Comparing viral, bacterial, and coinfections in community-acquired pneumonia, a retrospective cohort study.

OBJECTIVES

Despite the substantial rates of viral etiology in community-acquired pneumonia (CAP), empirical antibiotic therapy is regularly administered. This study compared the clinical presentation, antibiotic use, and outcomes of patients based on the identified causative pathogens.

METHODS

Patients that were immunocompetent and had CAP from the international, multicenter prospective cohort study on CAP (CAPNETZ) from 2007 to 2017 with available multiplex polymerase chain reaction testing for bacterial and viral pathogens from sputum were included. Patients were divided into four groups based on the detection of bacterial pathogens, viral pathogens, bacterial/viral coinfection, and no pathogen detected. Differences were analyzed using multivariate linear regression.

RESULTS

Patients with bacterial CAP were significantly younger (median age 60 years; adjusted odds ratio [aOR] 0.96 [0.94-0.98]) and reported less smoking (aOR 0.98 [0.97-1.0]). A higher CRB-65 score (confusion, respiratory rate, blood pressure, and age ≥65years) was associated with bacterial CAP (aOR 1.69 [1.1-2.58]). Bacterial CAP showed higher 180-day mortality (aOR 3.59 [1.09-11.8]) and viral CAP had higher 30-day mortality (aOR 15.79 [1.04-238.75]) than the other groups. Upon admission, the four groups could not be distinguished based on clinical presentation and showed no differences in CAP-related complications, length of hospital stay, or application/duration of antibiotic treatment (98.4% received antibiotics).

CONCLUSIONS

A pathogen-centered treatment algorithm for patients with CAP is required to avoid unnecessary antibiotic therapies, side effects and optimize patient outcomes and long-term morbidity.