Abstract
Although broad knowledge of influenza viral pneumonia has been established, the significance of non-influenza respiratory viruses in community-acquired pneumonia (CAP) and their impact on clinical outcomes remains unclear, especially in the non-immunocompromised adult population.
Hospitalised immunocompetent patients with CAP were prospectively recruited from 34 hospitals in mainland China. Respiratory viruses were detected by molecular methods. Comparisons were conducted between influenza and non-influenza viral infection groups.
In total, 915 out of 2336 adult patients with viral infection were enrolled in the analysis, with influenza virus (28.4%) the most frequently detected virus, followed by respiratory syncytial virus (3.6%), adenovirus (3.3%), human coronavirus (3.0%), parainfluenza virus (2.2%), human rhinovirus (1.8%) and human metapneumovirus (1.5%). Non-influenza viral infections accounted for 27.4% of viral pneumonia. Consolidation was more frequently observed in patients with adenovirus infection. The occurrence of complications such as sepsis (40.1% versus 39.6%; p=0.890) and hypoxaemia (40.1% versus 37.2%; p=0.449) during hospitalisation in the influenza viral infection group did not differ from that of the non-influenza viral infection group. Compared with influenza virus infection, the multivariable adjusted odds ratios of CURB-65 (confusion, urea >7 mmol·L−1, respiratory rate ≥30 breaths·min−1, blood pressure <90 mmHg (systolic) or ≤60 mmHg (diastolic), age ≥65 years) ≥3, arterial oxygen tension/inspiratory oxygen fraction <200 mmHg, and occurrence of sepsis and hypoxaemia for non-influenza respiratory virus infection were 0.87 (95% CI 0.26–2.84), 0.72 (95% CI 0.26–1.98), 1.00 (95% CI 0.63–1.58) and 1.05 (95% CI 0.66–1.65), respectively. The hazard ratio of 90-day mortality was 0.51 (95% CI 0.13–1.91).
The high incidence of complications in non-influenza viral pneumonia and similar impact of non-influenza respiratory viruses relative to influenza virus on disease severity and outcomes suggest more attention should be given to CAP caused by non-influenza respiratory viruses.
Abstract
The high incidence of complications in non-influenza viral pneumonia and similar impact of non-influenza viruses relative to influenza virus on disease severity and outcomes suggest more attention should be given to CAP caused by non-influenza viruses http://bit.ly/2vRTvFK
Introduction
Community-acquired pneumonia (CAP) is the leading cause of death worldwide among communicable diseases [1]. The development of molecular diagnostic techniques has markedly improved our ability to identify viruses, with numerous studies having reported a high prevalence of viral infections in patients with CAP [2]. A recent meta-analysis incorporating 31 observational studies showed that the proportion of viral infection was 24.5% in hospitalised patients with CAP, with influenza virus types A and B (IFV) most frequently identified, followed by human rhinovirus (HRV), respiratory syncytial virus (RSV) and human coronavirus (CoV) [2].
Previous studies have established the clinical features, complications and outcomes associated with influenza viral pneumonia [3, 4]. However, most data concerning non-influenza viral pneumonia were derived from case reports and case series, since non-influenza respiratory viruses were most frequently known as the cause of upper respiratory tract infections [3]. Additionally, the majority of studies focusing on non-influenza respiratory viruses in adults with pneumonia were conducted in populations with limited generalisability to CAP patients, such as immunocompromised or frail elderly patients, residents of long-term facilities, cases with underlying lung diseases, or critically ill people [5–11]. Therefore, the significance of non-influenza respiratory viruses in pneumonia and their specific impact on clinical outcomes has not been established in immunocompetent adult patients with CAP.
Here, we performed a national, prospective, multicentre study in order to evaluate the impact of non-influenza respiratory viruses compared with IFV on disease severity, proportion of patients achieving clinical stability within the first 3 days at the hospital, complications, intensive care unit (ICU) admission, length of hospital stay and 90-day mortality in immunocompetent adult patients with CAP.
Methods
Study setting and design
The CAP-China study, a prospective multicentre observational study, was conducted in 34 hospitals from 10 provinces of mainland China (figure 1), with all of the hospitals located in urban areas (supplementary table S1). From October 1, 2015 to June 30, 2017, patients with CAP admitted to the emergency department, general wards or ICU were screened for enrolment (supplementary figure S1). Written informed consent was obtained from all patients or their caregivers before enrolment. The CAP-China study was approved by the institutional review board of the China–Japan Friendship Hospital and was registered at ClinicalTrials.gov with identifier NCT02492425.
Patients were enrolled from five of the seven geographical locations in mainland China, including Beijing, Tianjin, Hebei and Shanxi provinces in North China; Shanghai, Shandong and Jiangsu provinces in East China; Henan province in Central China; Sichuan province in Southwest China; and Liaoning province in Northeast China.
Study population
Patients aged ≥14 years and hospitalised for CAP were enrolled on admission. CAP was diagnosed according to the 2007 Infectious Diseases Society of America (IDSA)/American Thoracic Society (ATS) guidelines [12].
Patients were excluded if one of the following criteria was met: 1) hospitalisation within the previous 90 days or enrolled in this study within the previous 30 days; 2) being immunocompromised: cancer with neutropenia (absolute neutrophil count <500 mm−3), haematological malignancies, solid malignancies receiving chemotherapy during the previous 3 months, solid organ or bone marrow transplant, active graft versus host disease, bronchiolitis obliterans, HIV infection, immunoglobulin deficiency, using immunosuppressive agents, or current treatment with systemic corticosteroids (≥20 mg prednisone per day or equivalent) for >30 continuous days before illness onset; 3) receipt of a tracheotomy, insertion of a percutaneous endoscopic gastrostomy tube or presentation of cystic fibrosis; or 4) had a clear alternative diagnosis at the end of follow-up that included lung cancer, pulmonary tuberculosis, interstitial lung disease, pulmonary embolism, pulmonary oedema or allergic bronchopulmonary aspergillosis.
2718 hospitalised CAP patients were eligible for enrolment during the study period. A total of 382 cases were excluded: refusal to consent (n=69), failure in sampling (n=158), being immunocompromised (n=37), final diagnosis of tuberculosis or non-pneumonia illness (n=58), or being lost to follow-up (n=60). Of the 2336 participants enrolled in the CAP-China study, 1421 were negative for viral detection. Finally, a total of 915 patients with positive viral detection were included in this analysis (figure 2).
Study flowchart. CAP: community-acquired pneumonia; ABPA: allergic bronchopulmonary aspergillosis.
Data collection and quality control
Before study initiation, all local principal investigators and study staff received extensive training on the protocol. Trained staff interviewed patients (or their caregivers) within 24 h after enrolment, with demographic information, comorbidities, clinical manifestations, vital signs, laboratory results and microbiology data, antimicrobial use, complications, and outcomes collected using a standardised case report form. During a 90-day follow-up period, all-cause death was recorded by telephone interview. All information was uploaded into an electronic database (www.chinapneumonia.cn) by trained clinical research coordinators and systematic data checks were performed automatically online in order to detect logical errors or extremes. For each case, 10% of the data were selected at random and checked against the original medical records by clinical research associates in order to confirm accuracy and completeness.
Specimen collection and testing
Two nasopharyngeal swabs were obtained from all patients (n=2332) immediately upon entry to the emergency department, general wards or ICU. One swab was collected for rapid influenza antigen testing on site; the other swab was placed in universal transport medium and stored at −80°C for PCR analysis in the central laboratory (National Clinical Research Center of Respiratory Diseases, China–Japan Friendship Hospital, Beijing, China). On-site urine antigen tests for Streptococcus pneumoniae and Legionella pneumophila were performed in all patients within 24 h after admission. Blood for cultures was obtained from 610 patients with a temperature ≥38.5°C within 48 h of admission, with 73.4% of them having received empirical antibiotics treatment. A total of 1229 patients had qualified sputum or endotracheal aspirate collected within 48 h after admission, with 868 of these patients having already received empirical CAP antibiotics. On-site Gram staining and bacterial cultures were performed according to routine microbiological guidance. Sputum, endotracheal aspirate and bronchoalveolar lavage fluid (BALF) samples were stored at −80°C until PCR testing in the central laboratory.
Pathogen detection
Viral aetiology was considered positive if one of the following criteria was met: 1) detection of respiratory virus in sputum, endotracheal aspirate, BALF or nasopharyngeal swabs by real-time PCR (TaqMan Array Microfluidic Cards; Applied Biosystems, Foster City, CA, USA) [13], including RSV, IFV, parainfluenza virus (PIV), HRV, CoV, human metapneumovirus (HMPV) and adenovirus (AdV); or 2) positive antigen for IFV (Clearview Exact Influenza A&B; Alere, Waltham, MA, USA).
Bacterial or atypical pathogens were considered positive if one of the following criteria was met: 1) positive bacterial culture from blood or pleural fluid; 2) positive urinary antigen for L. pneumophila (Binax Now; Trinity Biotech, Bray, Ireland) or S. pneumoniae (Binax Now; Emergo Europe, Amsterdam, The Netherlands); 3) detection of Mycoplasma pneumoniae or Chlamydia pneumoniae in sputum, BALF, endotracheal aspirate or nasopharyngeal swabs by real-time PCR (Shanghai Zhijiang Bio-Tec, Shanghai, China) [14]; 4) detection of L. pneumophila in sputum, BALF or endotracheal aspirate by real-time PCR (Shanghai Zhijiang Bio-Tec) [14]; or 5) bacteria with moderate to heavy growth (graded as >3+ growth) in qualified sputum or endotracheal aspirate, or quantified culture in BALF of ≥104 CFU·mL−1. In this study, qualified samples were defined as more than 25 leukocytes and less than 10 epithelial cells per magnified field (at ×100 magnification).
Definitions
Hypoxaemia was defined as arterial oxygen tension (PaO2)/inspiratory oxygen fraction (FIO2) <300 mmHg [15]. Sepsis was diagnosed based on the revised criteria defined by Singer et al. [16], i.e. Sequential Organ Failure Assessment score ≥2. Pneumonia severity was assessed by CURB-65 (confusion, urea >7 mmol·L−1, respiratory rate ≥30 breaths·min−1, blood pressure <90 mmHg (systolic) or ≤60 mmHg (diastolic), age ≥65 years) [17] and PaO2/FIO2. Clinical stability was defined according to the 2007 IDSA/ATS guidelines [12]. Viral and bacterial co-infection was defined as detection of both viruses and bacteria [18]. Viral and M. pneumoniae/C. pneumoniae co-infection was defined as detection of both viruses and atypical pathogens. Mixed viral infection was defined as detection of two or more viruses.
Statistical analysis
Viral CAP patients were classified into three groups: influenza viral infection group, non-influenza viral infection group and mixed viral infection group. For the influenza or non-influenza viral infection group, only one virus was detected. Clinical features, CURB-65 ≥3 and PaO2/FIO2 <200 mmHg on admission, 24 and 72 h clinical stability, hypoxaemia and sepsis during hospitalisation, ICU admission, invasive ventilation, and 90-day mortality were compared between the influenza and non-influenza viral infection groups. Further comparisons were performed in different viral infection groups (IFV, CoV, HMPV, HRV, AdV, PIV, RSV and mixed viral infection). Continuous variables are presented as median (interquartile range (IQR)) and categorical variables are presented as number (percentage). The normality of data distribution was assessed and a Mann–Whitney U-test was used for non-normally distributed variables. A Chi-squared test was used to compare categorical variables. Univariate logistic regression analysis was used to estimate the odds ratios and 95% confidence intervals of underlying diseases, clinical features, laboratory findings, severe disease, complications and mortality for the non-influenza viral infection group compared with the influenza viral infection group. A multivariable adjusted logistic regression model was used to estimate the effect of different viruses on disease severity (CURB-65 ≥3 and PaO2/FIO2 <200 mmHg on admission) and complications (sepsis and hypoxaemia during hospitalisation). Multivariable adjusted Cox regression was conducted to estimate the association between viral aetiology and 90-day mortality. All data were analysed with SPSS version 19.0 (SPSS, Chicago, IL, USA). A two-sided p<0.05 was considered statistically significant. Bonferroni correction was used for multiple comparisons.
Results
Patient characteristics
The median (IQR) age of the 915 patients with viral infection was 59.0 (36.0–72.0) years and 55.2% were male (table 1). A total of 426 (46.6%) patients had comorbidities, with cardiovascular disease the most commonly observed (31.3%), followed by diabetes (14.6%) and chronic obstructive pulmonary disease (COPD) (4.5%). The most common symptoms of patients with viral infection were cough (93.0%), fever (81.3%) and sputum (81.3%), and ∼33% of patients had dyspnoea. 151 patients had received antiviral therapy in the emergency department or on admission, among which 98.7% represented empirical treatment. Neuraminidase inhibitors were the most commonly used antivirals (86%), followed by ganciclovir (9%), acyclovir (7%) and ribavirin (2%). The median (IQR) length of hospital stay in patients with viral infection was 10 (8–14) days and the 90-day mortality was 3.1%.
Demographic characteristics, underlying diseases, clinical features and laboratory findings of hospitalised community-acquired pneumonia patients with viral infections
Among the 915 patients, influenza viral infection, non-influenza viral infection and mixed viral infection was found in 581, 240 and 94 cases, respectively. The baseline characteristics of age, symptoms, underlying diseases and laboratory findings were similar between the influenza and non-influenza viral infection groups, except that leukocytosis was more common in the non-influenza viral infection group (25.2% versus 16.0%; OR 1.77, 95% CI 1.23–2.56; p=0.002) and diffuse bilateral pulmonary infiltration was more common in the influenza viral infection group (36.8% versus 13.8%; OR 0.28, 95% CI 0.18–0.42; p<0.001) (table 1 and supplementary table S2).
Pathogen detection
Among 2336 hospitalised patients with CAP, IFV (28.4%) was the most frequently detected virus, followed by RSV (3.6%), AdV (3.3%) and CoV (3.0%); mixed viral infection was observed in 68 patients (table 2). AdV infection was found in 4.6%, 4.2%, 1.8%, 1.2% and 1.6% of patients aged 14–24, 25–44, 45–64, 65–84 and ≥85 years, respectively (p=0.002), with the positive rate of other respiratory viruses similar in different age groups (supplementary table S3). Influenza viral infection was more common in patients admitted to the ICU than patients in general wards (32.9% versus 24.3%; OR 1.52, 95% CI 1.06–2.18; p=0.023) (supplementary table S4). Typical bacteria were detected in 72 patients with viral infection, with S. pneumoniae and Klebsiella pneumoniae most commonly detected (supplementary table S5).
Patients with positive virus detection
Comparison of illness severity, complications and clinical outcomes between the influenza and non-influenza viral infection groups
Among 915 patients with viral infection, the proportions of patients with CURB-65 ≥3 (4.9% versus 3.8%; OR 0.76, 95% CI 0.35–1.64; p=0.486) and PaO2/FIO2 <200 mmHg (8.6% versus 5.1%; OR 0.57, 95% CI 0.30–1.10; p=0.092) on admission, and incidence of sepsis (40.1% versus 39.6%; OR 0.98, 95% CI 0.72–1.33; p=0.890) and hypoxaemia (40.1% versus 37.2%; OR 0.89, 95% CI 0.65–1.22; p=0.449) during hospitalisation in the influenza viral infection group did not significantly differ from those in the non-influenza viral infection group (table 3 and supplementary table S6). Additionally, there was no significant difference in ICU admission (8.3% versus 5.4%; OR 0.64, 95% CI 0.34–1.20; p=0.157), 90-day mortality (3.8% versus 1.7%; OR 0.43, 95% CI 0.15–1.26; p=0.174) and length of hospital stay (10 versus 10 days; p=0.986) between the two groups.
Disease severity, complications and outcomes in the influenza and non-influenza viral infection groups
Compared with the influenza viral infection group, the multivariable adjusted odds ratios of CURB-65 >3 and PaO2/FIO2 <200 mmHg on admission, and occurrence of sepsis and hypoxaemia during hospitalisation for the non-influenza viral infection group were 0.87 (95% CI 0.26–2.84), 0.72 (95% CI 0.26–1.98), 1.00 (95% CI 0.63–1.58) and 1.05 (95% CI 0.66–1.65), respectively. The hazard ratio of 90-day mortality was 0.51 (95% CI 0.13–1.91) (table 4). Additionally, diabetes and elevated procalcitonin (PCT) were associated with a higher proportion of CURB-65 ≥3 and sepsis (supplementary table S7).
Association of different viruses with CURB-65 ≥3 and arterial oxygen tension (PaO2)/inspiratory oxygen fraction (FIO2) <200 mmHg on admission, occurrence of hypoxaemia and sepsis during hospitalisation, and 90-day mortality in multivariable analysis#
Relationship between viral aetiology and disease severity, complications and outcomes
On admission, the proportions of patient with CURB-65 ≥3 (p=0.803) and PaO2/FIO2 <200 mmHg (p=0.476) were similar in the different viral infection groups (IFV, CoV, HMPV, HRV, AdV, PIV, RSV and two or more viruses) (figure 3). No significant difference was found in terms of demographic characteristics, underlying diseases, respiratory symptoms, C-reactive protein or PCT between the influenza viral infection group and any of the different non-influenza viral infection groups following Bonferroni correction. Higher proportions of consolidation on chest radiography (83.7% versus 50.4%; OR 5.04, 95% CI 2.31–10.98; p<0.001) but lower rates of diffuse bilateral pulmonary infiltration (4.1% versus 36.8%; OR 0.07, 95% CI 0.02–0.30; p<0.001) were found in the AdV group compared with the IFV group (supplementary table S8).
a, b) Illness severity on admission, c, d) complications during hospitalisation and e, f) outcomes of hospitalised community-acquired pneumonia patients with viral infections. CURB-65: confusion, urea >7 mmol·L−1, respiratory rate ≥30 breaths·min−1, blood pressure <90 mmHg (systolic) or ≤60 mmHg (diastolic), age ≥65 years; PaO2: arterial oxygen tension; FIO2: inspiratory oxygen fraction; IFV: influenza virus types A and B; CoV: human coronavirus; HMPV: human metapneumovirus; HRV: human rhinovirus; AdV: adenovirus; PIV: parainfluenza virus; RSV: respiratory syncytial virus. On admission, the proportions of patients with a) CURB-65 ≥3 (p=0.803) and b) PaO2/FIO2 <200 mmHg (p=0.476) were comparable in the different viral infection groups. During hospitalisation, no differences in the incidence of c) sepsis (p=0.406) and d) hypoxaemia (p=0.416) were found among the different viral infection groups. e) 90-day mortality (p=0.953) and f) length of hospitalisation (p=0.185) showed no statistically significant differences among the different viral infection groups.
Although more patients with AdV infection had a temperature >37.8°C, the proportion of patients achieving clinical stability within 24 h (p=0.205) and 72 h (p=0.230) after hospitalisation was similar irrespective of viral aetiology (figure 4).
Proportion of patients achieving clinical stability 24 and 72 h after hospitalisation and the day before discharge. IFV: influenza virus types A and B; CoV: human coronavirus; HMPV: human metapneumovirus; HRV: human rhinovirus; AdV: adenovirus; PIV: parainfluenza virus; RSV: respiratory syncytial virus. The proportions of patients that achieved clinical stability were similar irrespective of viral aetiology within 24 h (p=0.205) and 72 h (p=0.230) after hospitalisation.
During hospitalisation, the frequency of hypoxaemia was 45.1% in the influenza viral infection group and was similar among the different viral infection groups (p=0.416). Although the proportion of sepsis was lower in the RSV group than that in the IFV (24.6% versus 40.1%; OR 0.49, 95% CI 0.26–0.91; p=0.022), AdV (24.6% versus 44.6%; OR 0.40, 95% CI 0.18–0.90; p=0.025), HMPV (24.6% versus 46.4%; OR 0.38, 95% CI 0.14–0.97; p=0.036) and CoV (24.6% versus 48.3%; OR 0.43, 95% CI 0.14–0.98; p=0.042) groups, the difference was not significant following Bonferroni correction. Additionally, no significant difference in 90-day mortality (p=0.953) or length of hospital stay (p=0.185) was found between the different viral infection groups (figure 3).
Multivariable analysis showed no significant difference in the impact of different non-influenza respiratory viruses on risk of CURB-65 ≥3 or PaO2/FIO2 <200 mmHg on admission, and occurrence of sepsis or hypoxaemia during hospitalisation, and 90-day mortality was not significantly different from that of influenza viral infection, after adjustment for confounding factors (p>0.05 for all) (table 4).
Discussion
Here, we performed a large multicentre prospective observational study (conducted in 34 hospitals) in mainland China to systemically investigate the impact of non-influenza respiratory viruses in the adult immunocompetent population, relative to that of IFV. Our results indicated that the effect of non-influenza respiratory viruses on illness severity, complications and outcomes in immunocompetent patients with CAP is similar to that associated with IFV-related respiratory diseases.
In the GLIMP [19] and EPIC studies [20], viral CAP diagnosis was 37.2% in Asia and 23% in the USA, similar to our results (39.2%). Studies of adult CAP in China from 2001 and 2005 showed that bacterial aetiology (atypical and other bacteria) was much higher than viral causes (30.3–53.1% versus 10.6–19.0%) [21–25], whereas recent CAP aetiology studies from 2009 and 2016 in China revealed an increased detection of viral infections (21.1% versus 34.9%) and a decline of bacterial infections (7.8–24.8%) [14, 26–29], similar to the trends reported in European countries and the USA [30]. The use of new molecular diagnostic methods rather than paired serum antibody titres might explain the higher rate of viral detection. According to the recent epidemic surveillance data, with the exception of IFV, the primary causes of non-influenza viral pneumonia in China were AdV, RSV, CoVs, PIV, HRV and HMPV, with these viruses representing the main focus of this study [14, 26, 29, 31–34].
Non-influenza respiratory viruses accounted for nearly a third of the patients with viral pneumonia in this study. However, HRV was uncommon compared with the EPIC study, where it was the most common viral pathogen [20]. According to recent research and meta-analysis, HRV was the most common virus isolated from patients with COPD, with detection in 23.0% of exacerbated cases [35, 36]. However, only 5.1% (120 out of 2336) of the patients in our cohort had COPD compared with 42% (968 out of 2320) in the EPIC study [20] and 26% (759 out of 3149) in the GLIMP study [19]. This difference might be due to the younger age (median 59.0 versus 69.0 years) and fewer current/ex-smokers (31.3% versus 42.3%) among our patients [19]. Additionally, Radovanovic et al. [19] found a higher prevalence of HRV in North America relative to other continents, including Asia, Europe, Africa and Oceania (26.7% versus 2.8%; p<0.001), suggesting that climate, geographical position and local epidemiological trends might also influence the prevalence of viruses.
Consistent with previous studies, we observed similar clinical features among the different viral infection groups [37]. Although diffuse bilateral infiltrates were more common in IFV infection cases, further analysis showed that the difference only remained significant between IFV and AdV infection groups following Bonferroni correction. As previously reported, AdV infection was more commonly observed in young adults and more frequently appeared as consolidation than other viruses [38, 39]. The high rate of elevated lactate dehydrogenase associated with AdV infection might reflect the possibility of hepatitis [40]. Moreover, we found a high proportion of hypoxaemia associated with HMPV and HRV infection. However, none of the clinical factors was specific enough to reliably discriminate between pneumonia caused by IFV and other viral pathogens.
Our data showed that the severity of pneumonia caused by non-influenza respiratory viruses was similar to that caused by IFV, which agrees with reports by Gilca et al. [41] and Bjarnason et al. [42]. Previous studies had reported that the spread of virus into the bronchial epithelium could induce respiratory airway injury and release of cytokines. Although the induced cytokine profiles might vary by virus, they converge on a common end pathway and result in similar diffuse alveolar damage [43, 44]. In the present study, we found that comorbidities but not the type of viruses were independently associated with sepsis and hypoxaemia in patients with viral pneumonia. For elderly hospitalised patients with respiratory symptoms, RSV, HMPV and PIV were even associated with a higher mortality and more complications compared with IFV [45–49].
The impact of bacterial co-infection on disease severity and mortality had been reported in patients with viral infections [18, 49, 50]. However, detecting all of the concomitant bacterial infection was difficult due to the limitations in traditional methods, the few molecular diagnostic tests approved for bacteria identification [51] and the high rate of pre-sampling antibiotic use. According to a previous study, a PCT threshold of ≥0.5 ng·mL−1 resulted in a specificity of 72.5% for bacterial identification [52]. Additionally, fever, leukocytosis and lobar consolidation have also been used to predict bacterial pneumonia [30]. After adjusting for viral–bacterial co-detection, elevated PCT, leukocytosis, fever and lobar consolidation in multivariate analysis, the non-influenza viral infection group was still confirmed to be comparable with the influenza viral infection group regarding disease severity, complications and mortality.
Our study offers critical insights into and recognition of non-influenza respiratory viruses among patients requiring admission to the hospital and potentially the ICU. In some randomised controlled trials and observational studies, rapid recognition of viruses was not associated with reducing the proportion of antibiotic use, early de-escalation or decreased hospital stay [53–58]. However, the duration of antibiotic use was reduced in patients with influenza-like illness, asthma and infective exacerbation of COPD [53, 55]. We recently completed a randomised controlled trial study (ClinicalTrials.gov identifier NCT03391076) which demonstrated that timely virus detection via PCR contributes to a decreased duration of antibiotics treatment in hospitalised CAP patients [59]. Furthermore, success has been achieved in patients with a decreased antibiotic prescription rate along with increased antiviral use following detection of IFV by rapid point-of-care testing [60]. Recognising the role of non-influenza respiratory viruses will promote the rapid development of antivirals, along the same lines as anti-influenza drug development (e.g. balozavir [61], pimodivir [62] and favipiravi [63]). Additionally, more novel antivirals are currently undergoing pre-clinical trials, including DAS181 for influenza [64], HMPV [65] and PIV [66], GS5806, ALS-8176 and AK0529 for RSV [67], and cidofovir for AdV [68]. Timely diagnosis of viral aetiology might help clinicians to optimise the use of antivirals and antibiotics in the era of targeted therapy.
There were some limitations in our study. First, viruses identified from nasopharyngeal swabs cannot easily differentiate lower and upper respiratory tract infections. In this study, only patients with radiologically confirmed pneumonia were enrolled and those with only upper respiratory tract infection had been excluded. Second, bacterial aetiology (atypical and other bacteria) was lower than that reported in previous studies (30.3–53.1%) [21–25] in China from 2001 to 2005 [69]. This might be explained by increased pre-sampling antibiotic use, the relatively lower prevalence of comorbidities and less severe disease, resulting in lower bacterial detection in our study. Third, the PCR test was not used to detect typical bacterial pathogens, as it has not been recommended by the IDSA/American Society for Microbiology guidelines [51]. Fourth, we did not include immunosuppressed patients in this study and therefore our results cannot be generalised to that patient subset.
Conclusions
These findings suggest that complications were common in patients with non-influenza viral pneumonia and the impact of non-influenza respiratory viruses on clinical outcomes was comparable with that of IFV in immunocompetent adult patients with CAP. The increasing recognition of the significant role of non-influenza respiratory viruses might promote the development of CAP management procedures for immunocompetent adult patients.
Supplementary material
Supplementary Material
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
ERJ-02406-2018_SUPPLEMENTARY TABLES ERJ-02406-2018_SUPPLEMENTARY_TABLES
Acknowledgements
We would like to thank all the physicians from the 34 hospitals who participated in CAP-China.
Footnotes
This article has supplementary material available from erj.ersjournals.com
Author contributions: B. Cao, F. Zhou, Y. Wang, L. Gu, Y. Liu, B. Liu and C. Wang conceived of the project. B. Cao and F. Zhou performed the analysis and drafted the paper, and all authors critically revised the manuscript for important intellectual content and gave final approval for the version to be published. All authors agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. B. Cao takes full responsibility for the data and analysis.
Conflict of interest: F. Zhou has nothing to declare.
Conflict of interest: Y. Wang has nothing to declare.
Conflict of interest: Y. Liu has nothing to declare.
Conflict of interest: X. Liu has nothing to declare.
Conflict of interest: L. Gu has nothing to declare
Conflict of interest: X. Zhang has nothing to declare.
Conflict of interest: Z. Pu has nothing to declare.
Conflict of interest: G. Yang has nothing to declare.
Conflict of interest: B. Liu has nothing to declare.
Conflict of interest: Q. Nie has nothing to declare.
Conflict of interest: B. Xue has nothing to declare.
Conflict of interest: J. Feng has nothing to declare.
Conflict of interest: Q. Guo has nothing to declare.
Conflict of interest: J. Liu has nothing to declare.
Conflict of interest: H. Fan has nothing to declare.
Conflict of interest: J. Chen has nothing to declare.
Conflict of interest: Y. Zhang has nothing to declare.
Conflict of interest: Z. Xu has nothing to declare.
Conflict of interest: M. Pang has nothing to declare.
Conflict of interest: Y. Chen has nothing to declare.
Conflict of interest: X. Nie has nothing to declare.
Conflict of interest: Z. Cai has nothing to declare.
Conflict of interest: J. Xu has nothing to declare.
Conflict of interest: K. Peng has nothing to declare.
Conflict of interest: X. Li has nothing to declare.
Conflict of interest: P. Xiang has nothing to declare.
Conflict of interest: Z. Zhang has nothing to declare.
Conflict of interest: S. Jiang has nothing to declare.
Conflict of interest: X. Su has nothing to declare.
Conflict of interest: J. Zhang has nothing to declare.
Conflict of interest: Y. Li has nothing to declare.
Conflict of interest: X. Jin has nothing to declare.
Conflict of interest: R. Jiang has nothing to declare.
Conflict of interest: J. Dong has nothing to declare.
Conflict of interest: Y. Song has nothing to declare.
Conflict of interest: H. Zhou has nothing to declare.
Conflict of interest: C. Wang has nothing to declare.
Conflict of interest: B. Cao has nothing to declare.
Support statement: This study was supported by grants from the National Science Fund for Distinguished Young Scholars (81425001/H0104), the Innovation Fund for Medical Sciences from the Chinese Academy of Medical Sciences (2018-I2M-1-003), the National Key Technology Support Program from the Ministry of Science and Technology (2015BAI12B11), and the Beijing Science and Technology Project (D151100002115004). Funding information for this article has been deposited with the Crossref Funder Registry.
- Received December 19, 2018.
- Accepted May 2, 2019.
- Copyright ©ERS 2019
References
