Severity and mortality of respiratory syncytial virus vs influenza A infection in hospitalized adults in China

Abstract Background Respiratory syncytial virus (RSV) is an important cause of medically attended acute respiratory illnesses in older adults but awareness of the relevance of RSV in older people remains lower than that of influenza, which exhibits similar clinical characteristics to those of RSV. Objectives This study was performed to assess the clinical significance of RSV in respiratory samples from hospitalized adults. Methods Characteristics and outcomes in adults (≥18 years) hospitalized for RSV infection (n = 51) were compared with a cohort hospitalized for influenza A infection (n = 279) in a single‐center retrospective cohort study in Beijing, China. Results Respiratory syncytial virus patients were slightly older, with no significant differences in underlying chronic conditions. Lower respiratory tract infection and cardiovascular complications were more frequent (P < .05) in RSV patients. Rates of mortality in the RSV cohorts were significantly higher within 30 days (13.7% vs 5.0%, P = .019) and 60 days (17.6% vs 7.5%, P = .021). Bacterial co‐infection in respiratory samples was associated with reduced survival among RSV patients (log rank, P = .013). Conclusions Respiratory syncytial virus is a common cause of serious illness among hospitalized adults in China with greater mortality than influenza A. Increased awareness and the availability of antiviral agents might increase the scope for successful management.


| INTRODUC TI ON
Respiratory syncytial virus (RSV) used to be known primarily as a respiratory pathogen of young children and many laudable projects such as the World Health Organization RSV surveillance platform focus on pregnant women and young children. 1 However, in recent decades, awareness has grown of the importance of RSV infection to the health of older adults. In the United States, RSV infections occur at an annual rate of up to 10% in older adults, a rate which can exceed that observed for influenza in this population group. 2 In the older adult population, RSV infection can have serious consequences: RSV is responsible for around 12% of all medically attended acute respiratory illnesses in older adults 3 and the incidence of RSV-associated hospitalization increases with age. 4 Notably, whereas earlier data used a cutoff point of ≥65 years to define "older adult," 2 a recent study suggested that increased risk of severe RSV disease may commence at early as at 50 years of age. 5 Among older adults hospitalized with RSV, a mortality rate of 6%-8% has been reported. 3 RSV infection was shown to lead to severe lower respiratory complications and even respiratory failure in elderly in Hong Kong, with a mortality rate up to 11.9% within 60 days. 6 It is likely that even these alarming numbers represent an underestimation of the burden of RSV infection in older adults. 7 Despite these numbers, awareness of the relevance of RSV in older people remains lower than that of influenza, which exhibits similar clinical characteristics to those of RSV and which has been recognized for generations as cause of severe morbidity and mortality in older adults. 8 The need for awareness and distinction between the two diseases is illustrated by the fact that some 200 000 hospitalizations annually are associated with RSV infection compared with 300 000 hospitalizations secondary to influenza in the same population. 4,5 Low awareness is also reflected in a dearth of international data.
Most of the available studies were performed in the United States.
Particularly for China, with the world's largest population and an increasing proportion of elderly individuals, more data are urgently needed on the prevalence, clinical manifestations, complications, and outcomes of severe RSV infections in hospitalized adults. 6,9,10 The recent progress on antiviral treatments for RSV 11,12 has given such data unprecedented relevance to clinicians.
We performed a retrospective single-center study of a large co-

| Clinical data collection and definitions
Electronic and written medical records were reviewed for all subjects.
Data collected included demographic details, comorbid illnesses, presenting symptoms and signs, antiviral and antibiotic use, corticosteroid treatments received (intravenous or oral steroids), intensive care unit (ICU) admission, hospital length of stay, occurrence of complications, requirement for ventilatory support, exacerbation of chronic conditions, and all-cause death within 30 days and 60 days. Medical complications associated with RSV infection were defined as a new or exacerbated medical condition as confirmed by laboratory and radiographic studies. Lower respiratory complications were defined as radiologically confirmed pneumonia or exacerbation of asthma/bronchitis/chronic obstructive pulmonary disease (COPD). Cardiovascular complications were defined as the occurrence or exacerbation of cardiac symptoms (coronary syndrome, arrhythmia, myocarditis, and decompensated heart failure) and/or acute cerebrovascular events. 13,14 Bacterial superinfection is defined as the isolation of one or more bacterial pathogen from nasopharyngeal swabs, sputum, bronchoalveolar lavage fluid and/or blood and/or urine samples.

| Virus identification
Respiratory syncytial virus and influenza A virus infection were confirmed by analysis of nasopharyngeal swabs, sputum, bronchoalveolar lavage fluid and/or blood and/or urine samples using RSV Nucleic Acid Detection Kit (Liferiver) and Influenza A Virus Nucleic Acid Detection kit (Liferiver), respectively. 15

| Statistical analysis
Categorical variables are presented as frequencies and percentages.
Continuous variables are described as mean, standard deviation, and range. Comparisons of proportions were performed with chi-square and Fisher's exact tests; continuous variables were compared using Student's t test. All probabilities were 2-tailed, with statistical significance defined as P ≤ .05. Binary logistic regression was performed to estimate the odds ratio (OR) and 95% confidence interval (CI) for clinical hospitalization outcomes in RSV-infected patients compared with influenza A virus-infected cohorts. Survival curves were generated using the Kaplan-Meier method and compared using the log-rank test. All analyses were performed using PASW Statistics software, version 18.0.

| Study population
Demographic characteristics and comorbidities prior to admission of all hospitalized patients are presented in Table 1 (chi-square test; Figure 1). There were no significant differences between the groups in rates of comorbid conditions at admission although cardiac disease, respiratory disease, and cerebrovascular disease were more common in RSV-infected populations than those in influenza A-infected cohorts (Table 1).

| Clinical presentation and outcomes
Clinical symptoms and outcomes in the cohorts are presented in Table 2. Fever, cough, and sputum production were the most frequent presenting signs in both cohorts, but RSV cases were less likely than influenza A cases to report fever (P < .001; chi-square test) and cough (P = .026; chi-square test). The rates of bacterial superinfection in each kind of samples (respiratory samples or blood samples or urine samples) were similar between RSV-infected patients and influenza A patients (P > .05; chi-square test b Bacterial superinfection is defined as the isolation of one or more bacterial pathogen from respiratory samples (nasopharyngeal swabs, sputum, and bronchoalveolar lavage fluid) and/or blood and/or urine samples. c Lower respiratory complications included radiologically confirmed pneumonia or exacerbation of asthma/bronchitis/chronic obstructive pulmonary disease [13,14]. d Cardiovascular complications included the occurrence or exacerbation of cardiac symptoms (coronary syndrome, arrhythmia, myocarditis, and decompensated heart failure) and/or acute cerebrovascular events [13,14].
Rates of mortality in the RSV cohorts were significantly greater than that for influenza A-infected patients within 30 days (13.7% vs 5.0%, P = .019; chi-square test) and 60 days (17.6% vs 7.5%, P = .021; chi-square test), respectively. There were no differences in median time from admission to death between the groups nor in the median duration of hospitalization for survivors. In the binary logistic regression analyses, the odds of hospitalization outcomes (cardiovascular complications, pneumonia, lower respiratory tract complications, the need for invasive mechanical ventilation and 60-day mortality) in RSV cases were higher than in those hospitalized with influenza A infection, but the 95% CI crossed the boundary for all variables except for cardiovascular complications (Table 3).

| Analysis of RSV cases with fatal outcomes
Nine patients with RSV infection died during hospitalization. A comparison with survivors showed no differences in sex, comorbidities, blood biochemical indices, symptoms, and signs; notably though, bacterial superinfection in respiratory samples (nasopharyngeal swabs, sputum, or bronchoalveolar lavage fluid) was more common among non-survivors than that in survivors (P = .021, chi-square test; Table 4) and was showed to be related to lower survival ( Figure 2 Table 4). Survivors tended to have less cardiac disease and lower respiratory tract complications, but these differences did not reach statistical significance.

| D ISCUSS I ON
With RSV-specific antiviral therapy advancing in clinical development, the question of differentiating RSV infection from that of influenza in adult populations will likely become highly relevant to care decisions worldwide. 12,16 The current retrospective study is The differences in presentation at admission between RSV and influenza A infection are of interest. Fever and cough were less common among RSV cases, but rates of lower respiratory tract and cardiovascular complications, especially pneumonia, were greater in the RSV-infected population. The latter complications may partially explain the higher rates of mortality in this group. It is also possible that lower respiratory tract disease progression is more rapid in RSV infection, although further research into the mechanisms and natural history may be necessary. 19,26 The seasonal pattern of RSV infection in children in China has recently been shown to be very similar to those in the United States, 27 and it is reasonable to assume that this would also be the case for adult disease. The development of efficacious interventions against RSV b Cardiac disease included the occurrence or exacerbation of cardiac symptoms (coronary syndrome, arrhythmia, myocarditis, and decompensated heart failure) [13,14]; c Lower respiratory complications included radiologically confirmed pneumonia or exacerbation of asthma/bronchitis/chronic obstructive pulmonary disease [13,14]; d Bacterial superinfection is defined as the isolation of one or more bacterial pathogen from respiratory samples (nasopharyngeal swabs, sputum, and bronchoalveolar lavage fluid) and/or blood and/or urine samples.
should be a high priority as they could reduce mortality and morbidity, and burdens on the healthcare system. Furthermore, if early identification and diagnosis of RSV infection in hospitalized adults with bacterial co-infection enabled the timely implementation of appropriate therapies to reduce complications, this would reduce mortality, morbidity, and healthcare costs. An economic analysis in the United States estimated the average cost of RSV hospitalizations to be more than twice that of influenza A. 28 No economic data are available for our cohort but a substantial economic burden can be inferred from the demonstrated severity of the RSV-infected population.
There are limitations to this study. It was a single-center analysis with modest sample size, and the features of the setting may not be representative of China as a whole. As a retrospective study, causation, for example, between bacterial co-infection and mortality cannot be definitively determined. There was no analysis of RSV genotypes, which would be important for future epidemiological studies as well as to possibly assess future anti-RSV therapies. 29 In conclusion, RSV infection is a common cause of serious illness among hospitalized Chinese adults, with greater morbidity and mortality than influenza A virus infection. Greater awareness of the serious nature of RSV infection among healthcare professionals would enable adult RSV-infected patients, particularly those with bacterial infection or prior cardiac and pulmonary disease to be recognized in time and given appropriate treatments on admission. 30 If recent reports of successful antiviral treatment for RSV 12 are confirmed in further clinical trials these needs will take on a heightened relevance.

F I G U R E 2
Kaplan-Meier survival curves for patients with RSV infection with (n = 16) and without (n = 35) bacterial coinfection in respiratory samples (nasopharyngeal swabs, sputum, or bronchoalveolar lavage fluid), respectively