information to assist decision-makers in relation to vaccine introduction programmes.
This paper provides an example of development of an integrated ME surveillance system in Cambodia and, on the basis of this case study, presents the rationale and challenges for design and operation of such systems more generally in Asia.
System design for ME surveillance
A JE sentinel surveillance system was developed and embedded within the routine ME syndromic surveillance system in Cambodia in 2006. The sentinel system was designed so that it could provide a platform to incorporate laboratory testing for other central nervous system (CNS) infections in children.
The goal of the ME surveillance system in Cambodia is to assess disease burden due to neurological infection in children. Weekly reporting to the national level on the number of clinical syndromic ME cases and deaths is required from all district and provincial hospitals across the country. The syndromic ME surveillance is part of the national outbreak surveillance and response system.
The JE sentinel surveillance system was incorporated within this system. Six hospitals were chosen as sentinel sites from geographically diverse parts of the country. ME patients have epidemiological data gathered at these sites. When cerebrospinal fluid and blood are collected for routine case management, tubes are also collected for specific etiology testing. These additional samples are transported weekly to the National Institute of Public Health laboratory in the capital, Phnom Penh. Initially, only JE diagnostic testing by enzyme-linked immunosorbent assay (ELISA) was conducted. The system has been expanded recently and testing for vaccine-preventable etiologies of bacterial meningitis is now being added.
In the first year of surveillance, 47 of 275 (17.1%) ME cases reported from six sentinel sites were laboratory-confirmed as JE. The initial findings from the JE sentinel surveillance system are consistent with results of several previous research studies that have indicated approximately 20–30% of all acute encephalitis cases in Cambodia are attributable to JE virus infection.6,7 With this preliminary data, an estimate of national ME incidence of 42.6 cases per 100 000 children aged less than 15 years and a minimum JE incidence of 7.3 per 100 000 children aged less than 15 years were calculated. Although many factors limit the precision of this estimate, the figure provides a useful estimate of childhood JE incidence in Cambodia.
Many challenges were confronted in the planning and implementation of the surveillance system. In the following section, we outline these challenges and the responses made.
Challenges and responses with system development
Determination of case definition
The WHO standards for surveillance of vaccine-preventable diseases do not currently include a case definition for meningoencephalitis, although there are separate standards for bacterial meningitis and for acute encephalitis syndrome/JE.8
The clinical case definition that had been developed when syndromic ME surveillance was established in Cambodia in 2005 was “a person with acute onset of fever (≥ 38 °C) and at least one of the following: neck stiffness, altered consciousness, other meningeal sign”. This case definition more closely resembles the WHO-recommended case definition for bacterial meningitis than acute encephalitis.8 However, because the definition was considered to be sufficiently broad to include most presentations of encephalitis, and clinicians were already familiar with the definition, it was maintained when JE sentinel surveillance was incorporated in 2006.
Collection and testing of clinical samples
Collection of cerebrospinal fluid – the key biological sample required for ME surveillance – is often a challenge. Health systems in developing countries are not always equipped to routinely conduct lumbar punctures, staff have not always been trained to conduct them and hesitancy may occur among clinicians as well as family members because of perceived risks. The prioritization for use of a (potentially limited) cerebrospinal fluid sample must also be clearly defined. The first priority for all specimens must be to guide the immediate management and treatment of the patient. Logistics may also be a challenge. Some diagnostic tests cannot be conducted in the local hospital laboratory and samples need to be transported. Requirements for storage and transport may differ when different types of diagnostic tests are conducted. Good coordination of clinicians, laboratory technicians and others who may handle the specimens is required.
In addition to standard operating procedures for specimen collection and management, which included a schema for appropriate laboratory testing of specimens, ME case management guidelines and associated training programmes were developed as a complementary strategy to surveillance capacity-building. Indicators for assessing effective functioning of the surveillance system were defined and monitored.
Managing the limitations
Underestimation of population incidence of disease attributable to specific etiologies may occur as a result of limitations of currently available diagnostic testing methodologies. Timing of specimen collection or prior use of antibiotics can affect the likelihood of a positive result. In addition, different testing methodologies used (e.g. ELISA for JE diagnosis, or bacterial culture, polymerase chain reaction or latex agglutination antigen tests for diagnosis of causes of bacterial meningitis) have different sensitivities and specificities.
There are also challenges with relying on sentinel surveillance to provide data to determine the national picture of disease burden. In most Asian countries, laboratory testing of every ME patient nationwide is not feasible. Hence, the accepted model is sentinel surveillance, with testing of samples from patients at selected sites only. If sentinel sites are not truly representative, biases in disease burden estimates occur. For example, if there is less access to hospital facilities for rural rather than urban populations, then JE, a predominantly rural disease, may be under-represented.
To estimate national incidences of particular vaccine-preventable CNS diseases, accurate syndromic ME data are also needed. Under-reporting may occur due to multiple factors. For example, individuals who die before presentation or do not access hospital facilities will not be included. Lack of reporting from private-sector health services may also affect results.
Systems to ensure quality of laboratory data were implemented, including a laboratory quality assurance programme, and individual strategies such as encouraging collection of convalescent serum samples for JE diagnostic testing. The need for careful interpretation of laboratory data was reinforced, ensuring it was understood that the percentage of cases due to a particular viral or bacterial etiology could not necessarily be directly compared based on laboratory results.
Sentinel sites were selected after assessment visits and were carefully chosen from diverse geographical areas, also taking into account other factors including capacity of the hospital to collect and transport specimens. Monitoring of quality of data collection was undertaken during routine supervision visits.
Common causes of ME in Asia – JE, Hib and pneumococcal disease – are all vaccine preventable. With Hib and pneumococcal disease, the impact of immunization is not just on CNS disease but on respiratory and other invasive disease. These vaccines could therefore significantly reduce death and disability among children. Data from the recently established sentinel ME surveillance system and more substantial data from the region, and globally, suggest that 50% or more CNS infections in children in Cambodia could be prevented through vaccination.
There is a risk that a focus on individual neurological diseases may result in the emergence of a wide variety of vertical laboratory testing and reporting systems that do not coordinate with national health information systems, leading to fragmentation and inefficiencies in data collection and reporting. This is the single most important lesson learned from the development of other vaccine-preventable disease surveillance systems such as acute flaccid paralysis, measles and tetanus. Surveillance systems are often de-linked in terms of planning, financing and data collection and analysis, resulting in widespread inefficiencies and duplication of scarce financial and human resources.
In relation to feasibility of integration of meningitis and encephalitis surveillance, it is crucial to consider that we are dealing with one syndrome and one specimen collection procedure. In addition, the overall operating system – consisting of the patient, the clinician, the surveillance forms, the on-site laboratory testing, the transport system and the reference laboratory structure – remains constant. Based on experience from Cambodia, Table 1 summarizes the rationale for integration of ME surveillance, Fig. 1 provides an overview of an integrated ME surveillance structure and Box 1 summarizes the lessons learned.
Fig. 1. Overview of an integrated meningoencephalitis surveillance system
Box 1. Lessons learned
Sentinel surveillance systems, when linked to syndromic reporting systems, can characterize the epidemiology of meningoencephalitis and identify the proportion of hospital-based neurological infection in children that is vaccine preventable.
Integrated systems enable consistency in data collection, analysis and information dissemination and enhance the capacity of public health managers to provide more credible and integrated information to policy-makers.
This will assist decision-making about the potential role of immunization in reducing the incidence of childhood neurological infections.
Integrated ME surveillance is in line with WHO’s newly released strategy for vaccine-preventable disease surveillance, the Global Framework for Immunization Monitoring and Surveillance.10,11 Four of the seven goals of the framework include linking epidemiological and laboratory surveillance, building surveillance capacity at the country level for disease burden estimates and impact monitoring, expanding laboratory networks for viral and bacterial diseases and, finally, linking with other surveillance and monitoring systems for early detection and response to emerging infections. Integration of meningitis and encephalitis surveillance has the potential to successfully address these goals.
As demonstrated by this case study from Cambodia, an integrated ME surveillance system in Asia has the potential to better define the population incidence and proportion of infectious CNS disease in children that is vaccine-preventable. Public health priorities in the country, availability of viral and bacterial diagnostics, and local disease patterns may all determine the appropriateness of this approach. However, integrated surveillance has the potential to enhance the capability of public health managers to provide more credible and integrated information to policy-makers about the potential role of immunization in reducing childhood CNS infection-related death and disability. Additionally, it would provide a platform for surveillance and investigation of any new or emerging CNS-related infections, monitor the impact of new vaccine programmes and streamline information and reporting in resource-poor public health systems in developing countries. ■
We thank the laboratory staff at the National Institute of Public Health and at hospital sentinel sites, the various clinicians, communicable disease control staff, and national immunization programme staff and managers, who have supported the development of the surveillance system in Cambodia. We would also like to thank the following PATH staff: Kathy Neuzil, acting director of JE project, for her review of the manuscript; Jodi Udd for editing support; and the staff in the Cambodia office.
Funding: The work undertaken was funded by PATH’s Japanese Encephalitis Project, supported by the Bill & Melinda Gates Foundation. As a partnership programme, the Ministry of Health in Cambodia has full control of primary data.
Competing interests: None declared.
- Tsai TF. New initiatives for the control of Japanese encephalitis by vaccination: minutes of a WHO/CVI meeting, Bangkok, Thailand, 13-15 October 1998. Vaccine 2000; 18: 1-25 doi: 10.1016/S0264-410X(00)00037-2 pmid: 10821969.
- Wang CH, Lin TY. Invasive Haemophilus influenzae diseases and purulent meningitis in Taiwan. J Formos Med Assoc 1996; 95: 599-604 pmid: 8870429.
- Solomon T, Dung NM, Kneen R, Gainsborough M, Vaughn DW, Khanh VT. Japanese encephalitis. J Neurol Neurosurg Psychiatry 2000; 68: 405-15 doi: 10.1136/jnnp.68.4.405 pmid: 10727474.
- Chotpitayasunondh T. Bacterial meningitis in children: etiology and clinical features, an 11-year review of 618 cases. Southeast Asian J Trop Med Public Health 1994; 25: 107-15 pmid: 7824999.
- Beasley DW, Lewthwaite P, Solomon T. Current use and development of vaccines for Japanese encephalitis. Expert Opin Biol Ther 2008; 8: 95-106 doi: 10.1517/147125126.96.36.199 pmid: 18081539.
- Chhour YM, Ruble G, Hong R, Minn K, Kdan Y, Sok T, et al., et al. Hospital-based diagnosis of hemorrhagic fever, encephalitis, and hepatitis in Cambodian children. Emerg Infect Dis 2002; 8: 485-9 pmid: 11996683.
- Srey VH, Sadones H, Ong S, Mam M, Yim C, Sor S, et al., et al. Etiology of encephalitis syndrome among hospitalized children and adults in Takeo, Cambodia, 1999-2000. Am J Trop Med Hyg 2002; 66: 200-7 pmid: 12135294.
- WHO-recommended standards for surveillance of selected vaccine-preventable diseases. Geneva: World Health Organization; 2006.
- Disability, including prevention, management and rehabilitation. In: Fifty-eighth World Health Assembly, Geneva, May 2005. Resolutions and decisions. Geneva: World Health Organization; 2005. pp. 97-100.
- Dabbagh A, Eggers R, Cochi S, Dietz V, Strebel P, Cherian T. A new global framework for immunization monitoring and surveillance. Bull World Health Organ 2007; 85: 904-5 doi: 10.2471/BLT.07.048223 pmid: 18278243.
- Global framework for immunization monitoring and surveillance. Geneva: World Health Organization; 2007 (WHO/IVB/07.06).