The study was conducted at the primary health care centre “First Vienna Pediatric Medical Center.” The recruitment phase took place from November 2022 to March 2023, the typical RSV infection season in Austria before Covid-19 (13). Children without chronic infections, younger than the complete 36 months of age (≤3rd birthday), who presented with at least one of doctor's diagnosed acute respiratory infection symptoms (cough, rhinitis, nasal congestion, sore throat/pharyngitis, fever, otalgia) were eligible for this study. Fever was predefined as body temperature of at least 38 °C measured with an ear thermometer. In order to prevent a selection bias, swabs were taken on different days of the week during the normal practice opening hours (in total 66 days over the recruitment period) because due to limited resources not all of the eligible patients could be included. It was the aim to include as many patients as possible and at least 100 RSV positive children.

The following exclusion criteria were based on the RSV high-risk profile according to the RSV prevention guideline (14): preterm birth (<37 weeks of gestation), severe congenital pulmonary diseases (e.g., cystic fibrosis, primary ciliary dyskinesia, interstitial lung diseases), neuromuscular diseases with impaired lung clearance, severe hemodynamic cardiac diseases, or severe immunodeficiency (e.g., immunoglobulin deficiency); and in addition, older children (age >36 + 0 months).

The study design remains as described in the publications of the preceding seasons of 2020/21 and 2021/22 (4, 5), except for the following changes: the multiplex PCR was done at the Medical University of Vienna, Center of Virology within the RSV surveillance network project (13). Additionally, RSV subtypes and genotypes were characterised (15).

Telemedical follow-up was conducted one week after the first medical appointment to discuss the results and changes in the clinical status.

The patient's general clinical data (height, weight, chronic diseases), doctor's diagnosed acute respiratory symptoms (rhinitis, nasal congestion, cough, fever, pharyngitis), doctor's diagnosis (otitis media, bronchitis, bronchiolitis, pneumonia, upper airway infection (rhinitis, nasal congestion, throat infection), others (fever without further diagnosis, not further specified diagnosis), and general state of health [respiratory rate, oxygenation (SpO2), wheezing/crackles, accessory respiratory muscle utilisation] were documented in a case report form. Furthermore, prescribed symptomatic treatment (hypertonic saline spray, decongestant nasal spray, analgesics, antibiotics, corticosteroid treatment (inhaled or systemic), inhalation therapy of short-acting betamimetics (SABA) or 0.9% sodium chloride) were documented as well.

Parents signed the declaration of consent after being informed about the research aims and the study procedure.

The following data were gathered during the telemedical follow-up by the study team: respiratory symptoms and treatment, parent's subjective evaluation of the child's health (improved, deteriorated, or unchanged), and up-date on hospitalisation due to the respiratory infection within the last week.

Furthermore, information regarding the family's socioeconomic status (living situation, marital status, parent´s education, number of siblings) was determined by interviewing the corresponding parent either at the day of study inclusion or during the telemedical follow up control.

This study was approved by the Ethics Committee of the Medical University of Vienna (EK-No. 1864/2020). Due to the national regulations it was not possible to collect children's data without consent (e.g., total number of eligible children), as this violates the General Data Protection regulation.

During the doctor's appointment, a trained study team collected anterior nasal swabs. The samples were stored in 0.9% sodium chloride solution until further analysis via multiplex PCR. The following viral pathogens were tested: RSV including RSV-subtype and genotype, rhinovirus, adenovirus, metapneumovirus, enterovirus, SARS-CoV-2, human coronaviruses (229E, HKU1, OC43, NL63), influenza A (H1 2009, H3), influenza B (Yamagata, Victoria), influenza C, and parainfluenza virus 1–3.

The PCR evaluation and the RSV subgroup analysis were performed at the Center for Virology, Medical University of Vienna, and incorporated into the statistical report of the RSV Surveillance Network (13).

In case of recurrent acute respiratory infection in one patient, nasal swabs were repeated as long as there was a symptom-free interval of at least one week between visits.

To assess the severity of respiratory disease the Modified Tal Score (MTS) was used. This clinical ordinal scoring system has been used in multiple publications and is validated to categorise bronchiolitis especially among RSV infections and the severity of respiratory disease (16–18). Points were assigned depending on accessory muscle use (none (0 points)/mild/moderate/severe (3 points)), respiratory rate (<6 months of age: 0 points: <40, 1: 41–55, 2: 56–70; 3 points:>70/min; >6 months of age: 0: <30, 1: 31–45, 2: 46–60, 3 points > 60/min), degree of wheezing or crackles (none: 0 points/end expiratory with stethoscope/inspiration and expiration with stethoscope/3 points: audible without stethoscope), and SpO2 values (0 points: SpO2 > 95%, 1: 92%–94%, 2: 90%–91%, 3: <89%). Therefore, 0 points indicated no bronchiolitis, 1–5 points suggested a mild infection, 6–10 points a moderate one, and 11–12 points severe disease.

The principal aim of the analyses was to describe and characterize RSV infections over the winter season and the severity and risk for hospitalization in comparison to non-RSV and negative tested patients. In addition, we analysed the clinical course, the symptoms, prescribed therapy and the doctor's diagnosis.

Raw data was sorted in Google Sheets and Microsoft Excel and analzyed using IBM SPSS-Statistics version 29 and GraphPad Prism version 10. A two-sided p value of p < 0.05 indicated statistical significance. The Bonferroni test corrected the risk of higher frequency errors due to multiple testing, resulting in adjusted p-values (p-adj.).

Most data were categorised as nominal variables. Metric variables (age, weight, height, BMI, number of siblings) were used to describe the study population further.

Due to the low number of specific pathogens, the following groups were built: single infections with RSV, adenovirus, rhinovirus, enterovirus, metapneumovirus, human coronaviruses, influenza and parainfluenza. Co-infections constitute the category “co-infection”, and results without a detected viral pathogen the group “negative”.

Qualitative variables such as patient characteristics, manifestation of symptoms, recommended therapy, doctor's diagnosis and pathogen groups, as well as the comparison of these variables, were presented as descriptive statistics through absolute and relative frequency (percentages). As for metric variables, mean, median, interquartile range, and first and third quartiles were calculated.

The statistical analysis consists mainly of descriptive analysis. The sociodemographic data were reported for the entire study population using descriptive statistics.

Nominally and ordinally scaled variables such as pathogen groups/RSV subtypes, symptoms, treatments, diagnosis, and MTS were tested using the Chi-Square test or Fisher's Exact test. Associations of pathogen groups and symptoms, as well as pathogen group and treatment were analysed in contingency tables using the effect size Cramér V, which was interpreted using the Rea and Parker's classification (r = 0.10–0.20 weak, r = 0.20–0.40 moderate, and r = 0.40–0.60 relatively strong association) (19).

For detailed RSV characterization, we generated four groups: RSV A, RSV B (single infections only), non-RSV (single infections), and negative (negative PCR results). Co-infections were not included in this analysis. Sociodemographic data were reported for these groups. Associations of these groups and symptoms, treatment, severity (no bronchiolitis, mild-moderate-severe according to MTS) and state of health (improved, deteriorated, or unchanged) at the telemedical follow-up control were analysed in contingency tables using Cramér V. All p values were corrected using the Bonferroni method for multiple testing.

A binominal logistic regression was performed to ascertain the effects of RSV detection (yes/no, independent on the co-infection status), age (in months), gender (male/female), and MTS score (0–12) (all independent variables) on the likelihood of hospitalization [(yes/no); dependent variable]. The omnibus test of the model coefficients indicated the significance of the model and the Nagelkerke R2 was used to interpret the amount of explained variance. Wald test was used to test for significance and odds ratio (OR), 95% confidence interval (95% CI) and p-values were calculated.

Swabs were performed on 66 days throughout the study period. On these days, a total of 2,200 acutely ill children aged 0–36 months presented to the practice (Figure 1). After applying the inclusion and exclusion criteria, 329 children with acute respiratory infections participated in the study from November 2022 to March 2023. The median age was 14 months (Q1–Q3: 6.5–23.5 months); 47.4% of patients were male and 52.6% female. In total, 345 nose swabs were collected. During the study period, 314 children had one swab, 14 children had two swabs, and one child had three nose swabs. All sociodemographic and general clinical characteristics are summarised in Table 1.

The most common clinical symptom was cough (84.6%), followed by rhinitis (73.0%) and fever (53.3%) (Table 2). Patients were treated with analgesics (70.4%), decongestant and hypertonic saline nose sprays/drops (62.0%), short-acting betamimetics (13.9%), and antibiotics (13.3%). Of the 345 included visits, 291 cases received one clinical diagnosis (84.3%), 48 received two (13.9%) and 6 got three (1.7%). Most participants (279 of 345 cases, 80.9%) were clinically diagnosed by their attending physician with an upper airway infection, independent of the PCR result, followed by bronchitis (47/345, 13.6%) and otitis media (42/345, 12.2%) (details see Table 2).

Co-infections were found in 43 samples (12.5%) (41 of these with two pathogens and two with three pathogens), while no viral pathogen was detected in 57 cases (16.5%). The most common virus was RSV (110/345, 31.9%), with 89 single infections (25.8%), and 21 cases of co-infections including RSV (6.1%) (Figure 2A). The highest incidence of positive results and RSV cases was registered in weeks 46–52 (Figure 2B), followed by a decline with sporadic RSV diagnoses. In addition to RSV, influenza (17.5%) and rhinovirus infections (20.58%) were very common. The influenza vaccination rate was 19.5%.

Comparing the frequency of symptoms and all tested pathogen groups (including multiple viral infections and negative PCR results as separate category) revealed a significant and moderate association (Cramér's V: 0.346, p < 0.001, p-adj. <0.01) (Supplementary Table 1). The highest frequency of cough and rhinitis was found among RSV cases (96.6% and 76.4%), human coronaviruses (100% and 90.9%), parainfluenza (90.9% and 81.8%), and rhinovirus cases (85.7% and 81.6%). Fever was most common among individuals with adenovirus (90.9%), enterovirus (100%), and influenza infections (85.2%), whereas pharyngitis was mostly found in children with enterovirus (66.7%) and parainfluenza 1–3 infections (54.5%).

In contrast, no significant association emerged between prescribed treatment and pathogen group (Cramér V: 0.298, p = 0.396, p-adj. >0.999).

Out of the 110 positive RSV results (including co-infections), only 13 cases were subtype A (13/110, 11.8%), 96 were subtype B (96/110, 87.3%), and one RSV subtype could not be determined (<1%). Eight of the 13 RSV subtype A cases were of the ON1 genotype (8/110, 7.3%), 74 of the RSV subtype B cases were of the BA9 genotype (74/110, 67.3%), and in 28 cases, the genotype could not be identified (28/110, 25.5%). No other genotypes were detected.

Comparing occurrence of RSV cases and distribution of subtypes (Figure 3), the peak in RSV cases, specifically among subtype B patients, was recorded during the calendar weeks 46 to 2 (especially week 48), whereas subtype A cases were steadily low with 0–3 positive results per week throughout the season.

Four pathogen groups were established to examine correlations between participants’ symptoms, prescribed treatment, doctor's diagnosis, MTS and the children's state of health one week later at the follow-up appointment: RSV A, RSV B, non-RSV (infections without RSV), and negative PCR results. Only cases of single infections (absence of co-infection) were considered among those four groups. No significant differences of sociodemographic data were observed between these groups (Table 1).

Symptoms of patients with RSV subtype A and B were very similar, except for fever which was more common in the RSV A group (75.0% vs. 45.7% in RSV B). The participants' recorded symptoms revealed a moderate but not significant correlation with the corresponding groups. (Cramér V: 0.322, p = 0.068, p-adj. 0.612). Conversely, a moderate and significant correlation can be observed between participants’ treatment plans and the four groups (Cramér V: 0.380, p < 0.001, p-adj. <0.01) with a high proportion of RSV B patients being prescribed SABA inhalation (Supplementary Table 1). RSV A patients mainly received nose drops or analgesics. No significant differences were observed with regard to doctor's diagnosis between the four groups (Cramér V: 0.260, p = 0.235, p-adj. >0.999). Twenty-five percent of RSV single infection cases received the diagnosis of bronchitis (RSV A: 2 out of 8, RSV B: 21 out of 81); and 54.5% of all diagnosed bronchiolitis cases were infected with an RSV B single infection (6 out of 11 cases) (Table 2).

Comparing the MTS as a tool to ascertain the severity of bronchiolitis with the cases of RSV A, RSV B, non-RSV infections, and negative tests reveals a moderate correlation (Cramér V: 0.316, p < 0.001, p-adj. p < 0.01).

Out of the 89 RSV single infections, most participants infected with either subtype A or B showed no signs of bronchiolitis (Figure 4) or mild bronchiolitis, according to the MTS. One RSV A and 5 RSV B patients suffered from moderate bronchiolitis. There were no cases of severe bronchiolitis among our study subjects. Almost all children with a negative PCR test result or an infection without RSV, showed no sign of bronchiolitis (Table 2).

A logistic regression was performed to ascertain the effects of age, gender, MTS, and RSV on the likelihood that children are hospitalised. The logistic regression model was statistically significant [X2(4) = 27.68, p < 0.001]. The model explained 28% (Nagelkerke R2) of the variance in hospitalisation rate and correctly classified 96.1% of cases. RSV positive patients were 7.5 times (95% CI: 1.46–38.40) more likely to be hospitalised than patients without RSV. The odds of hospitalization was significantly lower with increasing age. A higher MTS score was trendwise (p = 0.08) associated with a higher hospitalization risk which, however was not significant (Table 3).

In nine cases the final data of hospitalization are missing because the telemedical follow-up was not possible. In general, thirteen hospitalisations occurred because of a respiratory infection during the observed interval (13/336, 3.9%). Out of these 13, in nine cases hospitalisation was necessary due to RSV single infections: one case of RSV subtype A and eight cases of RSV B. No participant infected with another single pathogen was admitted to the hospital. In two cases the hospitalised child was infected by multiple pathogens, in both cases including RSV (in combination with adenovirus and coronavirus OC43). In the last two cases of hospitalisation the pathogen was not detectable (PCR negative).

The one-week follow-up confirmed that the severity of symptoms had improved in almost all 110 RSV positive cases (including co-infections), non-RSV infections, and negative cases (100 out of the 110 RSV cases) with no significant difference within the RSV subtypes, non-RSV, and negative cases. (Cramérs V = 0.119, p = 0.214 p-adj. >0.999).

Although the “First Vienna Pediatric Medical Center” serves the highest annual number of patients of every socioeconomic status in Vienna, the monocentric design of this study is a limiting factor and could have potentially influenced its outcome. A national record of all RSV-positive cases from primary healthcare settings and hospitals would lead to more comprehensive data collection. The opportunity to participate in this study was generally well received; nevertheless, we lack the number of potential participants who decided not to be involved. Due to data protection regulations we are not able to specify the numbers of excluded patients.

As for the socioeconomic status of our population, participants with parents with university degrees are overrepresented compared to Vienna's general public. This factor could have potentially led to differing social circumstances in the study population regarding income, housing situation or number of siblings.

Anterior nasal swabs were preferred to the more invasive nasopharyngeal swabs as they reportedly provide valid results (51); they can be however, potentially inaccurate.

The Modified Tal Score, used to assess the severity of bronchiolitis, is validated and has been used in multiple publications especially for RSV bronchiolitis. However, it was shown that the MTS is not only applicable to RSV bronchiolitis, but also works validly for other respiratory disorders, although the data is limited.