Expanding severe acute respiratory infection (SARI) surveillance beyond influenza: The process and data from 1 year of implementation in Vietnam

Background In 2016, as a component of the Global Health Security Agenda, the Vietnam Ministry of Health expanded its existing influenza sentinel surveillance for severe acute respiratory infections (SARI) to include testing for 7 additional viral respiratory pathogens. This article describes the steps taken to implement expanded SARI surveillance in Vietnam and reports data from 1 year of expanded surveillance. Methods The process of expanding the suite of pathogens for routine testing by real‐time reverse transcriptase–polymerase chain reaction (rRT‐PCR) included laboratory trainings, procurement/distribution of reagents, and strengthening and aligning SARI surveillance epidemiology practices at sentinel sites and regional institutes (RI). Results Surveillance data showed that of 4003 specimens tested by the RI laboratories, 20.2% (n = 810) were positive for influenza virus. Of the 3193 influenza‐negative specimens, 41.8% (n = 1337) were positive for at least 1 non‐influenza respiratory virus, of which 16.2% (n = 518), 13.4% (n = 428), and 9.6% (n = 308) tested positive for respiratory syncytial virus, rhinovirus, and adenovirus, respectively. Conclusions The Government of Vietnam has demonstrated that expanding respiratory viral surveillance by strengthening and building upon an influenza platform is feasible, efficient, and practical.


| INTRODUC TI ON
In support of the World Health Organization (WHO) and its global network of National Influenza Centers (NIC), the U.S. Centers for Disease Control and Prevention (CDC) has provided pivotal technical assistance to more than 50 countries in the development of surveillance and laboratory capacity for influenza virus. 1,2 These efforts have helped nations describe influenza seasonality, identify influenza isolates for vaccines, and develop plans for pandemic preparedness. Regional and international initiatives such as the Asia Pacific Strategy for Emerging Diseases (APSED) 3

and the Global Health
Security Agenda (GHSA) 4  Many countries that conduct surveillance for severe acute respiratory infection (SARI) routinely test SARI specimens only for the presence of influenza virus 5 although other respiratory viruses clearly play an important role in causing severe respiratory infections. 6 Globally, influenza is identified as the etiologic agent in only about 10%-20% of SARI surveillance cases. [7][8][9] Without adequate reagents and routine procedures for testing non-influenza respiratory viruses, 10 the pathogens causing the remaining 80%-90% of cases from SARI sentinel surveillance are not identified. Data on noninfluenza respiratory viruses are essential for public health institutions to understand the burden of disease due to these pathogens.
In Vietnam, at the national level, the Ministry of Health General Until 2016, all SARI surveillance samples in Vietnam were tested for influenza A and influenza B, and positive samples underwent additional subtyping. Through this study, this algorithm was modified with the following change: all specimens that tested negative for influenza were tested for each of 7 non-influenza viral pathogensrespiratory syncytial virus (RSV); human metapneumovirus (hMPV); human parainfluenza viruses 1 (PIV1), 2 (PIV2), and 3 (PIV3); human rhinovirus (RV); and human adenovirus (AdV)-in a series of 7 singleplex reactions ( Figure 2). Following this modification, the SARI surveillance platform was referred to as expanded SARI surveillance to reflect the additional pathogens that were included in routine testing.
The International Reagent Resource (IRR, formerly known as the Influenza Reagent Resource, https://www.internationalreagentresource.org/), an existing online reagent portal managed by the Influenza Division at CDC, was modified to include primers and probes for these 7 non-influenza respiratory viruses in order to ensure availability of quality reagents. Vietnam's NIC laboratories, NIHE and PI-HCMC, already had access to the IRR for influenza surveillance; through this current study, TIHE and PI-NT were newly authorized to receive these reagents. All reagents procured through the IRR were available to laboratories at no cost.

| Training for non-influenza viral diagnostics and proficiency testing
To ensure accurate use of these diagnostic tests, CDC engaged the RIs in a series of trainings. This included a 4-day in-country training of didactic and laboratory instruction held at the PI-HCMC in August 2015. Topics covered laboratory safety; setup of a molecular diagnostics laboratory; appropriate specimen collection, transport, and storage; overview of real-time reverse transcriptase-polymerase chain reaction (rRT-PCR); software data analysis of rRT-PCR; and a discussion on the implementation of expanded SARI surveillance.
Regional institute personnel were trained to test for the panel of 7 non-influenza respiratory viruses as well as to test for Middle East Respiratory Syndrome coronavirus (MERS-CoV) to increase preparedness in the event of an outbreak. [11][12][13][14][15][16] Following this training, each RI had an additional site visit by CDC experts, who offered guidance that was specific to the needs of each laboratory.
After the first training, each of the RIs was provided with an external quality assessment (EQA) panel for respiratory viruses. This EQA panel provided the laboratory with a series of samples intended to mimic human respiratory specimens. While the types of pathogen were known by CDC, this information was not shared with the F I G U R E 1 A, List of the SARI sentinel sites selected at the harmonization workshop. B, Map illustrating SARI sentinel sites' geographic distribution across Vietnam and the Regional Institute that cooperates with each sentinel site  via patient interview and nasopharyngeal and oropharyngeal swabs were collected from each enrolled patient. In severe cases, endotracheal aspirates were also accepted as specimen types.

| Harmonization and expansion of SARI surveillance
At the harmonization workshop, unified, practical specimen transport guidance was agreed upon to maintain sample integrity and promote sustainability. According to this guidance, both nasopharyngeal and oropharyngeal swabs from an individual patient were stored in a single viral transport media (VTM) tube and transported once a week to the RIs by routine transportation systems.
Previously, the specimens were transported as soon as specimens were collected, resulting in an unsustainable and expensive transportation process.
At the harmonization meeting, sentinel sites for expanded SARI surveillance were selected. The following criteria were used The data reported in this article are from specimens collected and tested following the revised SARI Surveillance Guidelines.

| Laboratory testing and data reporting
Specimens were collected at the sentinel sites and delivered once a week to the laboratories of the RIs. Each of the RIs utilized the reagents procured through the IRR for specimen analysis. At the RI laboratories, total nucleic acid (TNA, ie, both ribonucleic acids and deoxyribonucleic acids) was extracted from the specimens. The 7 non-influenza pathogens were tested by singleplex rRT-PCR assays, designed by the Division of Viral Diseases at the National Center for Immunization and Respiratory Diseases, CDC. [11][12][13][14][15][16] Each week, epidemiology and laboratory data from the 4 RIs were sent to the epidemiology team at NIHE who prepared a weekly report, which was then distributed electronically to key public health stakeholders and partners. Two notable modifications were implemented to strengthen the data reporting process. GDPM designed improved SARI surveillance data reporting forms for RIs to reduce manual data cleaning at the regional and national levels, and GDPM, CDC, and PATH worked collaboratively with the RIs to integrate the SARI surveillance data into the digitized electronic data warehouse managed by the public health emergency operation center at GDPM. 17

| Harmonization and expansion of SARI surveillance
After the harmonization workshop, the SARI Surveillance   Table 3 for results by sentinel site   20 Studies of Vietnamese pediatric patients with acute respiratory infections identified 23.8% and 23% as positive for RSV. 27,28 As RSV is believed to be the most common pathogen causing acute lower respiratory infections in children, 29 it is predictable and consistent with our findings that the proportion of RSV in pediatric-   given the sensitivity of the rRT-PCR platform and the unclear role of non-influenza viruses in virulence, this conclusion warrants further research. [31][32][33][34] The 7 non-influenza respiratory viruses used in this study were selected based on historical data demonstrating their potential involvement in respiratory infections, as well as their potential for future vaccine development. [11][12][13][14][15][16][35][36][37][38][39] Of the 7 viruses, significant progress has been made toward RSV vaccine development; however, despite several potential candidates, some of which advanced to clinical trials, no vaccine so far is both safe and effective. 40 Establishing baseline data for these viruses in people with SARI helps to provide evidence to GDPM to guide public health decision-making. ters. 41 In addition, healthcare utilization is reported to be 3-4 times higher in urban areas than in the mountainous regions in the north or central regions of the country. 42  In spite of these limitations, the existence of a SARI surveillance network in Vietnam provided a platform for collecting data on important non-influenza respiratory viruses, and the process of expanding the system offered an opportunity to revisit the data collection and analysis practices that were routine for SARI surveillance. As a multistep process, the expansion and strengthening of SARI surveillance continued to develop throughout 2016 and 2017, and will likely continue to improve with further attention and refinement to the system. The future integration of the expanded SARI surveillance data into the GDPM data warehouse offers additional potential benefit of reduced delays due to data sharing, reduced need for manual data curation, as well as provision of data to the public health emergency operation center for routine data analysis. 17 The process of expanding SARI surveillance to test for noninfluenza respiratory viruses has engaged numerous organizations in Vietnam ranging from sentinel hospitals, RIs to GDPM and international donor organizations. Additionally, in the era of GHSA the study demonstrates an ability to enhance and expand surveillance by building and extending upon existing in-country capacities.

| D ISCUSS I ON
While these data provide valuable initial insight into the burden of severe respiratory disease, the real value will only be realized over time as data are collected and analyzed consistently through multiple years. Such data can be used to understand seasonality, contribution by various pathogens to respiratory disease burden, and the risk groups of these noninfluenza respiratory viral pathogens in Vietnam.

ACK N OWLED G EM ENTS
The authors would like to thank all of the individuals from GDPM, the 4 RIs, and the SARI sentinel sites who contributed to the im-

D I SCL A I M ER
The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.