FormalPara Key Summary Points

Why carry out the study?

Evidence on the burden of disease of respiratory syncytial virus (RSV) in Germany is based on few, mostly single centre studies, covering only a limited number of seasons or focusing on specific settings.

The current study provides a comprehensive analysis of public databases covering outpatient (between seasons 2014/2015 to 2022/2023), inpatient (years 2000–2022 and seasons 2019/2020 to 2022/2023) and mortality data (years 1998–2022).

What was learned from the study?

We found that RSV-related outpatient visits vary from 1,313,100 to 3,911,800 between the 2014/2015 and 2022/2023 seasons, and the number of RSV-related inpatient visits increase continuously from below 1000 in the year 2000 to up to 52,000 hospitalisations in the 2022/2023 season, with up to 1400 in-hospital deaths.

The inpatient data reveals significant underreporting that becomes apparent through the increased testing during the COVID-19 pandemic.

The underreporting is more pronounced in the older age groups and the age group 60+ comprises 13.1% of RSV-related inpatient visits, 24.6% of RSV-related inpatient stays and 93% of RSV-related in-hospital deaths.

Introduction

Respiratory syncytial virus (RSV) predominantly infects the lungs and respiratory tract, causing numerous cases of acute respiratory infections (ARI) across all age groups worldwide. It poses a significant public health challenge due to its high morbidity and mortality rates [1, 2], particularly among infants and young children, adults aged 60 years and over (60+) and individuals of all ages with pre-existing health conditions. While monoclonal antibodies are approved for use in infants, the German National Immunisation Technical Advisory Group (NITAG) currently does not recommend any vaccination for the elderly to mitigate the impact of RSV on this vulnerable group.

German medical societies acknowledge the burden of disease (BoD) of RSV in the 60+ age group and generally recommend vaccination against RSV [3]. A local immunisation recommendation in Saxony highlighted the need for maternal immunization and advised shared clinical decision-making for those aged 60+, citing uncertainties about the BoD of RSV [4]. To facilitate effective public health decision-making and vaccination recommendations at a national level, comprehensive data on the BoD of RSV in Germany is essential.

A systematic literature review, including data up to November 2021 on the burden of RSV in the 60+ age group, revealed limited evidence for Germany [5]. Since then, several studies have been published, but these either cover short and post-pandemic periods [6,7,8] or were conducted as single-centre studies [1, 9]. In addition to published studies, publicly available databases in Germany also provide evidence on the burden of RSV by age group over time.

This study aims to offer a detailed examination of the morbidity, measured by outpatient visits and hospitalisations, as well as mortality due to RSV, using public databases in Germany for the first time. It covers all age groups, with particular emphasis on those aged 60 years and over.

Methods

Study Design

This is a descriptive study analysing publicly available databases (see Table 1). To ensure a comprehensive and precise overview of the RSV BoD, data from various sources were combined, and datasets representing subsets of other datasets were excluded.

Table 1 Overview of the data bases used to extract data on the burden of RSV in Germany
Full size table

RSV Definitions

Definitions of an RSV case depend on the database and are either (1) PCR confirmed or (2) ICD based. It is unclear whether ICD-based cases have an underlying positive test and vice versa. To identify RSV cases, the following ICD-10 German Modification (GM) codes were used:

  • B97.4! RSV as the cause of diseases classified to other chapters

  • J12.1 RSV pneumonia

  • J20.5 Acute bronchitis due to RSV

  • J21.0 Acute bronchiolitis due to RSV

As B97.4 can only be used as a secondary diagnosis, cases identified via B97.4 were checked for their primary diagnosis. If the primary diagnosis was one of the three ‘J’-codes, those cases were further treated as ‘J’-codes and excluded to avoid double-counting in ICD-based data sources.

Data Sources

Outpatient Data

We utilised data from the sentinel surveillance network of the Arbeitsgruppe Influenza (AGI) at the Robert Koch-Institute (RKI) for outpatient visits. [10] This network comprises between 573 (season 2014/2015) and 700 (season 2022/2023) physician practices who report the weekly number of ARI consultations and test an age-stratified sample of their patients for various pathogens, including RSV. The AGI publishes weekly reports containing the ARI consultation frequency and the positivity rate of the samples taken [11]. The data are partly publicly available via online repositories [12] and partly only available from plots and tables in weekly reports. Data extraction from the latter is described in Appendix 1 in the Supplementary Material.

The database for notifiable diseases [13] of the RKI has been accessed to retrieve mandatorily reported RSV infections. RSV laboratory detection is notifiable in Germany since July 2023 [14]. Saxony is the only one of the 16 federal states with mandatory reporting of RSV infections since June 2002 (data only available since 2014). Since testing is performed irregularly, the mandatorily (laboratory) reported cases seem imperfect for BoD estimation [15] (a subanalysis on RSV-cases in Saxony can be found in Appendix 5 in the Supplementary Material).

Inpatient Data

Hospitalisation data were obtained from the German Federal Statistical Office, which provides information on all hospitalisations, stratified by main diagnosis, age, year, length of stay (LoS) and federal state. This was supplemented with information on intensive care units (ICU) and in-hospital mortality from the database of the Institut für das Entgeltsystem im Krankenhaus (InEK) [16], which collects a comprehensive survey of all Diagnosis-Related Groups (DRGs). This DRG statistic also allows for the inclusion of RSV cases with ICD-code B97.4 as secondary diagnosis (the top 20 main diagnoses are listed in Table A4-2 in Appendix 5 in the Supplementary Material).

Mortality

Mortality data were sourced from the German death registry and the in-hospital mortality from the aforementioned DRG statistic.

Data Analysis

To calculate crude and age-standardized incidence rates for the German population, we used population counts from the database of the German Federal Statistical Office destatis. More detailed age groups have been summarised to the lowest common denominator across all data sources for better comparability. Detailed data for each data source can be found in the appendices. No ethics approval was required as this article is based on routinely collected and published data and does not contain any new studies with human participants or animals performed by any of the authors. All analyses were performed using R version 4.1.2.

Results

Outpatient Visits

RSV-related outpatient consultations

Figure 1a illustrates the weekly incidence of medically attended RSV infections as reported by the AGI between October 2014 and September of 2023. The highest incidence per 100,000 inhabitants is observed in the age group of ‘0–4’, with lower incidences in the older age groups. The data indicates reduced RSV activity during the 2020/2021 COVID pandemic season and higher incidences in the subsequent 2021/2022 and 2022/2023 seasons.

Fig. 1
figure 1

Incidence per 100,000 inhabitants and absolute number of medically attended RSV infections (PCR tested) in the outpatient sector according to the sentinel surveillance system of the AGI at RKI by age group from the 2014/2015 to the 2022/2023 season

Full size image

Recalculating the same data to absolute values highlights a different aspect of the age-specific BoD. Figure 1b shows the absolute number of RSV cases by age group, aggregated by season. On average, between the 2014/2015 and 2022/2023 seasons (excluding the 2020/2021 pandemic season), the sentinel surveillance estimates 2,546,000 RSV-related outpatient consultations. Of these, 43.5% occur in the ‘0–4’ age group, 8.4% in ‘5–14’, 13.4% in ‘15–34’, 21.7% in ‘35–59’ and 13.0% in the ‘60+’ age group. The between-season variability shows a total of 1,313,100–3,911,800 cases overall [standard deviation (SD) 918,825] with cases in the ‘60+’ age group varying between 88,500 and 624,100 and averaging 331,000, with the proportion ranging between 6% and 17%.

Hospitalisations

Number of Hospitalisations

Figure 2 depicts the absolute number of hospitalisations (main diagnosis) per year between 2000 and 2022 for different age groups, stratified by ICD code, as available from the hospital statistics [19]. The absolute number of annual cases increases over the study period for all age groups and ICD codes. RSV-related bronchitis (ICD-code J20.5) and bronchiolitis (J21.0) are predominantly found in the youngest age group, ‘0–4’. This age group also shows the highest number of hospitalisations, reaching up to 30,000 in 2021. For the older age groups, pneumonia (J12.1) is the predominant hospital diagnosis.

Fig. 2
figure 2

Yearly number of hospitalisations in Germany due to RSV as main diagnosis for different age groups and ICD codes between 2014 and 2022

Full size image

While the annual case numbers steadily increase across all age groups, a significant and abrupt increase in case numbers can be observed in the age group 60+ from 257 cases in 2016 to 1317 cases in 2017. Another abrupt increase in 2019 is followed by 2 years of reduced but still overall high activity before another increase in the latest available year, 2022. The data is only available by calendar year, which distorts the generally low RSV activity during the early phase of the COVID-19 pandemic.

Hospitalisation Outcomes

The data from the hospital statistics also allow for the analysis of the length of stay (LoS) as depicted in Table A3-1 in Appendix 3 in the Supplementary Material. Patients with pneumonia (J12.1) show a longer LoS than patients with bronchitis (J20.5) or bronchiolitis (J21.0). The longest average LoS can be observed for pneumonia patients in the age group ‘35–59’ (11.5 days), followed by the age group 60+ (10.4 days, both groups median 8 days; 6/10 days 25%/75% percentile).

Figure 3 shows the results of the outcomes per season from the DRG statistic (numbers can be found in Table A4-1 in Appendix 4 in the Supplementary Material). Across all seasons, age groups and ICD codes, 17.0% of all RSV hospitalisations are treated in the ICU. The overall ICU rate was affected by the pandemic, with ICU rates of 7.5% for the 2019/2020 season, 6.5% for the 2020/2021 season, 30.8% for the 2021/2022 season and 9.5% for the 2022/2023 season. Excluding the two pandemic seasons (2020/2021 and 2021/2022), ICU rates are highest in the age group ‘40–59’ with 18.8% (followed by 16.1% in 60+) and lowest in the age group ‘1–2’ with 4.8%. ICU rates are generally higher for RSV-related pneumonia (13.3%) and B97.4-diagnoses (14.8%) compared with bronchitis and bronchiolitis (5.3%).

Fig. 3
figure 3

Number of hospitalisations for different age groups for the various RSV ICD codes and the seasons between 2019/2020 and 2022/2023, stratified by ICU stay (darker colour) and in-hospital mortality (brown colour)

Full size image

In-hospital mortality reveals a similar pattern: the overall mortality rate is 1.39% over all age groups and seasons. For the four seasons from the pre-pandemic 2019/2020 season to the post-pandemic 2022/2023 season, it is 0.82%, 0.09%, 0.32% and 2.76%, respectively. In-hospital mortality is consistently highest in the 60+ age group, at 9.4%.

Potential Underreporting in The Elderly in In-Patient Sector

The DRG statistic for seasons 2019/2020 to 2022/2023 reveals similar patterns as the outpatient sentinel surveillance (see Fig. 3). The time trend shows the effects of the measures taken during the initial phase of the COVID-19 pandemic and the generally low activity of ARIs in the 2020/2021 season. It also shows the rebound effect in the 2021/2022 season, especially in the younger age groups and for RSV-related bronchitis and bronchiolitis. In the 2022/2023 season, the case numbers appear similar to the pre-pandemic 2019/2020 season. An exception is the case numbers in the 60+ age group, showing a sevenfold increase between pre- and post-pandemic season. This increase can be observed for all outcomes, i.e. 6.9-fold increase for patients who survived RSV hospitalisation and were not admitted to the ICU, a 6.0-fold increase in patients who survived RSV hospitalisation and were admitted to the ICU, a 7.5-fold increase in in-hospital mortality of patients admitted to the ICU and an 11.1-fold increase in in-hospital mortality in patients not admitted to the ICU (see Table A4-1 in Appendix 4 in the Supplementary Material).

Stratifying the data by the hospital size in which the RSV patients were treated shows that the factor of increase varies between hospital sizes. Figure 4 shows the factor of post-pandemic (2022/2023) cases divided by pre-pandemic (2019/2020) cases across age groups per hospital size. A strong age pattern is visible, with an increasing ratio with increasing age after the age of 2 years. The data also show that the ratio increases more in smaller hospitals with less than 400 beds.

Fig. 4
figure 4

Factor comparing pre- and post-pandemic RSV cases by age groups across different hospital sizes

Full size image

Deaths

The mortality statistic [23] in Germany reports 214 deaths due to the RSV ICD codes J12.1, J20.5 and J21.0 between 1998 and 2022. Age information is available for 142 cases, with 96.5% of these occurring in individuals aged 60 years and over. No deaths are reported in the periods between 1998–2012 and 2014–2016. There are 5 cases reported in 2013, between 4 and 15 deaths annually between 2017 and 2021, and 90 deaths in 2022, all occurring in the 60+ age group. This figure is ten times lower than the corresponding numbers for in-hospital mortality derived from the DRG-statistic which shows a 10.5% in-hospital mortality rate.

Discussion

Summary

Our analysis presents the first comprehensive overview of outpatient visits, hospitalisations and deaths due to RSV in Germany, utilising extensive information from public data surveillance and registry sources over a prolonged period, including pre- and post-COVID-19 pandemic periods. It shows that RSV is responsible for 1.3–3.9 million outpatient consultations (out of 46.7 million ARI-related outpatient consultations overall) and up to 52,000 hospitalisations in the 2022/23 season (out of 17 million hospital stays overall), across all age groups. Annually, between 90,000 and 620,000 outpatient consultations (out of 6.9 million ARI-related consultations) and up to 12,800 hospitalisations (out of 0.99 million respiratory-related hospitalisations) in the 2022/2023 season occur in the 60+ age group. While data from the German mortality statistic indicate RSV-related mortality is close to zero, data from the DRG-statistic suggest that up to 1400 patients across all age groups die during their RSV-related hospital stay per season, with 93% of inpatient deaths from RSV occurring in the 60+ age group.

Our results for outpatient consultations align with other publications (our estimates have been transformed to the same unit of measurement and for the same age group where necessary and possible). Nacov et al. [7] estimate a cumulative incidence of symptomatic (non- and medically attended) RSV at 7.9% in the adult (18+) population between 7 December 2022 and 2 June 2023, compared with 2.5% in our estimation for medically attended symptomatic RSV in the same age group. The meta-analysis conducted by Savic et al. [5] finds an attack rate of 1.6% for 60+ in high income countries compared with an average of 1.3% in 60+ across all available seasons in our analysis.

Our analysis of the inpatient sector provides new insights compared with previous literature. Niekler et al. [24] use patient-level data from the hospital statistic, which are used in aggregated form in our study. They estimate hospitalisation incidence ranging between 0.10 and 11.08 per 100,000 inhabitants for 60+ over the period of January 2010 to December 2019. Our study estimates are identical for the same period but increase to 51.5 for the 2022/2023 season. This increase using the full sample of all hospital cases is also new compared with Kiefer et al. [9], which saw no difference between seasons in their single-centre study. Our ICU rates of 16.1% in 60+ are lower compared with Quarg et al. [2] (19.2% in adults in 2022/2023) but higher than the estimates from Cai et al. [25] (> 5.6% in 65+, 2013–2018), Kiefer et al. [9] (~10% in adults 2016/2017–2022/2023) and Niekler et al. [24] (11.0% in 60+ for 2010–2019). All studies have similar estimates for in-hospital mortality as the 10.4% for 60+ in our study: Ambrosch et al. around 10% in adults, Cai et al. 9.8% in 65+, Quarg et al. 11.1% in adults in 2022/2023, Niekler et al. 8.0% in 60+ for 2010–2019.

We hypothesize that the pandemic served as a natural experiment that changed the testing behaviour in hospitals due to COVID-19 and the availability of point-of-care (PoC) polymerase chain reaction (PCR) tests and multiplex PCRs that allow testing for multiple viruses simultaneously. This hypothesis is supported by the different increases of RSV diagnosis across different hospital types and age groups. While the increase is less pronounced in large hospitals such as university hospitals, which presumably had higher testing rates before the pandemic, smaller hospitals show a far stronger increase in RSV diagnoses as they may have only increased testing during the pandemic.

Limitations

Our findings have several limitations. For all data, the pre-test probability is unknown, i.e. whether only more severe cases or more high-risk patients are tested. The quality of outpatient data depends on the reporting of the voluntarily enrolled sentinel surveillance practices, with fluctuating numbers of practices and varying test samples over time. Reporting drops, especially in summer, complicating off-season positivity rate interpretation. Secondly, since the raw data is not available, data were extracted from PDFs, possibly resulting in deviating estimates, though no discrepancies were found when comparing with season totals and machine-readable data. Furthermore, legal requirements for medical sickness certificates for sick leave might lead to lower outpatient consultations from retired persons in the 60+ age group compared with other adult age groups with higher work force participation rates. Regarding inpatient data, we omitted data from the ICOSARI network and the DGPI as these datasets represent smaller samples from the sources we used with no additional information. While we were able to analyse full samples of the hospital statistic and the DRG statistic, no patient-level data could be analysed. The analysis of the German mortality statistic seems to drastically underestimate RSV as the cause of death and makes it difficult to differentiate between death due and death with an RSV infection. This might be due to complications of RSV such as heart failure or cardiopulmonary distress being documented as the cause of death rather than the RSV infection itself.

Conclusions

Our findings imply that the BoD of RSV in the 60+ age group has been underdetected regarding hospitalisations (sevenfold increase between 2019/2020 and 2022/2023) and in-hospital mortality (15-fold increase compared with the German mortality statistic). The limitations of the different testing methods [26, 27], make it challenging to assess how much of the underdetection has been addressed, especially in older adults. Nevertheless, our findings can provide a better understanding of the BoD of RSV, hel** public health decision-makers identify and assess the high unmet medical need for RSV prevention strategies that could help reduce hospitalisations and deaths. Further research is necessary to better understand the transmission dynamics of RSV infections in general and the risk factors of severe cases of RSV infection.