If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Corresponding author. Address: Institute of Microbiology and Immunology, Medical Faculty, University of Belgrade, Dr Subotica 1, 11000 Belgrade, Serbia. Tel.: +381 113643 3373.
We evaluated the prevalence, aetiologies and antibiotic resistance patterns of bacterial infections in hospitalized patients with laboratory-confirmed SARS-CoV-2. We also investigated comorbidities, risk factors and the mortality rate in COVID-19 patients with bacterial infections.
Methods
This retrospective observational study evaluated medical records of 7249 randomly selected patients with COVID-19 admitted to three clinical centres between 1st January 2021 and 16th February 2022. A total of 6478 COVID-19 patients met the eligibility criteria for analysis.
Results
The mean age of the patients with SARS-CoV-2 and bacterial infections was 68.6 ± 15.5 years (range: 24–94 years). The majority of patients (68.7%) were older than 65 years. The prevalence of bacterial infections among hospitalized COVID-19 patients was 12.9%, most of them being hospital-acquired (11.5%). Bloodstream (37.7%) and respiratory tract infections (25.6%) were the most common bacterial infections. Klebsiella pneumoniae and Acinetobacter baumannii caused 25.2% and 23.6% of all bacterial infections, respectively. Carbapenem-resistance in Enterobacterales, A. baumannii and Pseudomonas aeruginosa were 71.3%, 93.8% and 69.1%, respectively. Age >60 years and infections caused by ≥3 pathogens were significantly more prevalent among deceased patients compared with survivors (P<0.05). Furthermore, 95% of patients who were intubated developed ventilator-associated pneumonia. The overall in-hospital mortality rate of patients with SARS-CoV-2 and bacterial infections was 51.6%, while 91.7% of patients who required invasive mechanical ventilation died.
Conclusions
Our results reveal a striking association between healthcare-associated bacterial infections as an important complication of COVID-19 and fatal outcomes.
Coronavirus disease 2019 (COVID-19), an infection caused by severe acute respiratory syndrome coronavirus 2 (SARS-Co-V2), was declared a global pandemic in early 2020 [
]. The first case of COID-19 in Serbia was registered on 6th March 2020. Following this, several hospitals in the most populous cities were converted into COVID-19 care facilities.
Historically, respiratory viral pandemics have been associated with secondary bacterial infections (SBIs), resulting in higher mortality rates, and COVID-19 is no exception [
Etiology and antimicrobial resistance of secondary bacterial infections in patients hospitalized with COVID-19 in Wuhan, China: a retrospective analysis.
]. However, differentiating viral and bacterial infection can be challenging, leading to inappropriate or prolonged antibiotic use in patients with COVID-19, increasing the risk of antimicrobial resistance (AMR) [
]. Globally, the emergence of AMR jeopardizes infection treatment worldwide, leading to severe diseases and complications. Multi-drug-resistant (MDR) bacteria are estimated to be responsible for 10 million deaths and $100 billion in economic losses by the year 2050 [
Nevertheless, reports of bacterial superinfection in COVID-19 patients are scarce, especially from multi-centre studies, and they frequently omit details of the causative agents [
]. In Italy, the most abundant bacterial pathogens were Escherichia coli (12.5%), Klebsiella pneumoniae (11.6%), Enterococcus faecalis (9.8%) and Pseudomonas aeruginosa (9.6%) [
]. In Northeast Georgia, USA, 53.4% of recovered organisms in COVID-19 patients were Gram-negative bacteria such as P. aeruginosa, E. coli and K. pneumoniae [
]. In Romania coagulase-negative staphylococci and Enterococcus spp. were the predominant blood culture isolates, and Acinetobacter baumannii and Staphylococcus aureus were the most common species in tracheobronchial aspirates [
To respond to COVID-19, policymakers and clinicians need reliable data to support the decision-making processes that save or cost lives. A better understanding of bacterial infections in COVID-19 patients is crucial to improving guidelines for patient management and reducing inappropriate antibiotic prescribing. Thus, the present study aimed to evaluate the aetiology of bacterial infections, their AMR, antibiotic treatment approaches, comorbidities, risk factors, and overall and age-adjusted mortality rate of hospitalized patients with laboratory-confirmed SARS-CoV-2 infection.
Material and methods
Study design
This retrospective, observational study evaluated microbiological database records of 7249 randomly selected laboratory-confirmed COVID-19 patients. Enrolled participants were hospitalized from 1st January 2021 to 16th February 2022 in three regional hospitals in Serbia located in two biggest cities in the country: University Medical Hospital Centre Bežanijska kosa (N = 3592) and Clinical Hospital Centre Zemun (N = 2619), both located in Belgrade, and the Institute for Pulmonary Diseases of Vojvodina (N = 1038) in Novi Sad. All symptomatic patients with a diagnosis of COVID-19 confirmed by reverse transcription polymerase chain reaction (RT-PCR) and/or antigen-testing were evaluated. The study design, eligibility criteria and methods of selection of participants are illustrated in Figure 1. A prespecified sample size was not calculated, as all patients who met the inclusion criteria were enrolled in the study. Duplicated patients and those transferred to another hospital within the first three days following admission were excluded (N = 771). Patients were followed up from hospital admission to death, discharge or censoring, whichever occurred first.
Figure 1Flowchart of the study population. PCR, polymerase chain reaction.
This study was approved by the Ethics Committee of the University Medical Hospital Centre Bežanijska kosa (Number 4843/1, 22nd July 2022), Institute for Pulmonary Diseases of Vojvodina (Number 1-1/5, 3rd August 2022) and Clinical Hospital Centre Zemun (Number 13/1, 16th August 2022).
Data collection, bacterial identification, antimicrobial susceptibility testing and definitions used in the analysis
The following information was extracted from the medical records: date of patient admission, age, sex, type and number of bacterial infections, susceptibility profile of bacterial pathogens, date of microbiological result reporting, antibiotic treatment, invasive and non-invasive mechanical ventilation, comorbidities and in-hospital survival. Missing patient background information was noted as unknown.
Clinical samples were processed according to the laboratory standard operating procedures. The isolated bacteria were identified using VITEK2 system (bioMérieux, Marcy-l’Étoile, France), analytical profile index procedure (API test; bioMérieux, Brussels, Belgium), and matrix assisted laser desorption-ionisation time of flight (MALDI-TOF) mass spectrometry (Bruker Daltonics, Bremen, Germany), according to the manufacturer's instructions. C. difficile was identified using the CerTest Clostridium difficile GDH + Toxin A + B combo card test kit (Certest Biotec, S.L, Zaragoza, Spain), according to the manufacturer's protocol. Antimicrobial susceptibilities of bacteria were determined by the disk diffusion method (Bio-Rad, Hercules, CA, USA), gradient test (Liofilchem S.r.l, Roseto degli Abruzzi, Italy), VITEK2 system (bioMérieux, Marcy-l’Étoile, France) or broth microdilution test (Liofilchem S.r.l, Roseto degli Abruzzi, Italy), according to the European Committee on Antimicrobial Susceptibility Testing (EUCAST) recommendations for 2021 [
The European Committee on Antimicrobial Susceptibility Testing. Breakpoint tables for interpretation of MICs and zone diameters. Version 11.0. 2021. Available at: http://www.eucast.org [last accessed February 2022].
]. Bacterial isolates were classified as MDR (resistant to at least one agent in three or more antimicrobial categories), extensively drug-resistant (XDR, resistant to at least one agent in all but two or fewer antimicrobial categories), and pandrug-resistant (PDR, resistant to all agents in all antimicrobial categories tested) [
Multidrug-resistant, extensively drug-resistant and pandrug-resistant bacteria: an international expert proposal for interim standard definitions for acquired resistance.
Severe COVID-19 was defined as one or more of the following: development of acute respiratory distress syndrome, receipt of oxygen or invasive mechanical ventilation, treatment in high-care or (ICUs), or death, using a modified definition based on the recommendations from the World Health Organisation Clinical Platform External Clinical Advisory Group [
Bacterial infections were categorized as community-acquired infections (CAI) or co-infections or hospital-acquired infections (HAIs) or superinfections or SBIs. HAIs were defined according to the European Centre for Disease Prevention and Control criteria [
]. Low-virulence bacteria (Enterococcus species, non-pneumococcal α-haemolytic streptococci, coagulase-negative staphylococci) were not considered causes of ventilator-associated pneumonia [
The primary outcome of interest was in-hospital mortality. Out-of-hospital deaths were not evaluated. Secondary outcomes included invasive mechanical ventilation.
We followed Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline recommendations [
For statistical analysis, SPSS version 20.0 (SPSS Inc., Chicago, IL, USA) was used. Descriptive statistics were carried out for categorical variables and are presented as absolute frequencies and percentages. The χ2 and Fisher tests were used to assess possible associations between the nominal variables. Bivariate analysis was performed using logistic regression, where crude odds ratio (COR) was calculated. Variables with P≤0.2 were subjected to a multi-variate analysis where the adjusted odds ratio (AOR) was calculated, and significant variables were identified. The level of statistical significance was evaluated at a 95% confidence interval (CI). Kaplan–Meier methods were used for survival curve plotting and examined by log-rank test. A P-value < 0.05 was considered statistically significant.
Results
Baseline characteristics of the study population
Of 7249 randomly selected COVID-19 patients hospitalized between January 2021 and February 2022, 6478 were confirmed to be eligible for the study. 834 (12.9%) had BSIs and/or bacterial co-infections. The mean age of the patients with SARS-CoV-2 and bacterial infections was 68.6 ± 15.5 years (range 24–94 years). The majority of patients (68.7%) were older than 65 years; 456 (54.7%) of the patients were male.
Aetiology and types of bacterial infections
Of the 834 patients, 480 (57.6%) had a monomicrobial infection; 159 of the 354 patients with polymicrobial infections had two bacterial pathogens and 195 had ≥3 pathogens. Gram-negative bacteria (N = 977, 69.3%) outnumbered Gram-positive bacteria (N = 433, 30.7%) (Table I). K. pneumoniae and A. baumannii together accounted for 47% and 75.8% of the most common types of infections, bloodstream and respiratory tract infections, respectively.
Table IBacterial pathogens and sites of infections in 834 COVID-19 patients hospitalized from January 2021 to February 2022
Bacterial pathogen
Site of infection N (%)
Total
Type of infection N (%)
P
UTI
RTI
BSI
SSTI
GE
Other
HAIs
CAIs
Klebsiella pneumoniae
129 (34.3)
116 (28.4)
119 (23.5)
9 (12.3)
0 (0)
21 (46.7)
394 (25.2)
362 (24.9)
32 (21.2)
0.122
Acinetobacter baumannii
32 (8.5)
194 (47.4)
119 (23.5)
14 (19.2)
0 (0)
11 (24.4)
370 (23.6)
354 (24.5)
16 (10.6)
<0.01
Enterococcus spp.
141 (37.5)
5 (1.2)
113 (22.3)
13 (17.8)
0 (0)
4 (8.9)
276 (17.6)
238 (16.4)
38 (25.2)
0.013
Clostridioides difficile producing toxin A and/or B
a Other (peritoneal fluid, drainage sample, bioptate). Stenotrophomonas maltophilia: N = 6; Providencia stuartii: N = 6; Morganella morganii: N = 5; Streptococcus pneumoniae: N = 4; Haemophilus spp.: N = 2; Citrobacter freundii: N = 2; Streptococcus anguinosus: N = 1; Streptococcus mitis: N = 1; Streptococcus viridans: N = 1.
The overall prevalence of HAIs was 11.5%, with 1414 microbiologically confirmed infections identified in 743 patients. The most common HAIs were pneumonia (N = 240), bloodstream infection (N = 268), urinary tract infections (N = 169), gastroenteritis (N = 103) and skin and soft tissue infections (N = 29). Ninety-one (1.4%) patients had CAIs, including urinary tract infections (N = 27) bloodstream infections (N = 19) and skin and soft tissue infections (N = 19).
Antimicrobial resistance
The antimicrobial susceptibilities of the four most prevalent bacterial pathogens are shown in Figure 2. Overall, the prevalences of MDR, XDR and PDR bacterial pathogens were 24.2% (N = 341), 37.9% (N = 534), and 12.8% (N = 180), respectively. All PDR isolates were isolates from patients with HAI, and were either K. pneumoniae (N = 173) or A. baumannii (N = 7).
Figure 2Antimicrobial resistance of the most common isolated bacteria from 834 COVID-19 patients with bacterial superinfection and/or co-infection. AMC, amoxicillin-clavulanic acid; AMK, amikacin; AMP, ampicillin; CAZ, ceftazidime; CIP, ciprofloxacin; CST, colistin; FEP, cefepime; GEN, gentamicin; IPM, imipenem; LVX, levofloxacin; LZD, linezolid; MEM, meropenem; NIT, nitrofurantoin; PIP, piperacillin; SXT, trimethoprim-sulfamethoxazole; TEC, teicoplanin; TGC, tigecycline; TOB, tobramycin; TZP, piperacillin-tazobactam; VAN, vancomycin.
Invasive mechanical ventilation as an important risk factor for bacterial infection
Among 834 COVID-19 patients with bacterial infections, 350 (42%) were ventilated for more than 48 h. The majority of patients receiving invasive respiratory support were aged ≥60 years (N = 270; 77.1%).
Overall, 95% of intubated patients developed ventilator-associated pneumonia; Gram-negative bacilli accounted for the majority of infections (N = 293; 97.3%). Non-fermentative Gram-negative bacilli outnumbered Enterobacterales (58.5% vs 38.9%). The two most common species were A. baumannii (N = 147; 48.8%) and K. pneumoniae (N = 101; 33.6%). The prevalences of carbapenem resistance in A. baumannii, P. aeruginosa and Enterobacterales were 94.5%, 100% and 72.6%, respectively. A total of 284 (81.1%) intubated patients had comorbidities, including cardiovascular disease and hypertension (33.3%), diabetes (21.2%), chronic respiratory disease (10.9%), obesity (7.4%), malignancy (7.9%), neurological disease (5.3%), thyroid disorder (3.7%), immune-related disease (2.7%), chronic kidney disease (2.7%), haematologic disease (1.9%) and chronic liver disease (1.1%).
A total of 675 (80.93%) of patients with simultaneous SARS-CoV-2 and bacterial infections had severe COVID-19. Factors associated with severe disease are shown in Supplementary Table SI. Male gender, age >30 years, and invasive ventilation were significantly associated with negative outcomes in COVID-19 patients with bacterial infection (P<0.05) (Supplementary Table SI). However, comorbidities and the number of bacterial isolates did not influence outcomes.
Antibiotics used in management of COVID-19 patients with bacterial infections
All patients with bacterial infections received antibiotics. However, details of antibiotic therapy was only available for patients at one centre (N = 153); 138 (90.2%) patients were treated with ≥2 antibiotics. The most frequently prescribed antibiotics were metronidazole (69.3%), glycopeptide (66.7%), cephalosporins (49.6%), fluoroquinolones (29.5%) and carbapenems (26.9%).
In-hospital mortality rate of COVID-19 patients with community- and/or hospital-acquired infections
The in-hospital mortality rate of patients with SARS-CoV-2 and bacterial infection was 51.6%; for patients aged >60 years it was 55.3%. The mortality rates of patients with lower respiratory tract and bloodstream infections were 90% and 79.4%, respectively. The relationships between bacterial types and invasive mechanical ventilation and outcome are depicted in Figure 3.
Figure 3Kaplan–Meier plot showing survival of 834 COVID-19 patients with bacterial superinfection and co-infection: (a) Kaplan–Meier plot showing survival of COVID-19 patients with co-infections caused by various bacteria, P<0.01; (b) Kaplan–Meier plot showing survival of COVID-19 patients with and without invasive mechanical ventilation, P<0.01.
This is the first comprehensive multi-centre study of bacterial infections in hospitalized COVID-19 patients in the Western Balkan region of Europe. Compared with an extensive meta-analysis, we found a higher rate of bacterial infection (12.9% vs 7%) [
]. However, our findings accord with another Serbian study that reported that the prevalence of HAI had increased by >100% during the COVID-19 pandemic [
]. One possible contributor to this apparent increase may be that there was a low rate of microbiological testing in Serbia in the pre-pandemic period [
Empiric antibacterial therapy and community-onset bacterial coinfection in patients hospitalized with coronavirus disease 2019 (COVID-19): A multi-hospital cohort study.
]. Rates of C. difficile infection and non-C. difficile antibiotic-associated diarrhoea were comparable to those reborted by Maslennikov et al. (18.5% vs 16.7% and 3.4% vs 5.7%), respectively [
]. Contributors to these high rates may include patient transfers between hospitals and overprescription of antibiotics pre-admission and post-admission in Serbia [
We found that the prevalence of carbapenem-resistance in A. baumannii was extremely high, but that was also the case in a pre-pandemic study (93.8% vs 93.7%),. However, the rate of carbapenem-resistance in P. aeruginosa that was substantially higher than was reported pre-pandemic (69.1% vs 43.1%) [
A striking finding in our study was the widespread use of metronidazole (69.3%) and vancomycin (66%); the need for these antibiotics was not supported by the pattern of infections observed. Markovskaya et al. found that although the classes of antibiotics prescribed varied between studies, the most widely prescribed antibiotics were generally broad-spectrum agents (e.g., fluoroquinolones, β-lactams, and macrolides) [
In the present study, 51.6% of COVID-19 patients with bacterial infections died. Widely varying mortality rates have been reported previously, ranging from 6.5% to 66.7% depending on the study population and observation periods [
]. Notably, in our study, the lowest probability of survival was for patients with infections with Staphylococcus spp. and P. aeruginosa. The association between Staphylococcus spp. and bloodstream infections may explain the former observation. Our finding that non-survivors were more likely to be older, have polymicrobial infections and to have been ventilated accords with previous work [
A surge of COVID-19 cases, the extended use of common personal protective equipment, and an overwhelmed and understaffed healthcare system in Serbia may have contributed to the higher rates of HAIs [
]. Our results add to the growing evidence that existed pre-COVID-19 that infection prevention and control (IPC) should be focused on MDR bacteria and C. difficile, as well as on patients undergoing invasive mechanical ventilation. We recommend attention to the following IPC measures: (1) cleaning, disinfecting, and sanitizing protocols; (2) glove changing before and after contact with patients with MDR bacteria (use of alcohol gel on gloved hands is common practice in Serbia); (3) use of hypochlorite-based products for surface disinfection in wards where C. difficile infection has been detected; (4) improved adherence to hand hygiene protocols; (5) routine MDR bacteria screening of patients on admission and during their hospital stay, with use of contact precautions for those found to be positive and an early IPC response to evidence of hospital transmission of these bacteria [
World Health Organization Guidelines for the prevention and control of carbapenem-resistant Enterobacteriaceae, Acinetobacter baumannii and Pseudomonas aeruginosa in health care facilities.
Infection prevention and control measures and tools for the prevention of entry of carbapenem-resistant Enterobacteriaceae into healthcare settings: guidance from the European Centre for Disease Prevention and Control.
]; (6) strict antimicrobial stewardship inside and outside hospitals and outpatients – this is a crucial tool to prevent the emergence of antibiotic-resistant infections; (7) prevention and early diagnosis of ventilator-associated pneumonia.
Our study has some strengths and limitations. It was a large multi-centre study in a country where there are few published data, and no centralized monitoring of HAI rates. Data of this sort are vital for planning targeted IPC measures. Nevertheless, the study had important limitations. As a retrospective study it is possible that some patients were colonized rather than infected. Factors such as prior antibiotic therapy and use of corticosteroids and interleukin-6 receptor antagonists in treatment protocols may have confounded our results. There were some missing data that may have influenced our analysis of risk factos, but not seriously. Finally, we did not evaluate any possible influence of immunisation against COVID-19.
Despite the study's limitations, we noted high rates of hospital-acquired bacteria infection in COVID-19 patients, including with MDR pathogens and C. difficile. These observations point to the urgent need for evidence-based IPC policies and practices as we emerge from the pandemic.
Acknowledgements
We acknowledge hospital management for approving the study and Dragana Andjelkovic and Dusan Bozic for data extraction.
Appendix A. Supplementary data
The following are the Supplementary data to this article.
There were no conflicts of interest in the writing of this article.
Funding
This work was supported by the Ministry of Education, Science and Technological Development of the Republic of Serbia, Serbia [grant No. 451-03-68/2022-14/200110].
Etiology and antimicrobial resistance of secondary bacterial infections in patients hospitalized with COVID-19 in Wuhan, China: a retrospective analysis.
The European Committee on Antimicrobial Susceptibility Testing. Breakpoint tables for interpretation of MICs and zone diameters. Version 11.0. 2021. Available at: http://www.eucast.org [last accessed February 2022].
Multidrug-resistant, extensively drug-resistant and pandrug-resistant bacteria: an international expert proposal for interim standard definitions for acquired resistance.
Empiric antibacterial therapy and community-onset bacterial coinfection in patients hospitalized with coronavirus disease 2019 (COVID-19): A multi-hospital cohort study.
Guidelines for the prevention and control of carbapenem-resistant Enterobacteriaceae, Acinetobacter baumannii and Pseudomonas aeruginosa in health care facilities.
Infection prevention and control measures and tools for the prevention of entry of carbapenem-resistant Enterobacteriaceae into healthcare settings: guidance from the European Centre for Disease Prevention and Control.
00138-X&id=gr1.jpg){kind=link}
00138-X&id=gr2.jpg){kind=link}
00138-X&id=gr3.jpg){kind=link}
00138-X&id=figs1.jpg){kind=link}