Pseudomonas aeruginosa Infections in Patients with Severe COVID-19 in Intensive Care Units: A Retrospective Study

Patients hospitalized in ICUs with severe COVID-19 are at risk for developing hospital-acquired infections, especially infections caused by Pseudomonas aeruginosa. We aimed to describe the evolution of P. aeruginosa infections in ICUs at CHRU-Nancy (France) in patients with severe COVID-19 during the three initial waves of COVID-19. The second aims were to analyze P. aeruginosa resistance and to describe the antibiotic treatments. We conducted a retrospective cohort study among adult patients who were hospitalized for acute respiratory distress syndrome due to COVID-19 and who developed a hospital-acquired infection caused by P. aeruginosa during their ICU stay. Among the 51 patients included, most were male (90%) with comorbidities (77%), and the first identification of P. aeruginosa infection occurred after a median ICU stay of 11 days. Several patients acquired infections with MDR (27%) and XDR (8%) P. aeruginosa strains. The agents that strains most commonly exhibited resistance to were penicillin + β-lactamase inhibitors (59%), cephalosporins (42%), monobactams (32%), and carbapenems (27%). Probabilistic antibiotic treatment was prescribed for 49 patients (96%) and was subsequently adapted for 51% of patients after antibiogram and for 33% of patients after noncompliant antibiotic plasma concentration. Hospital-acquired infection is a common and life-threatening complication in critically ill patients. Efforts to minimize the occurrence and improve the treatment of such infections, including infections caused by resistant strains, must be pursued.


Introduction
Pseudomonas aeruginosa is an opportunistic Gram-negative bacterium that is largely associated with hospital-acquired infections (HAIs) [1,2].These infections can lead to ventilator-associated pneumonia (VAP), bloodstream infection (BSI), catheter-associated urinary tract infection (cUTI), and surgical infection [1].Worldwide, P. aeruginosa is the fifth most common microorganism responsible for bacterial infection, and it causes the most deaths and years of life lost in hospitals [3].The 30-day mortality for P. aeruginosa bacteraemia is approximately 40% [4].P. aeruginosa is one of the main bacteria responsible for HAIs [2], especially in intensive care units (ICUs), where 40% of P. aeruginosa infections are distributed [5].
A slight increase in the incidence of HAIs caused by P. aeruginosa was observed during the COVID-19 pandemic [6].The risk of acquiring HAIs significantly increases with the severity of COVID-19 and the duration of hospitalization [7].A large meta-analysis that described 355,579 COVID-19 patients across 138 studies performed between December 2019 and May 2021 revealed that the prevalence of bacterial coinfection and secondary infection among COVID-19 patients in ICUs were 8.4% and 39.9%, respectively [8].Patients with severe COVID-19 who were hospitalized in ICUs were significantly more likely to be coinfected with bacteria [9,10].In France, COVID-19 patients in ICUs were significantly more likely to be infected by P. aeruginosa bacteraemia (12.4% vs. 8.9%) and VAP (16.8% vs. 16.1%)than those without COVID-19 [11].Coinfections, including P. aeruginosa infections, which occurred among COVID-19 patients hospitalized in ICUs during the three initial waves of the COVID-19 pandemic (from March 2020 to May 2021) had a negative impact on patient prognosis, with a mortality rate of 74% [12].Mortality increased among hospitalized patients with resistant and multidrug resistant (MDR) P. aeruginosa infections compared to those with susceptible P. aeruginosa infections [13].
P. aeruginosa is complex to treat because it is intrinsically resistant to many antimicrobial agents, and this bacterium can additionally acquire other resistances [14].Among hospitalized COVID-19 patients, P. aeruginosa is among the main bacterial species with resistance to antibiotics, accounting for approximately 25% of resistant strains [15].In 2020, in Europe, 30.1% and 17.3% of 20,675 invasive P. aeruginosa isolates reported to EARS-Net were resistant to one and at least two antimicrobial groups under surveillance (piperacillin/tazobactam, fluoroquinolones, ceftazidime, carbapenems, and aminoglycosides) [14].
During the first waves of the COVID-19 pandemic, the consumption of penicillin with β-lactamase and carbapenems significantly increased in ICUs [16].The main antibiotics prescribed in hospitalized patients with COVID-19 were amoxicillin/clavulanic acid, piperacillin/tazobactam, cephalosporines, and carbapenems [17].The inappropriate use of antibiotic treatments was frequently described and could lead to infections by MDR bacteria, which have a significant impact on the management of hospitalized patients with COVID-19 infections, leading to a prolonged hospitalization time and increased mortality [7,8,15,18].
Due to the serious consequences of P. aeruginosa infections among hospitalized patients and, more particularly, patients admitted to ICUs with severe COVID-19, additional studies are needed.Moreover, a focus on the antibiotic resistance of P. aeruginosa is needed because the WHO has listed carbapenem-resistant P. aeruginosa as a pathogen of critical priority that requires further research and development [19].A recent review indicated that antibiotic resistant P. aeruginosa in patients with BSIs was rarely reported in the literature during the COVID-19 pandemic and needs to be further explored [20].
The main aim of this study was to describe the evolution of P. aeruginosa infections in patients hospitalized in ICUs at CHRU-Nancy (France) for Acute Respiratory Distress Syndrome (ARDS) due to COVID-19 during the three initial waves of the COVID-19 pandemic in France (1 March 2020 to 31 May 2021).The second aim was to analyze the resistance of P. aeruginosa strains that cause infection and to describe the antibiotic treatments.

Demographic and Clinical Data
A total of 645 adult patients were admitted for COVID-19 in ICUs and continuing care units in our hospital from 1 March 2020 to 31 May 2021.During this period, 185 adult patients admitted in the ICUs in our hospital were infected or colonized by P. aeruginosa, mainly males (75%) with a mean age of 62 ± 11 years.Among them, 53 adult patients met the inclusion criteria, but two refused to participate in the study.The characteristics of the 51 patients included in this study are summarized in Table 1.The participants were aged 65 ± 12 years (range: 36-82) with a body mass index of 30 ± 6 kg/m 2 , and 90% were male.Three-quarters of the patients (77%) presented with at least one comorbidity, mainly arterial hypertension (61%), obesity (45%), or chronic cardiac disease (45%).The median hospital stay and ICU stay were 36 (range: ) and 27 days (range: 6-172), respectively.Patients were mainly admitted from smaller peripheral hospitals (59%), and 43% died during their ICU stay.The two main causes of death were multiple-organ failure (55%) and respiratory failure (23%).Only two significant differences between survivors and patients who died were highlighted: the patients who died were older (p < 0.03) and the survivors had longer hospital stays (p < 0.03).Prior to hospital admission, it was documented that 41% of patients received at least one antibiotic treatment, mainly beta-lactams (37%)-including third-generation cephalosporins such as cefotaxime and ceftriaxone (26%), amoxicillin/clavulanate (22%), piperacillin/tazobactam (8%), and amoxicillin (4%)-and macrolides (26%) including spiramycin (22%) and azithromycin (4%).

Pseudomonas aeruginosa Infections
Among the 51 patients, 15 acquired their first P. aeruginosa infection during the first COVID-19 pandemic wave, 12 during the second, and 24 during the third (Figure 1).

Pseudomonas aeruginosa Infections
Among the 51 patients, 15 acquired their first P. aeruginosa infection during the first COVID-19 pandemic wave, 12 during the second, and 24 during the third (Figure 1).During these three waves, the median delay between the first diagnosis of COVID-19 and the identification of a P. aeruginosa infection among patients was 14 days (Q1-Q3: 10-25).The first identification of a P. aeruginosa infection occurred after a median ICU stay of 11 days (Q1-Q3: 6-18).After one month (30 days) in the ICU, 90% of patients were infected with P. aeruginosa, and 24% were infected with MDR P. aeruginosa (Figure 2).During these three waves, the median delay between the first diagnosis of COVID-19 and the identification of a P. aeruginosa infection among patients was 14 days (Q1-Q3: 10-25).The first identification of a P. aeruginosa infection occurred after a median ICU stay of 11 days (Q1-Q3: 6-18).After one month (30 days) in the ICU, 90% of patients were infected with P. aeruginosa, and 24% were infected with MDR P. aeruginosa (Figure 2).A total of 188 strains of P. aeruginosa were collected from the patients.After removing duplicate strains, 122 P. aeruginosa strains obtained from 106 samples and comprising 118 antibiograms were included in this study.The four strains without antibiograms corresponded to respiratory samples that were analyzed by multiplex polymerase chain reaction (m-PCR; FilmArray ® ).

Antibiotic Treatments
All patients (100%) were treated with antibiotics during their ICU stay for a median of 20 days (Q1-Q3: 12-27), representing approximately 70% of their ICU stay.
Among the seven patients who acquired BSIs by P. aeruginosa, four were initially treated by ineffective antibiotics (antibiograms revealed resistance to antibiotics previously prescribed), two had initiated their antibiotic treatments < 48 h before the blood samples, one was not treated by antibiotic before the blood sample, and one presented a BSI with a susceptible strain despite 10 days of piperacillin-tazobactam.After BSI identification and antibiogram results, all patients were treated with susceptible antibiotics for P. aeruginosa.

Discussion
In our hospital, 53 adult patients acquired P. aeruginosa infection in the ICU after admission for ARDS due to COVID-19 during the three initial waves of the COVID-19 pandemic.During this period, numerous patients infected with COVID-19 presented with ARDS and were hospitalized in ICUs [21].Up to 56% of these patients acquired bacterial coinfections, mainly VAP, BSIs, and UTIs, which often led to sepsis and septic shock [22,23].P. aeruginosa infections were mainly identified in respiratory samples and from male patients [5].Our results are in accordance with these previous studies [5,22,23] because, among the 51 included patients, 90% were males, and P. aeruginosa was mainly responsible for VAP (92%), BSIs (14%), and catheter-associated BSIs (10%).
Appropriate management of COVID-19 patients who develop HAIs in the ICU is crucial because these infections can occur in numerous patients, lead to difficulties in treatment, and have a high mortality rate (43% of patients who acquired P. aeruginosa died in our study).
Several clinical conditions may explain the increased rate of HAIs during the COVID-19 pandemic, including prolonged hospitalization in the ICU (the median ICU stay was 27 days in our study, with 50% of patients affected by P. aeruginosa after 11 days), worse clinical presentation at ICU admission (all the included patients in our study presented with ARDS, all received oxygenation support, and 84% required prone positioning), the use of systemic corticosteroid treatment for ARDS in the ICU (69% of patients were treated with corticosteroids in our study), the presence of immunosuppressive comorbidities such as diabetes (25% of patients had diabetes in our study), the administration of immunemodulator treatments such as tocilizumab, and the use of invasive mechanical ventilation (94% of patients were treated with invasive mechanical ventilation for a median duration of 30 days in our study) [7,24].One major explanation of the association between the duration of the ICU stays and the acquisition of P. aeruginosa HAIs is the exposure to at-risk invasive devices such as mechanical ventilation, vascular catheters, and urinary catheters [25].
Regarding antibiotic resistance, 61% of the patients included in our study were infected with resistant strains of P. aeruginosa, 27% with MDR strains and 8% with XDR strains.The resistance profiles of the first strains of P. aeruginosa identified in our study were comparable to those of P. aeruginosa strains in France in 2020: piperacillin-tazobactam resistance (13% vs. 17%), ceftazidime resistance (11% vs. 13%), carbapenem resistance (11% vs. 13%), aminoglycoside resistance (6% vs. 6%), and MDR (14% vs. 8%) [14].The subsequent strains sampled in our study were significantly more resistant, with the selection of resistant P. aeruginosa strains and the acquisition of resistances.P. aeruginosa is not the only concerning bacteria; all the ESKAPE pathogens (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, P. aeruginosa, and Enterobacter spp.) are causative bacteria of HAIs well-known to present antibiotic resistances [26].The main Gram-negative ESKAPE bacteria isolated from COVID-19 patients were A. baumannii, K. pneumoniae, and P. aeruginosa [27,28].Most of the studies highlighted an increase in the prevalence of carbapenem-resistant A. baumannii and K. pneumoniae [18].The highest prevalence of resistance was reported for MDR A. baumannii [8], with a prevalence from 40 to 95% in patients with HAI pneumonia and a mortality from 45 to 85% [26].
In their narrative review, Gaudet et al. reported that critically ill COVID-19 patients presented with a greater incidence of HAIs with MDR isolates in the ICU.These authors highlighted several pathophysiological factors that could explain these findings, including COVID-19-mediated post-aggressive immunoparalysis; exposure to antibiotics, steroids, and other immunomodulating agents; a longer ICU stay; exposure to invasive devices; and organizational constraints triggered by the pandemic [29].During the first waves of the COVID-19 pandemic, nurses in the ICU had an excessive workload; this increase in workload was associated with poor quality of care, unfavourable patient safety, and unfavourable infection prevention [30].Moreover, the lack of adequate personal protective equipment for frontline health care workers (including respirators, gloves, face shields, gowns, and hand sanitizers) was reported, potentially heightening the risk of MDR transmission [31].
Antibiotic exposure is a significant risk factor for the spread of antimicrobial resistance [8].Indeed, the use of antibiotics exerts selective pressure that isolates the resistant strains, particularly for longer treatments [32].A study performed in 260 United Kingdom hospitals that included nearly 49,000 hospitalized COVID-19 patients revealed that 37% of patients received antibiotics prior to hospital admission, and 85% received antibiotics during hospitalization [33].In our study, 41% of patients received antibiotics before hospital admission; this percentage could be underestimated due to the retrospective collection of data reported by patients at admission to our hospital.All included patients were treated with antibiotics during their ICU stay for a median of 20 days.The use of antibiotics in ICUs is crucial for treating patients infected with bacteria such as P. aeruginosa.However, the proper use of antibiotics is essential for fighting antibiotic resistance and reducing the side effects of treatment [34].Unfortunately, we highlighted the fact that probabilistic antibiotic treatment was prescribed for 96% of patients and was only adapted for 51% of patients after receiving an antibiogram and 33% of patients after receiving a noncompliant antibiotic plasma dosage.These percentages of adaptation of the antibiotic treatments could be underestimated due to the retrospective collection of data from electronic patient records; ICU teams were able to quickly adapt the treatment but only reported this change a few days later in the hospital software.To improve antibiotic prescription and to adapt treatments with relevant clinical and microbial information in real time (such as antibiogram and antibiotic plasma dosage results), our hospital installed a computerized decision support system for antimicrobial stewardship after the COVID-19 pandemic [35].This initiative was implemented to improve antibiotic treatments and the management of HAIs in a similar context to increase the quality of care.
This study describes the in-depth characteristics of patients admitted to a French hospital for ARDS due to COVID-19 during the three initial waves of the COVID-19 pandemic and who acquired P. aeruginosa infections during their ICU stay.Moreover, this study precisely describes the resistance of P. aeruginosa and provides new insights into the increase in the antibiotic resistance of strains during the patients' ICU stay.The two main limitations of this study are linked to its design because it was an observational retrospective monocentric study with a limited number of patients, which may have affected the analysis and generalizability of the results.On the one hand, the retrospective collection of data from electronic patient records could underestimate several outcomes such as the percentage of patients receiving antibiotics before hospital admission and the percentages of adaptation of antibiotic treatments after antibiogram and after antibiotic plasma dosage.This limitation is frequent in this field of research; Langford et al. reported in their systematic review that among 148 studies regarding antimicrobial resistance in patients with COVID-19, all were observational, and most were retrospective cohort studies (81%) [8].On the other hand, despite the inclusion of patients hospitalized in several ICUs in different buildings of our hospital with different medical teams, this study includes a small number of patients in a single hospital.So, this study presented a limited external validity, limiting the generalization of the data.

Study Design and Setting
We conducted a single retrospective cohort study in ICUs at CHRU-Nancy (Regional University Hospital of Nancy) during the three initial waves of the COVID-19 pandemic in France.The first wave occurred from 1 March to 1 August 2020, the second from 2 August 2020 to 1 January 2021, and the third from 2 January to 31 May 2021.During the initial waves of the COVID-19 pandemic, the CHRU-Nancy received critical COVID-19 patients in two ICUs and two continuing care units, which were upgraded to ICUs (up to 48 beds), and in one surgical ICU and two surgical continuing care units.
Ethics approval was obtained from the Ethics Committee of CHRU-Nancy (approval no.413).The study was registered on 21 November 2023 on ClinicalTrials.gov(Identifier: NCT06141837).

Patient Selection
The inclusion criteria were as follows: • Adult patients were hospitalized for at least 48 h in an ICU at CHRU-Nancy for ARDS due to COVID-19 (confirmed by reverse transcription polymerase chain reaction or an antigenic test).

•
Patients were hospitalized from 1 March 2020 to 31 May 2021.

•
Patients developed a HAI caused by P. aeruginosa during their ICU stay.
The exclusion criteria were as follows: • Patients < 18 years old.

•
Patients with P. aeruginosa isolated <48 h following ICU admission.

•
Patients who did not want to be included in the study.

Data Collection and Outcomes
Epidemiological and demographic data, medical history, and comorbidities at the time of hospitalization were reported.Data on antibiotic treatments received before hospital admission, the length of hospital and ICU stay, and mortality were also obtained directly from medical records.The records of all patients with positive P. aeruginosa results were reviewed, and infections occurring 48 h after hospital admission were categorized as BSI, VAP, or cUTI according to CDC definitions.The bacterial susceptibility of P. aeruginosa was interpreted according to the European Committee on Antimicrobial Susceptibility Testing (EUCAST) criteria [36].
As described by Lyu et al. [5], duplicate samples were eliminated, and when multiple strains were isolated from the same site, with the same antibiogram and from one patient within one month, they were designated as belonging to the same strain, with the first isolate being used as a representative sample.
P. aeruginosa relapse was defined as a positive culture within 30 days after discontinuing antibiotic therapy for the initial infection.
Following the P. aeruginosa categories listed by Magiorakos et al. [37], bacteria were classified as multidrug resistant (MDR) if they acquired nonsusceptibility to at least one agent in three or more antimicrobial categories, extensively drug resistant (XDR) if they acquired nonsusceptibility to at least one agent in all but one or two antimicrobial categories, and pandrug resistant (PDR) if they acquired nonsusceptibility to all agents in all antimicrobial categories.According to Kadri et al. [38], difficult-to-treat resistant (DTR) P. aeruginosa was defined as intermediate or resistant to all reported agents in carbapenems (meropenem and imipenem), β-lactam (ticarcillin-clavulanic acid, piperacillin-tazobactam, ceftazidime, cefepime, and aztreonam), and fluoroquinolone (ciprofloxacin) categories.
To describe the evolution of P. aeruginosa infections among patients, the number of newly infected patients was reported each month.To describe the resistance and susceptibility of P. aeruginosa strains, antibiograms following the EUCAST criteria were used [36].
The French guidelines [39][40][41] were used to analyze the antibiotic treatments, the appropriateness of the prescriptions (molecule, dose, route of administration, and duration), and the compliance of the antibiotic plasma dosages.

Statistical Analysis
The data were collected using Microsoft Excel and analyzed using RStudio (R version 4.3.2).The data are presented as numbers and percentages for categorical variables and as the means ± standard deviations (SDs) and ranges (min-max) or medians and interquartile ranges (Q1-Q3) depending on the distribution of continuous variables.Chi-squared tests were used for categorical variable analysis or Fisher's exact tests when the expected frequencies were less than five, and the Bonferroni correction was used for multiple comparisons.The Mann-Whitney U test was used for continuous variable analysis.The statistical significance was set at p < 0.05.

Conclusions
Patients who were hospitalized in ICUs for severe COVID-19 during the initial waves of the COVID-19 pandemic and who developed P. aeruginosa infections had difficult-totreat infections and high mortality.The acquisition and increase in antibiotic resistance among P. aeruginosa strains during the ICU stay complicated treatment.Efforts to minimize the likelihood of acquiring such infections, especially infections with MDR, XDR and DTR P. aeruginosa, must be pursued.Adaptation of antibiotic treatments according to antibiograms and antibiotic plasma dosage results should be quickly performed in infected patients to optimize antimicrobial treatments.

Figure 1 .
Figure 1.The number of first identified P. aeruginosa infections each month among 51 patients admitted to the intensive care unit (ICU) for ARDS due to COVID-19 during the three initial waves.

Figure 1 .
Figure 1.The number of first identified P. aeruginosa infections each month among 51 patients admitted to the intensive care unit (ICU) for ARDS due to COVID-19 during the three initial waves.

Figure 2 .
Figure 2. Occurrence of P. aeruginosa infections (red) and multidrug resistant (MDR) P. aeruginosa infections (blue) in 51 patients in the intensive care unit (ICU).

Figure 2 .
Figure 2. Occurrence of P. aeruginosa infections (red) and multidrug resistant (MDR) P. aeruginosa infections (blue) in 51 patients in the intensive care unit (ICU).

Table 1 .
Demographic and clinical characteristics of COVID-19 patients infected with Pseudomonas aeruginosa during their ICU stay and data based on survival during their hospital stay.

Table 2 .
Antibiotic resistance of the 118 Pseudomonas aeruginosa antibiograms among 51 patients.

Table 4 .
The ten most commonly prescribed antibiotics among the 51 patients during their ICU stay.
* Only piperacillin was dosed; NA: not applicable.