Viral community acquired pneumonia at the emergency department: Report from the pre COVID‐19 age

Abstract The role of viruses in community acquired pneumonia (CAP) has been largely underestimated in the pre‐coronavirus disease 2019 age. However, during flu seasonal early identification of viral infection in CAP is crucial to guide treatment and in‐hospital management. Though recommended, the routine use of nasopharyngeal swab (NPS) to detect viral infection has been poorly scaled‐up, especially in the emergency department (ED). This study sought to assess the prevalence and associated clinical outcomes of viral infections in patients with CAP during peak flu season. In this retrospective, observational study adults presenting at the ED of our hospital (Rome, Italy) with CAP from January 15th to February 22th, 2019 were enrolled. Each patient was tested on admission with Influenza rapid test and real time multiplex assay. Seventy five consecutive patients were enrolled. 30.7% (n = 23) tested positive for viral infection. Of these, 52.1% (n = 12) were H1N1/FluA. 10 patients had multiple virus co‐infections. CAP with viral infection did not differ for any demographic, clinic and laboratory features by the exception of CCI and CURB‐65. All intra‐ED deaths and mechanical ventilations were recorded among CAP with viral infection. Testing only patients with CURB‐65 score ≥2, 10 out of 12 cases of H1N1/FluA would have been detected saving up to 40% tests. Viral infection occurred in one‐third of CAP during flu seasonal peak 2019. Since not otherwise distinguishable, NPS is so far the only reliable mean to identify CAP with viral infection. Testing only patients with moderate/severe CAP significantly minimize the number of tests.


| INTRODUCTION
The role of viruses in lower respiratory tract illness has been long time underestimated especially in the emergency setting, where the priority is generally to focus on identifying critical patients and to rapidly start empirical antimicrobial therapy. 1 However, influenza-like illness (ILI) pathogens and seasonal influenza viruses causes significant morbidity and mortality worldwide each year, and their identification in patients admitted with community acquired pneumonia (CAP) has multiple implications on inhospital patient management. 2,3 Moreover, it has been postulated that intercurrent viral respiratory infections are able to modulate ACE2 receptors leading to upper airway mucosal damage and local immune impairment. 3,4 Therefore, ILI (mainly caused by influenza viruses, parainfluenza virus and respiratory syncytial virus) could represent a predisposing factor for subsequent severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection. 4,5 Finally, the identification of the causal pathogen of CAP is even more crucial now that the clinical and radiological differential diagnosis between coronavirus diasease 2019 (COVID- 19) and no-SARS-CoV-2 viral pneumonia is controversial. 6 For this reason, we reviewed the cases of CAP admitted at the emergency department (ED) of a large university hospital in a period of epidemiological peak for ILI and influenza. The primary aim of our study was to assess the prevalence of viral infections in patients with

| Study design
This is a retrospective, single center, observational study involving patients with CAP attending the adult ED of a 1200-bed in-town teaching Hospital (Rome, Italy) with a catchment area of 600,000-1,200,000 people. 7 The study period was set between January 15th and February 22th, 2019, during the peak of 2018-19 flu season. The Italian nationwide sentinel surveillance network (InfluNet) reported for 2018-19 influenza season: (1) 8 million cases of ILI in Italy (incidence of 13.6%), (2) a more significant co-circulation of influenza A(H1N1) pdm09 and A(H3N2) virus subtypes, (3) an influenza vaccination coverage in elderly population equal to 53.1%. 8

| Eligibility criteria and definitions
All adult (>18 years) admitted at the ED with a definitive diagnosis of CAP were selected and included in the study. Patients admitted to hospital for 48 h or more in the 90 days before this presentation were excluded. 9 CAP definition was compliant with current American Thoracic Society and Infectious Diseases Society of America guideline. 10

| Data collection
Demographic, anamnestic and clinical data, laboratory results (routine bloods analyses, including baseline arterial blood gas and Creactive protein), and radiological features (chest X-Ray/computrd tomography scan) along with pneumonia severity scores (CURB-65 and qSOFA) were obtained from ED patients' files. Nasopharyngeal swab (NPS) test results collected and analyzed at the time of ED hospitalization were obtained from the virology laboratory database. Bacteriological examination of the sputum samples was not available for all patients and therefore was not considered in this study.

| Statistical analysis
For categorical variables, either Pearson's χ 2 or Fisher's exact test were used to test the statistical difference in proportion between two or three independent groups. The level of agreement between tests was determined using Cohen's κ coefficient. Description of median with interquartile range (25%-75%), mean and SD, simple frequencies (n), proportions and rates of the given data on each variable was calculated. All data were analyzed using Statistical Package for Social Science version 20.

| Ethical considerations
Data exposed in this study were previously collected for diagnostic and clinical purposes by the medical staff of ED and the virology laboratory. The study was carried out in accordance with the Helsinki Declaration and data were collected and analyzed after receiving patients' informed consent. Ethical approval was not required since the study was based on data routinely collected and stored anonymized according to the Italian law on privacy. 11 3 | RESULTS

| Overall study population
The flow chart providing an overview of patient enrollment is showed in Figure 1. Overall, out of 157 patients with CAP presenting at the ED during the study period, only 75 patients met eligibility criteria and were eventually included in the study analysis. Demographic and clinical data of the study population along with outcomes are listed in Table 1.  Table 1). 86.6% (n = 65) of patients were eventually admitted on ward level whereas one was directly referred to ICU; ED intra-mortality rate was 1.3% (n = 1).

| Prevalence of viral infections in patients with CAP and etiology
Although not statistically relevant, all mechanical ventilation (MV) and intra-ED deaths were recorded into the group of CAP with viral infection. Notably, the only patient who died in the ED had received mechanical ventilation before.

| Impact of universal influenza test execution on ED activity
According to internal protocol, patient waiting for influenza rapid test' results were isolated in dedicated spaces and asked to wear a surgical mask as a precautionary measure. Considering that (1) the results of rapid test were available in 30-40 min, (2) only 16% of patients (those affected by influenza) required respiratory isolation,

| DISCUSSION
Despite traditionally underestimated, especially in the ED setting, more recent pre-COVID-19 evidence suggested that viruses play a relevant role in CAP etiology. [12][13][14][15][16] The Etiology of Pneumonia in the Community study was a large prospective US based surveillance study in which one or more viruses were detected in 26% of CAP requiring hospitalization. 14 [17][18][19][20] Accordingly, no significant differences were found in our study from the comparison of CAP with or without viral infection and with or without Influenza. Therefore, NPS molecular testing is so far the only reliable mean to detect viral infections.
Although not statistically significant, in our study all intra-ED deaths and MV were recorded into the group of CAP with viral infection. This data is in line with previous evidence underling the relation among viral infection and pneumonia severity.
T A B L E 1 Study population: Demographic and clinical data along with outcomes of overall study population and groups