Rationale and protocol for a prospective cohort study of respiratory viral infections in patients admitted from emergency departments of community hospitals: Effect of respiratory Virus infection on EmeRgencY admission (EVERY) study

Abstract

Introduction

Respiratory syncytial virus (RSV) is a causative virus for the common cold worldwide and can result in hospitalisations and even death in patients with high-risk conditions and older adults. However, the relationship between RSV or other incidental respiratory infections and acute exacerbations of underlying conditions has not been well investigated. The primary objective of this study is to estimate RSV prevalence, risk factors for adverse outcomes or hospitalisation and their effect on the hospital course of patients with acute respiratory symptoms admitted from emergency departments. Furthermore, we evaluate the prevalence of other respiratory viruses associated with respiratory symptoms.

Methods and analysis

We are conducting a multicentre prospective cohort study in Japan. We plan to enrol 3000 consecutive patients admitted from emergency departments with acute respiratory symptoms or signs from 1 July 2023 to 30 June 2024. A nasopharyngeal swab is obtained within 24 hours of admission and the prevalence of RSV and other respiratory viruses is measured using the FilmArray Respiratory 2.1 panel. Paired serum samples are collected from patients with suspected lower respiratory infections to measure RSV antibodies at admission and 30 days later. Information on patients’ hospital course is retrieved from the electronic medical records at discharge, death or 30 days after admission. Furthermore, information on readmission to the hospital and all-cause mortality is collected 180 days after admission. We assess the differences in clinical outcomes between patients with RSV or other respiratory viruses and those without, adjusting for baseline characteristics. Clinical outcomes include in-hospital mortality, length of hospital stay, disease progression, laboratory tests and management of respiratory symptoms or underlying conditions.

Ethics and dissemination

The study protocol was approved by the institutional review boards of participating hospitals. Our study reports will be published in academic journals as well as international meetings.

Trial registration number

NCT05913700.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• Prospective cohort study with 3000 consecutive patients aged 50 years or older admitted from emergency departments with acute respiratory symptoms or signs, to be examined for the prevalences of respiratory syncytial virus (RSV) and other respiratory infections.

• The comprehensive measurements of patient clinical backgrounds, frailty, composition of family members and symptoms of family members to explore the risk factors for RSV or other respiratory viruses.

• The large-scale cohort study can explore the effects of RSV and other respiratory infections on clinical outcomes including length of hospital stay, in-hospital mortality or 180-day mortality, adjusting for the patients’ characteristics.

• Although potential confounders are extensively registered, potential unmeasured factors may be present and affect the relationship between viral infection and outcomes.

Introduction

Respiratory syncytial virus (RSV) is one of the most common seasonal respiratory viruses that cause upper and lower respiratory infections, with significant morbidity and mortality risk for young children and high-risk adults.¹ The prevalence of acute respiratory diseases with RSV was 2.4% in the adult population aged 65 years or older in Japan.² In the USA, the incidence of RSV at outpatient or emergency department (ED) visits was high among adults aged 50 years or older, ranging from 195 to 1790 per 10⁵ population.^{3 4} Another community-based cohort study showed that the annual incidence rate of acute respiratory illness with RSV was 6.1 per 1000 persons for the overall population or 0.6 per 1000 persons in adults aged 65 years or older.⁵ However, there is limited evidence demonstrating RSV prevalence and burden among patients with risk factors of underlying diseases or conditions. One prospective cohort study reported that 4%–10% of the elderly with risk factors contracted RSV annually and that 10% of patients hospitalised during the winter season were infected with RSV.⁶

Many patients are admitted from the ED for acute exacerbations of underlying respiratory diseases or other chronic conditions that may have been triggered by respiratory viral infections, including RSV. Small studies in Japan have indicated that 3.7%–8.7% of patients with acute exacerbations of chronic lung diseases had RSV.^{7 8} The risk factors of RSV or other respiratory viral infections may differ, including patient age, comorbidities, frailty, family member composition and symptoms of family members. These differences may be associated with clinical outcomes, such as in-hospital mortality, length of hospital stay, disease progression and long-term prognosis. Therefore, the prevalence and risk factors of RSV and other respiratory viral infections are of clinical interest among patients requiring emergency admission.

In randomised clinical trials, single-dose RSV vaccination has demonstrated high efficacy in preventing symptomatic RSV-associated lower respiratory tract disease among adults aged 60 years and older.^9–12 Therefore, an RSV vaccination strategy presents an important public health opportunity for elderly populations at high risk of severe RSV infections, and investigating the local burden of hospitalisation due to RSV is important to determine the potential impact of an RSV vaccination strategy. The primary objective of this study is to estimate RSV prevalence, risk factors and their effect on the hospital course in adult patients with acute respiratory symptoms presenting to EDs and requiring hospitalisation. Additionally, we evaluate the prevalence of other respiratory viruses associated with respiratory symptoms.

Methods and analysis

Study design and setting

We are conducting a prospective cohort study of consecutive patients admitted from the EDs of three participating hospitals who present with acute respiratory symptoms or signs (figure 1). To ensure generalisability and representation of rural and urban burden of communicable disease, we chose three acute care hospitals that provide comprehensive care to residents in Japan, namely Shimane Prefectural Central Hospital (rural), Nara City Hospital (suburban) and Rakuwakai Otowa Hospital (urban). Patient enrolment starts on 1 July 2023 and ends on 30 June 2024, with the last study-related follow-up visit planned for 31 July 2024. There are no interventions in this study and patients’ hospital courses will be determined by the attending physician based on standard practices for both underlying diseases for admission and other conditions.

Figure 1.

Study flow chart. The blue box indicates the data collected in this study.

Patients and eligibility criteria

We plan to enrol consecutive patients admitted from EDs of the participating hospitals who meet the eligibility criteria and have acute upper or lower respiratory symptoms or signs (box 1).

Box 1. Patient eligibility criteria.

Inclusion criteria

• Aged 50 years or older.

• Admission from emergency department.

• Have at least one of the following respiratory symptoms or signs for 24 hours to 7 days:

  • Nasal congestion.

  • Rhinorrhoea.

  • Sore throat.

  • Cough.

  • Sputum.

  • Dyspnoea.

  • Wheeze.

  • Crackles or rhonchi.

  • Tachypnoea (≥20 breaths per minute).

  • Decreased saturation of oxygen (<95%).

  • Admission with oxygen supplementation.

Exclusion criteria

• Scheduled admission.

• Admission for trauma care.

• Have nasopharyngeal cavity diseases or deformities which interfere with nasopharyngeal sampling.

• Admission for end-of-life care.

• Declined study participation by informed consent or opt-out method.

Respiratory symptoms or signs must meet the case definition of acute respiratory infection as described in a previous report, and persist for at least 24 hours and less than 7 days before admission (box 2).¹¹

Box 2. Definitions of respiratory symptoms and signs.

Upper respiratory symptoms

• Nasal congestion/rhinorrhoea.

• Sore throat.

Lower respiratory symptoms

• New or increased cough.

• New or increased sputum.

• New or increased dyspnoea (shortness of breath).

Lower respiratory signs

• New or increased wheezing.

• New or increased crackles (rales)/rhonchi, based on chest auscultation.

• Respiratory rate ≥20 breaths per minute.

• Low or decreased oxygen saturation (=O₂ saturation <95% or ≤90% if baseline is <95%).

• Need for oxygen supplementation.

Ethics and informed consent

The study is conducted in accordance with the latest versions of the Declaration of Helsinki and Ethical Guidelines for Medical and Biological Research Involving Human Subjects issued by the Ministry of Health, Labour and Welfare, the Ministry of Education, Culture, Sports, Science and Technology, as well as the Ministry of Economy, Trade and Industry of Japan.¹³ The study protocol was approved by the institutional review boards of the Shimane Prefectural Central Hospital (Churin R23-003), Rakuwakai Otowa Hospital (Rakugaku-Rin-01-000208) and Nara City Hospital (NCH Rin 23-2). This study uses data from electronic medical records (EMRs) and a swab is used to sample micro-organisms without additional interventions. Therefore, informed consent will be obtained through an opt-out method or documented verbal consent in the EMR. In the opt-out method, investigators describe the study design, collected data and study contacts on each participating hospital’s webpage. Patients who would not like to be registered in the study can request the removal of their individual data. The institutional review boards of the participating hospitals have approved this method based on current ethical guidelines.¹³ Written informed consent will be obtained from patients with suspected lower respiratory infections (bronchitis and pneumonia) who will undergo paired serum RSV antibody measurements. The study was registered at ClinicalTrials.gov (NCT05913700).

Data collection and follow-up

All data are recorded in EMRs by the physicians in charge. Respiratory symptoms are universally assessed in EDs using the standard EMR format when patients are admitted. Information is retrieved from the EMRs related to medical history, baseline characteristics, living status, physical findings, laboratory tests, chest radiography and electrocardiogram on admission (online supplemental table 1). As preventive measures against COVID-19 and other communicable diseases in wards, nasopharyngeal swabs are obtained within 24 hours after admission as a standard practice and will be sampled at EDs or hospital wards. The swab is transferred to the onsite laboratory office to be measured with the FilmArray Respiratory 2.1 panel by trained technicians or attending physicians. If the immediate measurement is not conducted due to workflow or medical care congestion, the swab is stored at 2°C–8°C for up to 3 days before measurement. Information for patients who undergo echocardiography, head computed tomography (CT) or chest CT for clinical indications is retrieved from the EMRs.

Serum RSV antibodies are obtained from patients with suspected lower respiratory infections (bronchitis and pneumonia) at admission and 4 weeks after. All patients with suspected lower respiratory infections are asked to undergo serum RSV antibody testing and those who provided written informed consent are enrolled. Paired serological tests are conducted with the neutralising antibody method. Patients who are suspected to have lower respiratory infection have at least one of the following lower respiratory tract symptoms or signs (cough, sputum, dyspnoea, wheezing, crackles or rhonchi, tachypnoea (≥20 breaths per minute), decreased oxygen saturation (<95%), admission with oxygen supplementation), and these symptoms or signs have to be explained by the presence of lower respiratory infections.¹¹ Patients with infiltrates on chest radiographs or chest CT are considered to have pneumonia, while those without are considered to have bronchitis. The first sample is obtained within 1 week of symptom onset and the second sample 4 weeks after onset. The allowed interval between paired serum samplings is set between 3 and 6 weeks.

At the time of discharge, death or 30 days after admission, hospital course information is retrieved from the EMRs, including baseline characteristics, living status and hospital course (online supplemental table 1). For patients admitted until 31 January 2024, hospital readmission and all-cause mortality are measured 180 days after admission.

All data are electronically retrieved from EMRs and anonymously stored in an electronic data capturing (EDC) system. Data are centrally monitored by data managers of an independent data centre, which regularly queries inconsistent or missing data. Two investigators review all data and adjudicate the clinical course of each patient. The physicians in charge and clinical research coordinators at participating hospitals supplement data in response to queries.

Outcome measures

The primary outcome is RSV presence on admission, measured using the FilmArray Respiratory 2.1 panel, which has a sensitivity and specificity of 0.911 and 0.987 for RSV, respectively.¹⁴ The secondary outcomes are summarised in box 3.

Box 3. Primary, secondary and safety outcomes.

Primary outcome

• Respiratory syncytial virus presence on admission measured by the FilmArray Respiratory 2.1 panel.

Secondary outcomes

• Respiratory virus and micro-organisms on admission measured by the FilmArray Respiratory 2.1 panel.

• Lower respiratory tract infections.

• All-cause mortality.

• All-cause readmission.

• Length of hospital stay.

• Changes in Clinical Frailty Scale.

• Changes in functional oral intake score.

• Changes in modified Rankin Scale.

• Presence of nasal congestion or rhinorrhoea.

• Presence of sore throat.

• Presence of cough.

• Presence of sputum.

• Presence of dyspnoea.

• Presence of wheeze.

• Presence of crackles or rhonchi.

• Presence of tachypnoea.

• Presence of decreased oxygen saturation.

• Presence of oxygen supplementation.

• Length from onset to admission of acute respiratory infection symptoms.

• Presence of family member who attends preschool or school.

• Presence of symptoms of family member.

• Use of antimicrobials.

• Admission to intensive or high care unit.

• Respiratory complications.

• Cardiovascular complications.

• Cerebrovascular complications.

Safety outcomes

• Insert site bleeding or peripheral nerve injury by blood drawing.

• Any adverse events related to study procedure.

Respiratory viruses and micro-organisms are measured on admission using the FilmArray Respiratory 2.1 panel (online supplemental table 2). We also determine the positivity of paired serological tests using the neutralising antibody method for RSV infection.

Sample size

We plan to register 3000 patients over a 1-year enrolment period. The sample size is determined using the accuracy of prevalence and the number of adjusters for multivariable linear or logistic regression models, exploring the potential effects of RSV infection on clinical outcomes. The lower bound of the expected RSV prevalence was assumed to be 3% based on the incidence data of a previous study.⁶ The sample size of 3000 yields a 0.5% confidence interval width (ranging from 2.5% to 3.5%), ensuring accurate prevalence estimates at a 95% confidence level. As the prevalence of 3% will yield 90 patients with RSV, 9 potential variables are examined to be risk factors for RSV.

A previous cohort study at one of the study sites demonstrated a 5% in-hospital mortality among general admissions of patients aged 15 years or older.¹⁵ Additionally, a meta-analysis reported a 7.13% in-hospital case fatality rate among RSV-associated acute respiratory infections in adults aged 60 years and older in high-income countries, a setting similar to this study.¹⁶ Considering the eligibility of high-risk patients with advanced age admitted from EDs, the in-hospital mortality rate was assumed to be 5%–10%, which results in 150–300 mortalities among 3000 patients. The multivariable regression model also allows up to 300 variables for a linear model and 15–30 variables for logistic or Cox proportional hazard models. These models allow us to evaluate the effect of RSV and other respiratory viruses on length of hospital stay or mortality, adjusting for clinically relevant variables.

Statistical analysis

The statistical analysis plan (SAP) will be specified before data analyses and all analyses will be conducted according to the SAP. The full analysis set (FAS) includes all patients for whom RSV infection is confirmed using the FilmArray Respiratory 2.1 panel while the per-protocol set (PPS) includes patients for whom planned outcome data as per the protocol are available. The study flow will present how the FAS and PPS are defined, along with detailed information for excluded patients. Descriptive statistics with appropriate expressions will be presented for all study participants and subgroups. Continuous variables will be described as means and standard deviations, or medians and interquartile ranges, depending on the data distribution. Categorical variables will be described as frequencies with percentages.

RSV prevalence will be estimated for the overall cohort and for subgroups defined according to clinical background (age group, sex, underlying diseases, family composition or other) or lower respiratory tract infections (LRTIs). After determining RSV prevalence, we will compare the clinical backgrounds of patients with and without RSV infection, in the overall cohort and LRTI subgroups, using appropriate statistical testing for differences. Additionally, we will compare the use of antibiotics or antiviral drugs between patients with and without RSV infections. As clinical backgrounds are potential risk factors for RSV infection, a multivariable logistic model using these variables will be constructed to explore the risk factors for RSV infection. The same multivariate regression models will be constructed in subgroups of interest (age group, underlying disease or other).

The incidence of LRTIs, in-hospital mortality and length of hospital stay will be determined for the overall cohort and by subgroup based on the presence of RSV infections, using appropriate statistical testing for differences. Risk factors for LRTIs, in-hospital mortality and length of hospital stay will be explored separately by constructing appropriate multivariate regression models.

Changes in symptoms, signs, biomarkers and clinical frailty scale during the hospital stay will be estimated for the overall cohort and subgroups, based on the presence of RSV infection or LRTIs, using appropriate statistical testing for differences. Multivariate linear regression models will be constructed to explore the factors associated with these changes. An exploratory clinical prediction model for predicting RSV infections among patients admitted from EDs will be developed. Further exploratory analyses will be conducted in conjunction with discussions among investigators.

All statistical analyses will be performed by study statisticians and members of the data centre (Institute for Clinical Effectiveness, Kyoto, Japan) using JMP Pro V.17.0 (SAS Institute) and SAS V.9.4 (SAS Institute). As this study aims to estimate RSV prevalence, imputation of missing variables is not planned and data will be analysed as observed. For the multivariable analyses, independent variables with a significant fraction of missing data will be added with missing codes as sensitivity analyses. Other missing observations will be excluded from corresponding analyses. All p values are two sided and p values less than 0.05 will be considered statistically significant in this study. Considering the exploratory nature of this study, p values will not be adjusted; p values less than 0.05 may not be sufficient to indicate statistically significant differences.

Study oversight

This study is planned and designed by a steering committee consisting of physicians affiliated with academic institutions and hospitals (online supplemental table 3). The steering committee will oversee all study aspects, including protocol development, study operation, data collection, statistical analyses and reporting. The study statistician is responsible for planning, conducting and reporting statistical analyses. Data managers will centrally monitor the data throughout the study period.

Representatives from GlaxoSmithKline participate in the study design and manuscript review, but not in the study operation, data collection or data analysis; the latter tasks are solely conducted by the academic authors. Representatives from GlaxoSmithKline participate in the discussion and interpretation of the results derived from the data. The academic authors are responsible for making the final decisions related to manuscript submission. GlaxoSmithKline contracts with the study secretariat (located at the Institute for Clinical Effectiveness in Kyoto, Japan) and funds the operations and supplies for this study, but does not fund any compensation or profit for investigators or participating hospitals under the contract.

Patient and public involvement

All investigators should make an effort to incorporate patient and public opinions or knowledge when reporting the study. There is no patient or public involvement in study protocol development.

Timeline

This study was approved in June 2023 by the Institutional Review Boards of the Shimane Prefectural Central Hospital (Churin R23-003), Rakuwakai Otowa Hospital (Rakugaku-Rin-01-000208) and Nara City Hospital (NCH Rin 23-2). We started patient enrolment on 1 July 2023. On admission, all included patients are assessed for respiratory viruses and micro-organisms using the FilmArray Respiratory 2.1 panel, baseline characteristics, living status, comorbidities, physical conditions, symptoms, signs, drugs and biomarkers. Serum antibodies against RSV are measured in patients with suspected LRTIs within 1 week of admission. At discharge or 30 days of admission, patients are assessed for their hospital course, discharge route, physical conditions, symptoms, signs, biomarkers and prescription. Patients discharged in less than 30 days are assessed for respiratory symptoms, signs, physical condition and hospital readmission 30 days after admission. Secondary serum antibodies against RSV are measured in patients with suspected LRTIs within 3–6 weeks of admission. Patients admitted until 31 January 2024 are assessed for death and readmission to the hospital 180 days after admission. Patient enrolment ends on 30 June 2024 and the last study-related follow-up visit date is 31 July 2024. The primary data collection completion and analysis date is 31 December 2024 and the study completion date is 31 March 2025.

Ethics and dissemination

To ensure the adequacy of the study performance, the study protocol and progress will be monitored. Central monitoring is continuously performed using the EDC system and on-site or off-site monitoring is employed if needed. If necessary, the principal investigator should perform an audit to ensure the reliability of the study operation. Auditors should be persons who are not involved in study operations or monitoring. If protocol amendments are required after implementation, the steering committee is responsible for revising and implementing the protocol as early as possible. The principal investigator must promptly report the protocol revision and the reasons for revision to the hospitals’ institutional review boards in written form.

This investigator-initiated study is conducted in collaboration with GlaxoSmithKline Biologicals SA. All investigators are responsible for managing potential conflicts of interest under their institutional rules. Potential conflicts of interest will be disclosed in the report of this study.

The reports of this study will be published in academic journals and international or national meetings of scientific societies. After publication in academic journals, the principal and site investigators will disclose the results and clinical implications of the study on the participating hospital homepages.

Ethics statements

Patient consent for publication

Not applicable.