Seasonality and Co-Detection of Respiratory Viral Infections Among Hospitalised Patients Admitted With Acute Respiratory Illness-Valencia Region, Spain, 2010-2021

Abstract

Background: Respiratory viruses are known to represent a high burden in winter, yet the seasonality of many viruses remains poorly understood. Better knowledge of co-circulation and interaction between viruses is critical to prevention and management. We use > 10-year active surveillance in the Valencia Region to assess seasonality and co-circulation.

Methods: Over 2010-2021, samples from patients hospitalised for acute respiratory illness were analysed using multiplex real-time PCR to test for 9 viruses: influenza, respiratory syncytial virus (RSV), parainfluenza virus (PIV), rhino/enteroviruses (HRV/ENV), metapneumovirus (MPV), bocavirus, adenovirus, SARS-CoV-2 and non-SARS coronaviruses (HCoV). Winter seasonal patterns of incidence were examined. Instances of co-detection of multiple viruses in a sample were analysed and compared with expected values under a crude model of independent circulation.

Results: Most viruses exhibited consistent patterns between years. Specifically, RSV and influenza seasons were clearly defined, peaking in December-February, as did HCoV and SARS-CoV-2. MPV, PIV and HRV/ENV showed less clear seasonality, with circulation outside the observed period. All viruses circulated in January, suggesting any pair had opportunity for co-infection. Multiple viruses were found in 4% of patients, with more common co-detection in children under 5 (9%) than older ages. Influenza co-detection was generally observed infrequently relative to expectation, while RSV co-detections were more common, particularly among young children.

Conclusions: We identify characteristic patterns of viruses associated with acute respiratory hospitalisation during winter. Simultaneous circulation permits extensive co-detection of viruses, particularly in young children. However, virus combinations appear to differ in their rates of co-detection, meriting further study.

Keywords: RSV respiratory syncytial virus; co‐detection; co‐infection; influenza; respiratory virus; rhinovirus; seasonality; severe acute respiratory syndrome coronavirus 2.

FIGURE 1

Observed seasonal curves for each virus in each season, per capita in the total catchment population. Each virus is shown in a different colour. The top panel shows the viruses with maximum incidence greater than 2 per 100,000 per week and the bottom those with lower maximum incidence. The greyed out areas are those in which data were not collected, and the text underneath the graph shows the dates of each data collection season. Each virus curve is displayed separately on a separate panel in Figure S1.

FIGURE 2

(A) Seasonality trends for each virus, with seasons overlaid. Each row is a different virus, with colours representing seasons, and the black line showing the weekly mean incidence per virus. (B) The number of mono‐detections and co‐detections by age group and all ages combined. Each panel represents an age category, with the panel label indicating both the age and number of patients with valid tests. Size of points indicates average number of detections per active season (SARS‐CoV‐2 was considered active for one season, all other viruses for 10 seasons in ages < 5 and 11 seasons otherwise) on a log scale. The diagonal represents the number of mono‐detections, while the off‐diagonal represent co‐detections, with colour representing the proportion of detections of Virus1 that are co‐detections (capped at 30%). The matrix is symmetrical for point size but not colour.

FIGURE 3

The point estimates and bootstrapped credible intervals of observed versus expected number of co‐detections for each pair of viruses by age group and all ages combined. The x‐axis represents the observed/expected ratio, with the point indicating the point estimate and the error bars the 2.5% and 97.5% percentile of the bootstrapped sample. The colour of the point and the error bar represent the observed expected ratio in the point estimate. The size of the point indicates the number of expected co‐detections, as a measure of the extent of co‐circulation. The y‐axis and the panel columns represent each pair of viruses, while each row of panels represents an age group.