Severity of disease and mortality for hospitalized patients with community-acquired viral pneumonia compared to patients with community-acquired bacterial pneumonia

Background: There exists a large body of literature to help identify, diagnose, treat, and manage community-acquired pneumonia (CAP). Despite this, there is little data that directly compares the clinical syndromes and complications of pure bacterial pneumonia to pure viral pneumonia. Our study compares the clinical presentation, morbidity and mortality of viral vs. bacterial etiologies of CAP. Methods: This was a secondary data analysis of the Community-Acquired Pneumonia Organization (CAPO) international study database. Data was collected concerning patient demographics, physical examination findings, laboratory findings, radiological findings, severity of illness, and clinical outcomes and stratified according to the two study groups, CAVP and CABP. A microbiological diagnosis of CABP was based on the isolation of a bacterium from a respiratory sample, blood culture and/or identification of a urinary antigen for Streptococcus or Legionella; microbiological diagnosis of CAVP was based on polymerase chain reaction or antigen detection from respiratory samples. Results: Our study included 1,913 patients. Of these, 286 (15.0%) had viral infection, while 1,627 (85.0%) had CAVP. We found that bacterial CAP patients are older, more frequently male, and suffer from a higher proportion of comorbidities when compared to viral CAP patients. Comparison of physical exam findings and laboratory values failed to find a clinically significant difference between bacterial and viral CAP patients. When comparing severity of illness, bacterial CAP patients had greater frequency of PSI ≥ class IV; however, viral CAP patients more frequently needed ICU admission, ventilator support, vasopressor support, and had higher rate of in hospital mortality. Conclusions: Our study confirms the extreme difficulty differentiating CABP from CAVP using demographics, physical exam, or x-ray findings. We found no major clinical or laboratory findings distinguishing CABP from CAVP. The increased severity of illness of CAVP compared to bacterial etiologies shows that PSI scores may not be an accurate indicator of severity of disease. More studies are needed to identify the best process of care for patients with CAP, including the potential benefits of routine respiratory viral panel testing and empiric antiviral therapy. DOI: 10.18297/jri/vol3/iss1/3 Received Date: October 26, 2018 Accepted Date: December 18, 2018 https://ir.library.louisville.edu/jri/vol3/iss1/ Affiliations: 1Division of Pulmonary, Critical Care and Sleep Disorders, University of Louisville 2Division of Infectious Diseases, University of Louisville 3Saint Louis University Center for Health Outcomes Research (SLUCOR) This original article is brought to you for free and open access by ThinkIR: The University of Louisville’s Institutional Repository. It has been accepted for inclusion in The University of Louisville Journal of Respiratory Infections by an authorized editor of ThinkIR. For more information, please contact thinkir@louisville. Recommended Citation: Kim, Richard Y.; Chandler, Thomas; Furmanek, Stephen P.; Wiemken, Timothy Lee; and Cavallazzi, Rodrigo (2019) “Severity of disease and mortality for hospitalized patients with community-acquired viral pneumonia compared to patients with community-acquired bacterial pneumonia,” The University of Louisville Journal of Respiratory Infections: Vol. 3 : Iss. 1, Article 3. *Correspondence To: Richard Y Kim, MD Work Address: 401 E Chestnut Street, Suite 310 Louisville, KY, United States, 40202 Work Email: richard.kim@louisville.edu ORIGINAL RESEARCH Copyright: © 2019 The author(s). This is an open access article distributed under the terms of the Creative Commons Attribution 4.0 International License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.


Introduction
Pneumonia is an extremely common, yet exceedingly dangerous condition that makes up 423,000 emergency room visits and is the 8th leading cause of death in the United States, according to the CDC [1].Inpatient community-acquired pneumonia (CAP) treatment is often empiric and laboratory testing may not result in a definite organism [2].It has been demonstrated that the microbial etiology of roughly 30%-60% of CAPs treated in the inpatient setting remains unidentified [3][4][5][6].Of those admitted to the hospital, 10-20% of patients are admitted to the ICU [7].While imaging and prognosticating tools like the CURB-65 and Pneumonia Severity Index assist in the diagnosis and severity assessment of pneumonia, distinguishing between viral and bacterial pneumonia remains a challenge [8].
Bacterial pneumonia is a major cause of pathogen identified CAP.For example, Streptococcus pneumoniae is the most common cause of bacterial pneumonia, causing significant morbidity and mortality [7].Tools like procalcitonin levels are helpful in the diagnosis and management of bacterial pneumonia [9].On the other hand, viral pneumonia, a significant contributor to the prevalence of CAP, is identified by polymerase chain reaction (PCR) in more than 25% of cases using nasopharyngeal swabs and up to 40% when using lower respiratory tract samples [10].Tools like rapid antigen testing and real time PCR assist in the diagnosis of several viral pneumonias [11].Furthermore, morbidity and mortality of bacterial pneumonia after a preceding viral infection increases [12].
There is a large body of literature to help physicians identify, diagnose, treat, and manage CAP.Despite this, there is little data that directly compares the clinical syndromes and complications of bacterial pneumonia to viral pneumonia.Our study aims to compare the clinical presentation, morbidity and mortality of purely viral to purely bacterial etiologies of CAP.

Study Design
This was a secondary data analysis of the Community-Acquired Pneumonia Organization (CAPO) international study database.This multinational database is coordinated by the University of Louisville School of Medicine, Department of Medicine, Division of Infectious Diseases.Investigators from 130 hospitals across 30 countries perform data collection designed by the University of Louisville.Data is electronically transferred and validated by research associates at the University of Louisville [13].

Subjects
Patients were eligible for inclusion in analysis if they were hospitalized with CAP and a microbiological diagnosis of a viral or bacterial infection was established.Patients were enrolled from 2001 to 2017 and categorized into two groups: those with confirmed bacterial infection and those with confirmed viral infection.Patients without an identified bacterial or viral organism were excluded from the study.Additionally, patients with any coinfection were excluded.

Community-acquired pneumonia (CAP):
A patient was defined as having CAP when the following 3 criteria were met: 1) presence of a new pulmonary infiltrate on chest radiograph and/or chest computed tomography scan at the time of hospitalization, defined by a board-certified radiologist's reading; 2) at least 1 of the following: a) new cough or increased cough or sputum production, b) fever >37.8°C (100.0°F) or hypothermia <35.6°C (96.0°F), c) changes in leukocyte count (leukocytosis: >11000 cells/μL; left shift: >10% band forms/mL; or leukopenia: <4000 cells/μL); and 3) no alternative diagnosis at the time of hospital discharge that justified the presence of criteria 1 and 2 [14].

Hospitalization with community-acquired bacterial pneumonia (CABP) vs. community-acquired viral pneumonia (CAVP):
A microbiological diagnosis of CABP was based on the isolation of a bacterium from a respiratory sample, blood culture and/ or identification of a urinary antigen for Streptococcus or Legionella; microbiological diagnosis of CAVP was based on polymerase chain reaction or antigen detection from respiratory samples.

Coinfection:
A patient was defined as having coinfection if more than one microorganism was identified.All patients with coinfections were excluded from analysis.

Measurements
Data was collected concerning patient demographics, physical examination findings, laboratory findings, radiological findings, severity of illness, and clinical outcomes and stratified according to the two study groups, CAVP and CABP.

Statistical Analysis
Descriptive statistics were performed, with frequencies with percentages as well as medians with interquartile ranges reported for categorical and continuous variables, respectively.Chi-squared tests and Mann-Whitney U-tests were performed to compare categorical and continuous variables.P-values of less than 0.05 were considered statistically significant.R v 3.4.3was used for all analyses [15].

Physical Examination Findings
Comparison between vitals measurements are shown in Table 3.While many variables were statistically significant, the differences between the CABP and CAVP groups were not considered clinically significant.Other differences were clinically or statistically non-significant.

Severity of Illness
Severity of disease on admission was characterized by frequency of pneumonia severity index (PSI) class IV or greater, need for ICU admission, altered mental status, need for ventilator support, and need for vasopressors, as shown in

Discussion
Using the international CAPO database, we were able to retrospectively evaluate 1,913 patients admitted to the hospital for CAP to compare and contrast the clinical presentation, morbidity, and mortality of CAVP and CABP.Demographically, when compared to viral CAP, we found that bacterial etiologies are more commonly found in males and older patients.Bacterial pneumonia also is more frequent than viral pneumonia in patients with chronic obstructive pulmonary disease, congestive heart failure, human immunodeficiency virus, stroke, neoplastic disease, and those on home oxygen.Physical exam characteristics were found to be clinically non-significant between both groups.Furthermore, the blood urea nitrogen level, a nonspecific marker of plasma volume status, and frequency of PSI scores ≥4 were found to be higher in patients with CABP.However, the overlapping similarity in findings when comparing physical exam, laboratory data, and radiographic findings is in line with the extreme difficulty in diagnosing between bacterial vs. viral pneumonias.This is important to address as CAVP patients were observed to have a greater frequency of overall disease severity when comparing the frequency of need for ICU admission, ventilator support, vasopressor therapy, and mortality.Our findings are both supported and refuted by currently established literature, reiterating the ambiguity between and difficulty in distinguishing the clinical presentation and severity of illness of bacterial vs. viral etiologies of CAP.For example, Johnstone found that viral pneumonias, in comparison to nonviral causes of pneumonia, were found to be more prevalent in older and frailer patients [16].Other studies have identified a similar age difference between the two groups [6,17].This is contrary to our finding that CABP patients are older than CAVP patients.Also, patients with COPD, CHF, and cerebrovascular disease were also found to be more commonly affected by viral CAP than bacterial CAP [17].Similar to our findings, on the other hand, malignancy was observed to be more commonly found in patients with bacterial etiologies than viral [17].HIV was also observed to be more common in bacterial CAP than viral CAP [18].Despite these findings, it is important to also note that multiple studies found no significant difference in age or comorbidities between bacterial and viral etiologies of CAP [19][20][21][22].
Medical literature also supports our findings that physical exam and laboratory values are often unable to definitively distinguish between bacterial vs. viral pneumonias [17].We did not find drastic differences between the two groups when comparing physical exam findings.These minor, yet significant differences were seen when comparing respiratory rate, blood pressure, and degree of tachycardia.We found higher BUN levels in patients suffering from CABP compared to CAVP.Higher levels of BUN have been shown to be associated with increased mortality in patients with CAP [23,24], however there were no studies found comparing BUN levels in CAVP vs. CABP.While we observed a significant difference in serum procalcitonin, the large proportion of patients missing those values should cause caution when interpreting these results.Though we observed significant, yet minor differences in serum glucose, hematocrit, and bicarbonate values between CABP and CAVP, Johnstone went on to state, "it seems unlikely that any constellation of symptoms, signs, and routine laboratory findings will ever reliably differentiate between the presence or absence of a virus.16Furthermore, though there are several studies showing the clinical and laboratory findings of pure bacterial, pure viral, and combined bacterial-viral CAP infections [16,20,25], diagnosing and differentiating purely viral from bacterial CAP remains a problem as the etiology for a significant proportion of CAP remains unknown.It is estimated that 40-60% of CAP remains unidentified [3,26,27].Caglayan states that bacterial CAP infections resemble combined bacterial-viral CAP in terms of mean age, immune status, leukocyte count, C-reactive protein (CRP) values, hospitalization duration and CURB-65 score [25].
We also evaluated disease severity in the setting of purely viral vs. purely bacterial CAP.CABP more frequently had PSI ≥ class IV, however CAVP significantly showed higher frequencies of ICU admission, intubation, vasopressor support, and in hospital mortality.Though it has been previously demonstrated that PSI ≥ class IV indicates increased disease severity and is strongly associated with ICU admission [28,29], our results show that PSI scores cannot be used to accurately prognosticate viral CAPs.Furthermore, literature suggests that PSI often underestimates the risk of patients with Influenza A H1N1 pneumonia [30,31] and neither PSI nor CURB-65 scores can be used to predict ICU admission or need for mechanical ventilation in influenza patients31.The CDC recommends early antiviral therapy for patients who are suspected of suffering from influenza [32].
There exist limitations to our study.First, this was a retrospective study.Secondly, while the multicenter and international nature of the study increases the strength and generalizability of results, the data collection and other differences in process of care provide an unmeasurable confounding element that may have significant impact on data collection.Also, a large portion of patients included were afflicted by the H1N1 pandemic in 2009, that may explain the increased mortality for CAVP.Patients affected by H1N1 are observed to be younger in age and with fewer comorbidities [33].Furthermore, pathogens identified from respiratory samples may represent colonization or active infection.
In conclusion, our study confirms the extreme difficulty differentiating CABP from CAVP using demographics, physical exam, or x-ray findings.We found no major clinical or laboratory findings distinguishing CABP from CAVP.The increased severity of illness of CAVP compared to bacterial etiologies shows that PSI scores may not be an accurate indicator of severity of disease.More studies are needed to identify the best process of care for patients with CAP, including the potential benefits of routine respiratory viral panel testing and empiric antiviral therapy.

Table 1
Most common pathogens isolated in CAVP

Table 2
Most common pathogens isolated in CABP 3. Laboratory FindingsComparison of laboratory values are shown in Table3.Largest difference between the two groups was observed in the blood urea nitrogen (BUN), with bacterial pneumonia having a higher BUN value compared to that found in viral pneumonia(30 [IQR

Table 4 .
There was a greater frequency of PSI ≥ class IV in CABP patients than

Table 3
Patient demographics, physical examination findings, laboratory findings, and radiological findings

Table 4
Severity of illness and clinical outcomes