Our aim with this meeting was to explore establishment of a representative national serosurveillance network for Vietnam to continue seroepidemiology work and to assess the need for, and how best to use, serosurveillance in policy making in Vietnam. We envisaged that this network would help assess coverage of vaccination programs and identify at-risk regions, model health and economic benefits of future vaccine implementation, and provide information about the presence and risk of novel pathogens.

At this meeting, we discussed the present design of and results from serum collections established in Vietnam by NIHE, OUCRU, and their partners and by Public Health England and its partners; public health questions relevant for Vietnam that such a network should be able to answer and availability of assays to address these questions; ways of collecting serum (hospital, outpatient clinics, community) and their benefits and biases and how to best understand and avoid these biases; and feasible ways to collect representative samples for national serosurveillance in Vietnam, making optimal use of existing serosurveillance programs (OUCRU) and demographic surveillance sites.

We present a summary of the discussions and the main outcomes of the meeting. This report provides a roadmap for the implementation of a nationally representative serosurveillance network.

Meeting participants decided that the main focuses of serosurveillance in Vietnam would be to test for antibodies to pathogens for which concerns exist about vaccination uptake or current schedules (e.g., measles, pertussis, tetanus, and diphtheria); vaccines may potentially be introduced in the future (e.g., enterovirus A71, influenza virus, and pneumococcal infection); and the extent of transmission is currently unknown or not fully quantified in Vietnam (e.g., Zika virus, avian influenza virus, or other emerging pathogens). For all of these considerations, use of serosurveillance relies on there being commercially available assays or availability of easily implementable, interpretable, and replicable in-house assays. For many vaccine-preventable diseases (e.g., measles or tetanus), these assays have been developed and are widely used, although interpretation may still be complicated (32). For other pathogens (e.g., enterovirus A71 or Bordetella pertussis), careful thought must be given to the assays used.

For setting up serosurveillance or serologic studies, sampling techniques are debated. In general, collections can be done in 2 ways: random population-based sampling (e.g., the NIHE hepatitis B and C studies) and convenience sampling (e.g., the OUCRU serum banks). Random population-based collection has the advantage of being more representative of the population being sampled; however, it is very expensive and labor-intensive and therefore cannot be conducted regularly. Convenience sampling introduces bias because the sampled population may not be representative. However, this type of sampling is less expensive and easier to undertake and therefore can be conducted more regularly and may still provide relevant information, as has been illustrated for measles and tetanus in Vietnam (14,31).

Because of the pros and cons of each sampling approach, meeting participants decided to use a hybrid approach in Vietnam, extending previous studies undertaken in the country and learning from lessons in other countries. The hybrid approach will help clarify the biases in the convenience sampling methods, some of which may be specific to Vietnam. These biases include health-seeking behaviors and their association with factors such as socioeconomic status, ethnic group, and vaccination uptake. The locations will be chosen across the country to be as representative as possible, which is particularly necessary in a country with wide geographic and demographic variation.

The backbone of the network was envisioned to consist of hospital-based residual serum collections with the following characteristics, extending and bringing together the 2 existing OUCRU studies in Vietnam. For the hospital-based collection of residual serum, we envisage collecting serum, stratified by patient age, from 20–25 centers. We intend to have age-stratified collection of serum with 1-year bands for patients 1–20 years of age and 5-year bands for patients 20–70 years of age, collecting a total of 200 samples every 3 months and storing them at −20°C, as has been done in previous studies (14,25). This method will ensure that collection is in line with the standards of the European Sero-Epidemiology Network (16). The centers (2–4 centers per region) will represent the 8 ecologic regions in Vietnam. These samples will be stored in Vietnam, with the possibility of sample sharing if approved by NIHE. This study will be submitted for approval by the NIHE, Oxford Tropical Research Ethics Committee, and OUCRU institutional review boards for testing for antibodies to pathogens relevant to Vietnam and with waived consent, as in previous studies in Europe (9). The serum bank will be funded on a per-project basis, but a business plan will be developed for national support through NIHE moving forward.

In tandem with setting up this hospital-based network, we will compare it with population-based studies already conducted by NIHE in Vietnam. This comparison will enable us to assess the biases of this type of collection by testing samples from the same time and location from both hospital-based and population-based serum collections concurrently for antibodies to the same antigen. The results will then be compared across patient ages and locations to look for areas where results are discordant, and we will attempt to summarize associated factors. Possible antigens for this testing include those against C. tetani and measles, hepatitis B, and hepatis C viruses. This comparison could be of interest to other countries considering the different ways to set up serum collection. We will also continue to assess any possibility of adding serum collection and testing to other data collection efforts.

For all testing of samples, we will work with Public Health England and other national reference institutes to choose sample storage, platforms, and assays that can be implemented in Vietnam for high throughput multiplex analysis of locally relevant diseases. In past work on this serum bank in Vietnam, individual ELISA-type tests have been used for the vaccine-preventable diseases (14), and multiplex assays have been used within pathogen groups (24). However, moving forward, to decide what best to use, we will explore use of multiplex assays (such as a Luminex assay) across multiple pathogens, considering cost, sample volume used, laboratory equipment required, ease of standardization, and information gained.

The results of studies of this serum bank will be communicated in scientific publications and in regular discussion with NIHE and the Vietnam Ministry of Health. OUCRU will tailor its public engagement program on vaccination to results of this project.

We believe that serosurveillance could form an integral part of Vietnam’s future surveillance strategy for infectious diseases, particularly for vaccine-preventable diseases and emerging infections. Such systems have been of great use globally. In Vietnam, we are able to build on the current strengths of research and public health bodies to develop a nationwide surveillance system. Creating this system within existing public health bodies would additionally benefit from preestablished and functional relationships with policy makers. This system will be a vital tool for Vietnam in the next era of infectious disease control.