key: cord-0965254-evpg537d
authors: Takazono, Takahiro; Mukae, Hiroshi; Izumikawa, Koichi; Kakeya, Hiroshi; Ishida, Tadashi; Hasegawa, Naoki; Yokoyama, Akihito
title: Empirical antibiotic usage and bacterial superinfections in patients with COVID-19 in Japan: a nationwide survey by the Japanese Respiratory Society
date: 2021-10-07
journal: Respir Investig
DOI: 10.1016/j.resinv.2021.09.004
sha: 378757169f103801e933d2fea0d9cc70e880cd51
doc_id: 965254
cord_uid: evpg537d

An internet questionnaire survey for investigating empirical antibiotic usage and bacterial superinfections in patients with coronavirus disease-2019 (COVID-19) in Japan was conducted among the chief physicians of respiratory disease departments of 715 Japanese Respiratory Society-certified hospitals using Google Forms between January 28, 2021 and February 28, 2021. Responses to the questionnaire survey were obtained from 198 of 715 hospitals (27.6%). The survey revealed that the complication incidences of community-acquired pneumonia; hospital-acquired pneumonia, including ventilator-associated pneumonia; and sepsis were 2.86, 5.59, and 0.99%, respectively, among patients with moderate/severe and critical COVID-19. Bacterial co-infection and secondary infection rarely affected patients with COVID-19 in Japan, and the isolated pathogens were not specific to these patients. Moreover, the anti-inflammatory effects of macrolides for COVID-19 were not observed in several studies. These results might be useful in clinical practice for COVID-19.

The worldwide spread of the coronavirus disease-2019 (COVID-19), caused by the severe acute respiratory syndrome coronavirus 2, has resulted in more than four million deaths as of July 2021 [1] . Bacterial infections complicate viral airway infections, and those secondary to severe influenza virus infections are fatal. Therefore, it is crucial to focus on addressing secondary infection or co-infection in patients with COVID-19.

Moreover, it is essential to investigate their epidemiology in patients with COVID-19 to practice antibiotic stewardship and reduce unnecessary antibiotic prescriptions. Several studies have reported low incidence rates [2, 3] . However, nationwide large-scale data are limited. Therefore, we conducted a nationwide internet questionnaire surveillance study in Japan to clarify empiric antibiotic usage and incidences of bacterial pneumonia and sepsis, complicating COVID-19.

An internet questionnaire survey was conducted among the chief physicians of the respiratory disease departments of 715 Japanese Respiratory Society-certified hospitals using Google Forms between January 28, 2021 and February 28, 2021. The questionnaire included the total number of COVID-19 cases diagnosed, based on reverse transcription-polymerase chain reaction testing in each hospital from the beginning of the COVID-19 pandemic to January 27, 2021. The questionnaire included COVID-19 severity according to the National Institute of Health treatment guidelines [4] , use of empirical antibiotic treatment, type and duration of empirical antibiotic treatment, and complications of community-acquired pneumonia (CAP), hospitalacquired pneumonia (HAP) excluding ventilator-associated pneumonia (VAP), and J o u r n a l P r e -p r o o f sepsis. The contents of the questionnaires related to this study are summarized in supporting information. Informed written consent was not required because the study utilized anonymized patient data and was approved by the Japanese Respiratory Society.

Responses to the questionnaire survey were obtained from 198 of 715 hospitals Tables 1 and 2. This report investigated empirical antibiotic usage and frequency of bacterial superinfections in patients with COVID-19 in Japan. Empirical usage rate was as high as 51% of the facilities, including for the answer option "if patients were in a severe condition due to COVID-19." Antibiotic usage rate was not as high as those in previous studies by Rawson et al. [2] and Kabara et al. [3] , in which the rates of antibiotic usage were 72% and 69%, respectively. Since the present study was conducted a couple months after these previous studies, the physicians may have already been aware that the co-infection rate was not as high as expected in the early phase of the COVID-19 pandemic. Macrolides were the third most frequently used antibiotic in our survey, and their anti-inflammatory effects have been utilized for COVID-19 treatment worldwide.

However, azithromycin failed to improve survival rate or clinical outcomes in hospitalized patients with COVID-19 [5] and did not alleviate symptoms in outpatients with COVID-19 [6] . Therefore, empirical macrolide administrations should be avoided unless pneumonia due to atypical pathogens, such as Mycoplasma pneumoniae and Legionella spp., is highly suspected. Moreover, Petit et al. reported that reducing empiric antibiotic therapy including administration period did not affect the outcome of COVID-19 in a single-center study [7] .

The complication rate of CAP was only 2.86% in patients with moderate/severe and critical COVID-19 (336/11,714). This did not differ from the result of a previous metaanalysis, which reported a complication rate of 3.5% (95% confidence interval [CI] 0.4-6.7%) [8] . Nosocomial pneumonia, including HAP and VAP, was observed in 5.59%

(655/11,711), and sepsis was observed in 0.99% (116/11,711) . The isolated bacterial pathogens in our survey were identical to those of previous studies on patients with COVID-19, which were common pathogens in CAP and HAP [9, 10] . The secondary J o u r n a l P r e -p r o o f infection rate of these two complications was 6.58% (771/11,711), but overlapping cases were possibly included. Our data indicated a lower complication rate compared to 14.3% (95% CI, 9.6-18.9%), obtained from a previous meta-analysis [8] . The pathogens isolated from patients with sepsis in our survey were the typical pathogens of catheter-related bloodstream infections.

Our study had several limitations. First, the available data were limited to the nature of the questionnaire-based study, and the accuracy of data is lower compared to observational studies. Second, since the response rate from the survey was relatively low (27.6%) and biased toward referral centers, the results of this study might not be generalized. Third, the diagnosis of pneumonia was based on the judgment of each physician, rather than the defined diagnostic criteria. Fourth, our survey data were obtained from an internet questionnaire survey. Therefore, direct comparisons with previous observational studies are not applicable. J o u r n a l P r e -p r o o f 

Bacterial and fungal coinfection in individuals with coronavirus: A rapid review to support COVID-19 antimicrobial prescribing

Prevalence of co-infection at the time of hospital admission in COVID-19 patients, A multicenter study

NIH. COVID-19 treatment guidelines

Azithromycin in patients admitted to hospital with COVID-19 (RECOVERY): A randomised, controlled, open-label, platform trial

Effect of oral azithromycin vs placebo on COVID-19 symptoms in outpatients with SARS-CoV-2 infection: A randomized clinical trial

Reducing the use of empiric antibiotic therapy in COVID-19 on hospital admission

Bacterial co-infection and secondary infection in patients with COVID-19: A living rapid review and meta-analysis

Treatment of community-acquired pneumonia during the coronavirus Disease 2019 (COVID-19) pandemic

Coinfections in patients hospitalized with COVID-19: A descriptive study from the United Arab Emirates

We appreciate the Japanese Respiratory Society-certified hospitals that participated in this survey.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. All authors are required to disclose any COI within 3 calendar years preceding the current year, prior to the submission of any manuscript in the subject matter of which any company, entity, or organization has an interest.

No If Yes: List the name(s) of authors and commercial entity(ies) and use as much space as necessary This statement will be kept for 2 years after the publication of the manuscript.