The changing landscape of respiratory viruses contributing to respiratory hospitalisations: results from a hospital-based surveillance in Quebec, Canada, 2012-13 to 2021-22

BackgroundA comprehensive description of the combined effect of SARS-CoV-2 and respiratory viruses (RV) other than SARS-CoV-2 (ORV) on hospitalisations is lacking.

AimTo compare viral etiology of acute respiratory infections (ARI) hospitalisations before and during two pandemic years from a surveillance network in Quebec, Canada.

MethodWe compared detection of ORV and SARS-CoV-2 during 2020-21 and 2021-22 to 8 pre-pandemic influenza seasons in patients hospitalised with ARI who were tested systematically by a multiplex PCR.

ResultsDuring pre-pandemic influenza seasons, overall RV detection was 92.7% (1,493) (48.3% respiratory syncytial virus (RSV)) in children and 62.8% (4,339) (40.1% influenza) in adults. Overall RV detection in 2020-21 was 58.6% (29) in children (all ORV) and 43.7% (333) in adults (3.4% ORV, 40.3% SARS-CoV2, both including coinfections). In 2021-22 overall RV detection was 91.0% (201) in children (82.8% ORV, 8.1% SARS-CoV-2, both including coinfections) and 55.5% (527) in adults (14.1% ORV, 41.4% SARS-CoV-2, both including coinfections).

Virtually no influenza was detected in 2020-21 and in 2021-22 up to epi-week 2022-9 presented here; no RSV was detected in 2020-21. In 2021-22, detection of RSV was comparable to pre-pandemic years but with an unusually early season. There were significant differences in ORV and SARS-CoV-2 detection between time periods and age groups.

ConclusionSignificant continuous shifts in age distribution and viral etiology of ARI hospitalisations occurred during two pandemic years. This reflects evolving RV epidemiology and underscores the need for increased scrutiny of ARI hospitalisation etiology to inform tailored public health recommendations.


Introduction
Stringent mitigation efforts such as border closures and travel restrictions, lockdowns, social distancing, use of masks in public spaces, school and business closures, and teleworking have been implemented worldwide to reduce the transmission of severe acute respiratory syndrome virus 2 (SARS-CoV-2) and its impact on hospital bed capacity. While first weeks of 2020 in the Northern hemisphere were dominated by respiratory viruses other than SARS-CoV-2 (ORV), SARS-CoV-2 almost completely replaced seasonally circulating ORV within several weeks (1-3), with a subsequent alteration of traditional seasonality of some of them and virtual disappearance of others during extended periods of time in different parts of the world (4)(5)(6)(7)(8)(9). One of the collateral consequences observed early during the pandemic was the decrease in pediatric visits and hospitalizations overall and especially of those associated with acute respiratory infections (ARI) (10,11), bronchiolitis (12,13), as well as pediatric asthma exacerbations associated with ARI (14).
The pandemic had an impact on resources with decreased volumes of performed tests and changes in the propensity to test for ORV, as well as on the health-seeking behavior, which may complicate the interpretation of ORV surveillance. Detection of ORV in hospitalised patients may be more informative since admission requires a certain degree of severity and propensity to be tested for a larger panel of respiratory viruses is higher. However, because of high demand of SARS-CoV-2 tests on hospital laboratories, the testing for ORV has been reduced even in hospitalised patients during the pandemic. A number of reports described detection of ORV in patients hospitalised with ARI or with COVID-19 during the pandemic (8,(15)(16)(17)(18)(19). However, to our knowledge none described results of systematic detection of both ORV and SARS-CoV-2 (not only at physician request) using a panel of multiple RV in a multicenter network including both pandemic years and comparing them with as long as 8 pre-pandemic years. The characterisation of combined impact of both SARS-CoV-2 and ORV on hospital capacity during the two pandemic years and its comparison with pre-pandemic seasons may provide insightful information on postpandemic period when the SARS-CoV-2 will cocirculate along with ORV.
In Quebec, Canada, a prospective hospital-based surveillance network with systematic testing for a panel of 17 respiratory viruses in pediatric and adult patients admitted for ARI has been in place since 2012-2013 during periods with high influenza circulation (20)(21)(22)(23). The same network was used for the surveillance during the pandemic, by adding the SARS-CoV-2 to the panel. We report . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 2, 2022. ; https://doi.org/10.1101/2022.07.01.22277061 doi: medRxiv preprint here the results for two pandemic winter seasons (2020-21 and 2021-22) and compare them to eight previous seasons.

Study Population
The study design during the pre-pandemic years has been described in detail elsewhere (20)(21)(22)(23).
In brief, four regional hospitals (2 community, 2 academic/tertiary; all of them serving both children and adults) with a catchment area of nearly 10% of the Quebec population (approximately 8.6 million in 2021) participated in the surveillance during eight influenza seasons since 2012-13. In 2020-21, one of the 4 hospitals was not able to participate (≈15% of the population included in previous years) because of the challenges with hospital resources during the pandemic; this hospital rejoined the network in 2021-22. Two additional tertiary hospitals, one adult and one pediatric, joined the network in 2021-22, for a total of 6 hospitals (≈15% of the Quebec population). Results from these two hospitals are included only in the description of virus detection per week and are not used for the comparison with pre-pandemic period. All patients presenting to their emergency department with ARI were systematically swabbed during high influenza activity weeks of pre-pandemic years or during periods with increasing hospitalisations due to SARS-CoV-2 during the pandemic. Eligible patients were those admitted for ≥24 hours with a standardized definition (fever/feverishness not attributed to other illness or cough or sore throat), expanded in 2020-21 to include symptoms specific for COVID-19 (fever/history of fever not attributed to other illness, or cough (or exacerbation of cough)/difficulty breathing (or exacerbation of difficulty breathing), or sudden extreme fatigue, or at least two of the following symptoms: rhinorrhea/nasal congestion, sore throat, myalgia/arthralgia, or sudden anosmia/ageusia). Nurses collected demographic and clinical details from the patient or legal representative on a standardized questionnaire and reviewed patients' charts at discharge for additional clinical information.

Surveillance period
For the pre-pandemic years, the surveillance period started when the positivity rate for influenza in respiratory specimens from the provincial sentinel laboratory surveillance was ≥15% for two consecutive weeks and stopped the week after this rate dropped below 15% or when the planned . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 2, 2022. ; https://doi.org/10.1101/2022.07.01.22277061 doi: medRxiv preprint sample size for the season was achieved (800-1000 specimens depending on the season). The provincial laboratory surveillance included >40 laboratories across the province of Quebec with >100,000 respiratory specimens per year. Surveillance lasted from 7 to 12 weeks per season (median of 8,5 weeks), between epi-weeks 49 (earliest) and 14 (latest) (Supplementary Figure 1).
In 2020-21, the surveillance started on September 27, 2020 (epi-week 40), during the ascending phase of the 2 nd COVID-19 wave in Québec and was halted on May 29, 2021 (epi-week 21) (overall duration 35 weeks) (Figure 1, Supplementary Figure 1 Because of challenges with hospital resources during periods of high SARS-CoV-2 circulation, the surveillance was paused during some weeks and sampling of enrolment during pre-determined days of week was adopted by some hospitals (Supplementary Figure 1).

Laboratory Analysis
Nasal specimens collected on flocked swabs from eligible patients were sent to the provincial public health laboratory (Laboratoire de Santé Publique du Québec, LSPQ) and tested using the is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 2, 2022. ; https://doi.org/10.1101/2022.07.01.22277061 doi: medRxiv preprint 3, adenovirus, rhinovirus, enterovirus, hCoV 229E and OC43, RSV, and hMPV was used by the adult center (25). In house PCR using LightMix® Modular Assays according to the manufacturer's recommendations (26) was used by the pediatric center to detect influenza A and B, RSV, hCoV (not differentiated), hMPV, adenovirus, hPIV (not differentiated), enteroviruses/rhinoviruses (not differentiated), and SARS-CoV-2.

Statistical Analysis
Proportions were compared by using chi-square or Fisher exact test when appropriate. Statistical significance was set at p<0.05. Statistical analyses were conducted using SAS, version 9.4 (SAS Institute, Cary, NC). Similar hospitalisation rate and viral etiology distribution was assumed for days with and without enrolment during weeks with only three enrolment days.

Ethics
Institutional Review Board approval was obtained from all participating hospitals (Hôpital régional de Rimouski, Hôpital de Chicoutimi, Hôpital de la Cité-de-la-Santé -Laval, Centre hospitalier universitaire régional de Trois-Rivières) for the first 3 years and a signed informed consent was used. A waiver was obtained for the following years when the project was conducted as a sentinel surveillance mandated by the Ministry of Health from the Research Ethics Board of the Centre hospitalier universitaire de Québec-Université Laval.  Table 1). In 2020-21, significantly less children were hospitalised with ARI (4% of all hospitalised patients) compared to previous seasons (26%, p<0,0001). During both pre-pandemic and pandemic winter seasons, very few young adults were hospitalised with ARI, with approximately 1% of 18-29-year-olds and 2% of 30-39-year-olds and a subsequent gradual increase with age ( Figure 2). Compared to the pre-pandemic period, during . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
In 2021-22, more patients (728, 62.2%) were positive for at least one virus, but the detection rate was still significantly lower than during the pre-pandemic ( Table 1) (Table 2A and 2B). Hospitalisation due to RSV was detected much earlier (>10% during epi-weeks 40 to 48, with a maximum of 33% during epi-week 41) compared to pre-pandemic seasons (>10% during epi-weeks 52 to 10) (Figure 1).
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Adult hospitalisations
During the pre-pandemic seasons, at least one respiratory virus was detected in 63% (2,723) adults (6.1% coinfections); influenza was by far the most predominant virus (40.1%).
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 2, 2022. ; https://doi.org/10.1101/2022.07.01.22277061 doi: medRxiv preprint Similar to the 1 st pandemic season, the SARS-CoV-2 was the most frequent virus in adults (41.4%, 393). ORV with or without SARS-CoV-2 were detected in all examined adult age groups except the 40-49-year-olds, for an overall proportion of 16.5% (157) ( Table 1). Compared to the 1 st pandemic season, the SARS-CoV-2 was detected more often in adults younger than 50 years (the only significant difference in 18-29-year-olds, Table 1); detection was similar in adults 50 years and older. Compared to pre-pandemic period, significantly more adenoviruses, and less hMPV, hPIV1-4, and hCoV were detected ( Table 2B).  Table 1).

Results
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 2, 2022. ; https://doi.org/10.1101/2022.07.01.22277061 doi: medRxiv preprint Another potential contributive factor was the SARS-CoV-2 variant evolution. For example, ancestral and alpha SARS-CoV-2 variants circulated during the 2020-21 season and affected mostly adults; while delta, and then omicron variants, both affecting more importantly pediatric population, predominated during the autumn and the winter of 2021-22 (24). Also, an unhabitual surge of RSV in August-September 2021(27) resulted in higher proportion of children hospitalised with both ORV (predominantly RSV) and COVID-19 compared to the first pandemic season. Finally, COVID-19 vaccines and outpatient antiviral availability and uptake, effective to prevent hospitalisations, varied by age group. During the second pandemic year, older patients were prioritized in COVID-19 vaccination campaigns and more patients with comorbidities were offered outpatient antiviral treatment. For children, however, COVID-19 vaccines were available later and in young adults vaccine uptake was lower than in older population (28).
Overall during the pre-pandemic influenza seasons, the two most frequently detected viruses among patients hospitalised with ARI were influenza and RSV. The SARS-CoV-2 was the most important respiratory virus during the first pandemic year; in 2021-22 its contribution decreased while contribution of ORV increased. However, there were differences in pediatric and adult population. RV (mostly RSV) affected more children than adults during pre-pandemic winter seasons, and although their overall impact was lower during pandemic period, children remained mostly affected by ORV and not by the SARS-CoV-2 during both pandemic years. In adults, influenza was the most important virus during the pre-pandemic years, while SARS-CoV-2 was by far more important than ORV during the two pandemic years.
The important reduction of pediatric respiratory hospitalisations, virtual absence of influenza and RSV during first pandemic year and increased role of COVID-19 and ORV, especially RSV, in hospitalised children during second pandemic year in our study are in line with reports from some other countries and other provinces of Canada (13,16,17,29,30). Of note, the timing of the unusual surge of RSV varied by region of the world: in Europe, Israel and USA increases started to be reported in the Spring-Summer of 2021 (13,16,17,29). Major shifts in the epidemiology of RSV with large-scale outbreaks were reported in New Zealand (corresponding to usual season) (31) and Australia (out-of-season) (32). In Canada, the first province to report a surge of RSV was Quebec in August 2021, while the other provinces followed with a delay of 2-3 months(30).
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 2, 2022. This study has limitations. First, the pre-pandemic surveillance occurred during the peaks of influenza activity and therefore relative contribution of other RV may be underestimated as compared to entire winter seasons. Surveillance periods during pandemic years were mostly tailored to the increase in respiratory hospitalisations following increased SARS-CoV-2 circulation and were much longer. However, we think the comparison is still valid because it includes periods with most strain on hospital capacity due to intensive circulation of either ORV, SARS-CoV-2, or both. Second, ARI definition used during pandemic years was broader than during pre-pandemic years and may have contributed to lower detection of RV. Third, the first pandemic year was limited to only 3 hospitals and periodic pauses and sampling of the enrolment during the 2 pandemic years were necessary given the stretch of resources which limited the sample size. In conclusion, important shifts in age distribution and viral etiology of ARI hospitalisations in children and adults were observed during the two first pandemic years. While the first pandemic year was significantly different from the pre-pandemic winter seasons, the second year was more comparable, both in terms of age distribution and RV contribution. The complex interplay between mitigation measures, intrinsic seasonality and secular trends of ORV, changes in circulating SARS-CoV-2 variants and their severity, COVID-19 vaccine uptake and effectiveness, outpatient antiviral treatments and potential viral interference (33,34) are all factors that may contribute to variations of ORV and SARS-CoV-2 role in hospital morbidity. Even if we do not report here the results for the entire 2021-22 season, our study underscores importance of surveillance in understanding altered seasonal patterns of RV and shows that the role of SARS-. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 2, 2022. ; https://doi.org/10.1101/2022.07.01.22277061 doi: medRxiv preprint 12 CoV-2 relative to ORV is continually changing. Current situation may reflect a transition period until it will find its ecological niche in human population. With the continued easing of mitigation measures, an increasing role of usual winter viruses is expected. Although new SARS-CoV-2 variants may emerge and cause occasional increases in hospitalisations, in the long run it may establish itself as another respiratory virus. At this point, it is difficult to foresee its role compared to ORV. Increased scrutiny to continuing changes in the etiology of ARI hospitalisations is needed in order to inform mathematical modelling and appropriate public health recommendations.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Acknowledgments:
We acknowledge France Bouchard who coordinated the surveillance and Sophie Auger who entered and cleaned collected data throughout all years. We are extremely grateful to the frontline surveillance staff from the participating hospitals who collected and provided data in . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 2, 2022. ; https://doi.org/10.1101/2022.07.01.22277061 doi: medRxiv preprint Table 1 Number and proportion of patients hospitalised for acute respiratory infections by age group and detected respiratory virus    . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 2, 2022.  . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 2, 2022. ; https://doi.org/10.1101/2022.07.01.22277061 doi: medRxiv preprint  CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 2, 2022. ; https://doi.org/10.1101/2022.07.01.22277061 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 2, 2022. ; https://doi.org/10.1101/2022.07.01.22277061 doi: medRxiv preprint