Retrospective Observational Multicenter Study on the Burden of Hospitalized Patients With Respiratory Tract Infection

Background Respiratory tract infections (RTIs) are among the leading causes of morbidity, mortality, and healthcare utilisation worldwide. A wide range of viral and bacterial pathogens contribute to the incidence and severity of RTIs, with influenza viruses, respiratory syncytial virus (RSV), and SARS-CoV-2 being among the most prominent agents. The epidemiology and clinical impact of these pathogens vary considerably by age group, comorbidity status, and season. In the Netherlands, comprehensive real-world data on the burden of RTI hospitalisations, stratified by pathogen, age group, and risk category, are limited. National surveillance figures (RIVM, Nivel) are based on sentinel networks and model-based extrapolations without direct hospital-level data on healthcare resource utilisation, ICU outcomes, or costs per pathogen. For RSV specifically, available burden estimates are largely derived from European models (ECDC ERVISS) with considerable uncertainty margins. Key policy decisions, including the potential expansion of the National Immunisation Programme (NIP) with RSV vaccination strategies (e.g. nirsevimab), are currently made without an empirical Dutch evidence base on RTI hospitalisation burden. This study directly addresses this evidence gap. Study Design Retrospective, observational, multicenter cohort study using routinely collected hospital data from up to 10 participating Dutch hospitals, comprising a mix of academic, large peripheral, and smaller peripheral centres. Observational period: 1 April 2018 to 31 July 2026. This period is fully closed prior to any data extraction. Data are extracted, pseudonymised, and linked within each participating hospital before transfer to the research team. No patient contact occurs and no new data are collected. Data sources: administrative claims data (Diagnose Behandel Combinatie, DBC), electronic prescribing data (add-on medications), and laboratory/microbiology databases. Study Population All patients (all ages, paediatric and adult) with at least one hospital encounter carrying a primary or secondary RTI ICD-10 diagnosis code (see protocol Appendix 1 for full code list) during the observational period. Patients who have formally objected to the use of their data for scientific research are excluded. Primary Research Question (PICO) In adult and paediatric patients hospitalised in Dutch hospitals with a respiratory tract infection (RTI) as primary or secondary ICD-10 diagnosis (April 2018 - July 2026): what is the burden of RSV-related hospitalisations - measured by incidence, length of stay, ICU admission rate, and in-hospital mortality - compared to influenza and SARS-CoV-2, stratified by age group and comorbidity class? Study Hypotheses H1: The burden of RSV-related hospitalisations (as measured by length of stay and ICU admission rates) differs significantly between paediatric and adult patients in Dutch hospitals. H2: Pathogen-specific RTI incidence varies significantly across annual seasonal cycles (April-March). H3: Microbiological testing rates show significant inter-hospital variation for comparable patient populations. Objectives Primary: To quantify the incidence, demographic distribution, and clinical outcomes (length of stay, ICU admission rate, ICU length of stay, in-hospital mortality) of patients hospitalised for RTIs - including RSV, influenza, SARS-CoV-2, and other viral or bacterial pathogens - across all ages and over multiple consecutive annual cycles (2018-2026). Secondary: (1) To measure healthcare resource utilisation (HCRU) during and 30 days post-discharge; (2) to characterise microbiological testing patterns across participating hospitals; (3) to quantify direct hospital costs per RTI-related hospitalisation. Statistical Analysis Descriptive statistics for all outcomes, stratified by pathogen, age group, comorbidity class, and annual seasonal cycle. Continuous variables summarised as mean (SD) or median (IQR); categorical variables as frequencies and percentages. Sensitivity analyses using alternative RTI case definitions and time windows. All analyses prespecified and conducted in R. Results reported only for strata with n≥5 to prevent identification of individual patients. Ethical and Legal Framework The study is reviewed by the Medical Ethics Committee Utrecht (MEC-U), registration number W26.018. Processing of pseudonymised health data is based on Article 9(2)(j) GDPR in conjunction with Article 24 UAVG (scientific research), Article 89(1) GDPR (appropriate safeguards), and Article 7:458 BW (WGBO). Individual informed consent is not obtained; patients retain the right to object via their treating hospital. The pseudonymisation key remains exclusively at the source institution and is not accessible to the research team. Results are reported in aggregate only (minimum group size n≥5). Dissemination Publication of results in peer-reviewed journals constitutes the primary and unconditional dissemination objective. Results will be submitted for publication irrespective of their direction or magnitude, in accordance with ICMJE authorship criteria. Initial results are anticipated by end of 2026, timed to inform the 2027 NIP evaluation cycle.

Inclusion criteria

• All patients with any care activity during a hospitalisation registered under any primary or secondary RTI ICD-10 code; hospital admission beginning within April 2018 - July 2026.

Exclusion criteria

• Patients who have formally objected to the use of their hospital data for scientific research.