key: cord-0740835-qtwuh2bl
authors: Subissi, Lorenzo; Bossuyt, Nathalie; Reynders, Marijke; Gérard, Michèle; Dauby, Nicolas; Lacor, Patrick; Daelemans, Siel; Lissoir, Bénédicte; Holemans, Xavier; Magerman, Koen; Jouck, Door; Bourgeois, Marc; Delaere, Bénédicte; Quoilin, Sophie; Van Gucht, Steven; Thomas, Isabelle; Barbezange, Cyril
title: Spotlight influenza: Extending influenza surveillance to detect non-influenza respiratory viruses of public health relevance: analysis of surveillance data, Belgium, 2015 to 2019
date: 2021-09-23
journal: Euro Surveill
DOI: 10.2807/1560-7917.es.2021.26.38.2001104
sha: 6b4d9c43a16dee2c1a773ef0e80cdde179645731
doc_id: 740835
cord_uid: qtwuh2bl

BACKGROUND: Seasonal influenza-like illness (ILI) affects millions of people yearly. Severe acute respiratory infections (SARI), mainly influenza, are a leading cause of hospitalisation and mortality. Increasing evidence indicates that non-influenza respiratory viruses (NIRV) also contribute to the burden of SARI. In Belgium, SARI surveillance by a network of sentinel hospitals has been ongoing since 2011. AIM: We report the results of using in-house multiplex qPCR for the detection of a flexible panel of viruses in respiratory ILI and SARI samples and the estimated incidence rates of SARI associated with each virus. METHODS: We defined ILI as an illness with onset of fever and cough or dyspnoea. SARI was defined as an illness requiring hospitalisation with onset of fever and cough or dyspnoea within the previous 10 days. Samples were collected in four winter seasons and tested by multiplex qPCR for influenza virus and NIRV. Using catchment population estimates, we calculated incidence rates of SARI associated with each virus. RESULTS: One third of the SARI cases were positive for NIRV, reaching 49.4% among children younger than 15 years. In children younger than 5 years, incidence rates of NIRV-associated SARI were twice that of influenza (103.5 vs 57.6/100,000 person-months); co-infections with several NIRV, respiratory syncytial viruses, human metapneumoviruses and picornaviruses contributed most (33.1, 13.6, 15.8 and 18.2/100,000 person-months, respectively). CONCLUSION: Early testing for NIRV could be beneficial to clinical management of SARI patients, especially in children younger than 5 years, for whom the burden of NIRV-associated disease exceeds that of influenza.

Acute viral infections of the respiratory tract are common in humans. According to the World Health Organization (WHO), complications such as lower respiratory tract infections and pneumonia are among the main causes of mortality in children and elderly people worldwide [1] . The burden attributed to seasonal influenza virus has long received most of the attention [2] , but the involvement of non-influenza respiratory viruses (NIRV), such as respiratory syncytial virus and human metapneumovirus, is increasingly documented [3, 4] . However, the burden of NIRV [5] [6] [7] compared with seasonal influenza [8] is not sufficiently estimated [9] . These data would allow to better understand the need for enhanced surveillance of these viruses during winter seasons with the aim of improving patient management.

The WHO implemented the influenza surveillance network in 1952 to monitor the continuous antigenic changes of the virus and to guide vaccine composition. Following the 2009 influenza pandemic, the scope of the surveillance was extended to better assess the severity and burden of influenza viruses. In Belgium, influenza virus surveillance is organised by the national public health institute, Sciensano, which also hosts the National Influenza Centre (NIC). Two sentinel networks are in place: (i) surveillance of influenza-like illness (ILI) in the community, based on general practitioners (GPs) providing information on mild cases (in place since the mid-1970s) and (ii) surveillance of severe acute respiratory infection (SARI) based on six sentinel hospitals providing information on hospitalised patients (set up after the 2009 influenza pandemic and operating every season since 2011). Starting with the 2015/16 influenza season, the NIC introduced systematic testing of NIRV by qPCR in addition to testing and characterising influenza viruses. We report here the results of systematic NIRV testing of ILI and SARI surveillance samples for four seasons (2015/16 to 2018/19) and estimate the burden and severity of NIRV compared with influenza.

Sentinel-based ILI surveillance was performed through a network of GPs widely distributed throughout Belgium, representing ca 1.5% of all Belgian GPs and covering more than 1.1% of the population. Patients were enrolled during each influenza season surveillance period (from week 40 to week 20). Recommendations were that nasopharyngeal swabs were obtained weekly from the first two ILI patients belonging to two different households. SARI patients were recruited through the sentinel network of six hospitals in Belgium, two in each administrative region of the country: Flanders, Wallonia and Brussels-Capital. It was recommended that paediatric and adult units systematically collect detailed clinico-epidemiological data and a respiratory specimen from all patients meeting the SARI case definition during the epidemic period of seasonal influenza (starting and ending weeks between January and April, depending on influenza activity).

Based on WHO guidelines [10] , ILI and SARI cases were defined as acute respiratory illness with onset within the last 10 days, with measured or reported fever of ≥ 38 °C and with cough and/or dyspnoea. In addition, overnight hospitalisation was a required criterion for SARI cases. Patients who did not give informed consent (either directly or through a parent or legal guardian) were not enrolled in the study nor sampled.

Standardised questionnaires were used to collect data on age, sex, clinical signs included in the case definition, status of vaccination against influenza viruses, administration of a neuraminidase inhibitor antiviral and/or antibiotic treatment, and co-morbidities known as risk factors of severity. Follow-up data during hospitalisation were also reported for the SARI cases to evaluate disease severity and included the detection of pneumonia based on chest radiography, and/or the development of acute respiratory distress syndrome, and/or the requirement for respiratory assistance and/ or for extracorporeal membrane oxygenation, and/or the admission to an intensive care unit, and/or death (all-cause death).

Flow diagram for participant inclusion in the study for surveillance of influenza-like illness and severe acute respiratory infections, Belgium, 2015-2019 (n = 9,734) 

Respiratory samples were analysed at the Belgian National Influenza Centre. Viral nucleic acids were extracted using NucliSENS EasyMag (BioMerieux Benelux, Brussels, Belgium). The following respiratory virus targets were detected using several in-house multiplex qPCRs: influenza virus types A and B (and subsequent subtype/lineage), respiratory syncytial virus types A and B, human metapneumoviruses, parainfluenzavirus types 1, 2, 3 and 4, coronaviruses CoV-OC43, CoV-NL63 and CoV-229E, adenoviruses, picornaviruses (Rhinovirus and Enterovirus genera), specific enterovirus D68, parechovirus and bocavirus as previously described [11] .

To evaluate some aspects of the burden of disease of virus-associated SARI, the catchment population estimates from 2017 (mid-year of the study period) were used to calculate monthly incidence rate per 100,000 population during the winter season, based on WHO recommendations [10] . Catchment populations of the SARI hospitals by age group were estimated based on hospital admission data (i.e. the proportion of the different municipalities) and the total population of these municipalities; this rough estimation has the advantage that it is easily extracted without extra workload for the hospitals. The number of months covered by the study period (i.e. 14.9 months), used for the denominator, was calculated by adding up the number of weeks with active SARI surveillance for each season (i.e. 64 weeks), multiplied by 7 days and divided by the average number of days in a month during the winter period (December to April, i.e. 30 days). Age groupspecific catchment populations of the six hospitals were summed up to calculate the monthly incidence rate by age group (< 5, 5-14, 15-64 and ≥ 65 years) for virus-associated SARI. The 95% confidence intervals (CI) were calculated using the Rothman-Greenland method [12] .

The SARI surveillance protocol was approved by a central Ethical Committee (reference AK/12-02-11/4111; in 2011: Centre Hospitalier Universitaire St-Pierre, Brussels, Belgium; from 2014 onwards: Universitair Ziekenhuis Brussel, Brussels, Belgium) and the local ethical committees of each participating hospital. Informed consent was obtained from all participants or parents/guardians.

Over the four seasons under study 1,791 ILI ( Figure  1A ) and 4,774 SARI ( Figure 1B) Table S1 ). Table  S2 ). Co-detections were more common among SARI patients than ILI patients. They accounted for 11.3% of the overall number of tested SARI samples (541/4,774), with influenza virus-positive (n = 250) and influenzanegative (n = 291) co-detections representing 5.2 and 6.1%, respectively.

The proportion of NIRV detection (excluding co-infections with influenza viruses) varied between 12.8% in 2017/18 and 24.3% in 2018/19 in the ILI samples ( Figure 2A) . In SARI patients, the proportion was higher and varied between 24.2% in 2017/18 and 33.9% in 2018/19 ( Figure 2B ). Within a season, the proportions of picornaviruses and coronaviruses were similar among ILI and SARI patients. On the contrary, hMPV and RSV detection rates were systematically higher in SARI than in ILI samples. Except for season 2015/16, these two viruses represented the largest groups in SARI patients, whereas picornaviruses and Figure S1 ) and RSV in the SARI surveillance, with the detection of the end of the RSV epidemic, which usually occurs before the influenza epidemic in western Europe [13] including in Belgium [14] (Figure 3 ). 

In children (younger than 15 years), the proportion of samples positive for NIRV (single or co-detection) was higher in SARI than ILI patients ( respectively, Supplementary Table S3 ), but also from older adults (21.9%; 52/237 and 29.0%; 20/69, respectively). Finally, SARI samples with co-detection of NIRV or of NIRV with influenza viruses were largely from children younger than 5 years (86.3%; 251/291 and 51.6%; 129/250, respectively, Supplementary Table S3 ).

The burden of NIRV captured by influenza surveillance was partially evaluated by calculating the incidence rates of SARI associated with each virus within specific age groups and for all the samples ( Figure  4 and Supplementary Table S4 ). NIRV incidence rates were noticeably higher in the youngest age group (< 5 years) than in the other age groups or the overall population, with RSV, hMPV and picornaviruses contributing the most as single infection (13.6, 15.8 and 18.2 per 100,000 person-months, respectively), followed by adenoviruses, bocavirus, coronaviruses and parainfluenzaviruses (7.4, 5.8, 5.0 and 4.6 per 100,000 personmonths, respectively). In this age group, the incidence rate of SARI associated with all kinds of influenza-negative NIRV infections was almost double that associated with influenza (103.5 vs 57.6 per 100,000 personmonths). Incidence rates of SARI associated with NIRV were low in the other age groups, in contrast to those associated with influenza ( Figure 4 ). Despite that, the incidence rates of SARI associated with RSV, hMPV and coronaviruses in older adults were more than twice that in the overall population (3.9, 4.4 and 3.2 vs 1.8, 2.1 and 1.2 per 100,000 person-months, respectively).

Here, we report on the use of several in-house multiplex qPCR assays on respiratory specimens collected through two sentinel networks of influenza surveillance (ILI and SARI) to detect a broad range of respiratory viruses of public health relevance during the winter season. Systematic testing of NIRV on samples was introduced in Belgium from the winter season 2015/16 onwards, with the use of in-house multiplex qPCRs offering the flexibility to change the panel of viruses/virus groups depending on the situation. We found that NIRV testing may explain up to one third of infections and becomes particularly useful for diagnostic purposes of paediatric SARI patients. By estimating incidence rates, we showed that NIRV even represent a higher burden than influenza viruses in the SARI patients under the age of 5 years.

A limitation of our study is that these sentinel surveillance networks collect data only during the winter season in Belgium. This is particularly true for the SARI network which operates only during the period of influenza virus activity. In addition, since the objective of the SARI surveillance is primarily to evaluate the severity of influenza viruses, the definition used might not be optimally adapted for NIRV. The criteria 'fever' and 'cough' were chosen to specifically target influenza viruses [15] , but they are also characteristic of most NIRV, including RSV and hMPV [16] , the two main NIRV identified in our study. Despite the differences in the case definition and the period of surveillance, our results remain comparable to those obtained in other countries regarding the NIRV identified and their proportions [17, 18] , but the detection rates observed here for the different NIRV might not represent their whole circulation patterns. For example, it was shown that the intensive circulation period for RSV usually precedes that of influenza viruses in western Europe [13] . Although RSV, hMPV, coronaviruses and other NIRV do not strictly follow the same pattern as influenza viruses [19] , many tend to also circulate during the winter season in temperate regions. Interactions between respiratory viruses have been reported [20, 21] , including a negative association between rhinovirus and RSV infection in children [22] . There have also been reports of an increased risk of NIRV infections in children who received inactivated influenza vaccine in a small-scale study [23] , although this was not observed in a larger study encompassing data from seven seasons [24] . Influenza sentinel surveillance thus appears as a relevant system for the study of NIRV [25, 26] and we consider that extending the period of SARI surveillance in Belgium would be beneficial to better estimate their burden but also, on a long-term perspective, the potential impact of the new severe acute respiratory syndrome coronavirus 2 after the current pandemic phase, if it establishes itself as a seasonal virus.

Another limitation is that paediatric and geriatric samples are underrepresented in ILI patients, probably due to sampling bias (GPs in Belgium being less likely to have children or older people as patients or to sample them).

The systematic testing for NIRV allowed the detection of a pathogen in only one eighth of the ILI patients, suggesting that the additional effort of testing NIRV in this study population may be of limited interest and not cost-effective. In contrast, samples were NIRV-positive in almost a third of the SARI patients, in whom the morbidity and mortality associated with respiratory infection is the highest, and in almost half of the paediatric SARI patients. The burden associated with NIRV in SARI patients was age-dependent and particularly high for young children (< 5 years). While co-infections with several NIRV were rarely detected among ILI patients, they were almost as common as influenza virus infection among paediatric SARI patients younger than 5 years. The non-negligible burden associated with RSV should also be taken into consideration in the coming years, when RSV vaccines become available and vaccination is implemented [27] . This clearly shows the benefit that testing SARI patients for NIRV infections can have on the surveillance trend analysis but also, to a lesser extent, on clinical management.

Despite an important decrease in the number of specimens that were reported as negative in SARI patients following the introduction of NIRV testing, more than one fourth of SARI patients remained negative for all tested viruses. Although viruses are considered as the main cause of acute respiratory tract infections [28] [29] [30] and our methodology covers the main respiratory viruses that are involved [31] , it is possible that some samples were positive for viruses that were not targeted by our multiplex qPCRs (e.g. coronavirus HKU-1, influenza C virus, cytomegalovirus, herpes simplex virus), or for bacterial or fungal respiratory pathogens, known to substantially contribute to complications in lower respiratory tract infections [32] [33] [34] . In a prospective study in 11 European countries that looked at the aetiology of lower respiratory tract infections among adults, a bacterial pathogen was detected in one fifth of the participants [35] . The difference in the proportion of 'negative' samples that we observed between age groups (one third among adults and one fifth among children) suggests that the pathogens causing SARI differ between children (more broadly covered by our multiplex PCRs) and adults, which could be due to the clinical signs of the SARI case definition being less specific for adults, especially older adults [36, 37] .

Our study demonstrates the added value of testing NIRV with a flexible method and confirms that the burden of disease associated with NIRV in SARI patients should not be underestimated, especially in children under the age of 5 years for whom it is even higher than that of influenza viruses. Although more expensive than other surveillance systems, SARI sentinel surveillance offers the advantage that it relies on a precise case definition and collects data on complications and co-morbidities. Throughout the years, as data obtained in a standardised manner accumulate, such system could provide better estimates of the burden of NIRV.

Forum of International Respiratory Societies. The global impact of respiratory disease -Second edition

The risk of lower respiratory tract infection following influenza virus infection: A systematic and narrative review

The economic burden of non-influenza-related viral respiratory tract infection in the United States

Other respiratory viruses are important contributors to adult respiratory hospitalizations and mortality even during peak weeks of the influenza season

Burden of human metapneumovirus and respiratory syncytial virus infections in asthmatic children

Respiratory viruses and influenza-like illness: Epidemiology and outcomes in children aged 6 months to 10 years in a multi-country population sample

Burden and risk factors for coronavirus infections in infants in rural Nepal

FLUVAC Study Group. Non-influenza respiratory viruses in adult patients admitted with influenza-like illness: a 3-year prospective multicenter study

Global burden of acute lower respiratory infections due to respiratory syncytial virus in young children: a systematic review and meta-analysis

A manual for estimating disease burden associated with seasonal influenza. Geneva: WHO

Monitoring of human coronaviruses in Belgian primary care and hospitals, 2015-20: a surveillance study

Philadelphia: Lippincott-Raven

European Influenza Surveillance Network. Seasonality and geographical spread of respiratory syncytial virus epidemics in 15 European countries

Capturing respiratory syncytial virus season in Belgium using the influenza severe acute respiratory infection surveillance network, season 2018/19

Clinical signs and symptoms predicting influenza infection

Can we distinguish respiratory viral infections based on clinical features? A prospective pediatric cohort compared to systematic literature review

Viral etiology of acute respiratory infections in hospitalized children in Novosibirsk City

Severe Acute Respiratory Infection (SARI) sentinel surveillance in the country of Georgia

RSV Global Epidemiology Network, RESCEU investigators. Global patterns in monthly activity of influenza virus, respiratory syncytial virus, parainfluenza virus, and metapneumovirus: a systematic analysis

Early occurrence of influenza A epidemics coincided with changes in occurrence of other respiratory virus infections. Influenza Other Respir Viruses

Using routine testing data to understand circulation patterns of influenza A, respiratory syncytial virus and other respiratory viruses in Victoria

Interference between respiratory syncytial virus and human rhinovirus infection in Infancy

Increased risk of noninfluenza respiratory virus infections associated with receipt of inactivated influenza vaccine

Influenza vaccine does not increase the risk of coronavirus or other noninfluenza respiratory viruses: retrospective analysis from Canada

Epidemiology of seasonal coronaviruses: Establishing the context for the emergence of coronavirus disease 2019

Clinical characteristics, predictors, and performance of case definition-Interim results from the WHO global respiratory syncytial virus surveillance pilot. Influenza Other Respir Viruses

Preparing for the future implementation of respiratory syncytial virus vaccines

Community-acquired pneumonia requiring hospitalization among U.S. adults

Community-acquired pneumonia requiring hospitalization among U.S. children

Respiratory syncytial virus and other noninfluenza respiratory viruses in older adults

Viral infections of the lower respiratory tract: old viruses, new viruses, and the role of diagnosis

Invasive aspergillosis in patients admitted to the intensive care unit with severe influenza: a retrospective cohort study

The impact of bacterial and viral co-infection in severe influenza. Influenza Other Respir Viruses

Communityacquired bacterial co-infection predicts severity and mortality in influenza-associated pneumonia admitted patients

Aetiology of lower respiratory tract infection in adults in primary care: a prospective study in 11 European countries

The diagnosis of viral respiratory disease in older adults

Influence of age on symptoms and laboratory findings at presentation in patients with influenzaassociated pneumonia

License, supplementary material and copyright

The authors would like to thank Jeannine Weyckmans, Ilham Fdillate, Assia Hamouda and Reinout Van Eycken from the NIC for their excellent technical assistance throughout the surveillance seasons; Evelyn Petit from AZ Sint-Jan, Françoise Antoine, Tiphaine Mouquet and Christelle Nguepi from CHU St-Pierre, Lore-Lien Roels and Prof. Elke De Wachter from UZ Brussels, Marlies Blommen, Annemie Forier and Dr. Luc Waumans from Jessa Ziekenhuis, Catherine Sion and Mohamed El Kaissi from Grand Hôpital de Charleroi for their dedication in the follow-up of the SARI cases; and all the doctors and nurses of the six hospitals who took part in the recruitment of the SARI cases, without whom the SARI surveillance would not be possible. N.D. is a Post-doctorate Clinical Master Specialist of the FRS-FNRS (Fond National de la Recherche Scientifique).

Funding: This work was supported by the Belgian Federal Public Service 'Health, Food Chain Safety, and Environment', the Belgian National Insurance Health Care (INAMI/RIZIV), and the Regional Health Authorities of Flanders (Agentschap Zorg en Gezondheid, AZG), of Brussels (Commission communautaire commune de Bruxelles-Capitale, COCOM), and of Wallonia (Agence pour une Vie de Qualité, AVIQ).

None declared.