Urinary Tract Infections in Hospitalized COVID-19 Patients, What’s Up, Doc?

The SARS-CoV-2 pandemic might have increased the risks of healthcare-associated infections (HAIs); however, several studies of HAI such as urinary tract infections (UTIs) and catheter-associated urinary tract infections (CAUTIs) have shown contradictory results. The aim of this study is to assess the clinical features of UTIs and bacterial isolates from urine samples of hospitalized COVID-19 patients. We conducted a retrospective observational study including 87 COVID-19 patients with UTIs admitted to our centre. Bacterial UTIs presented were 87: 9 (10.3%) community-acquired UTIs (coinfection group) and 78 (89.6%) hospital-acquired UTIs (superinfection group). In the coinfection group, the most frequent type was non-CAUTI with 5 (55.5%) patients; however, the most frequent UTI in the superinfection group was CAUTI, with 53 (67.9%) patients. The median number of days of hospitalization in coinfected patients was lower than superinfection patients: 13 (IQR 11, 23) vs. 34 days (IQR 23, 47) p < 0.006. All UTI patients admitted to ICU, 38 (43.7%), belonged to the superinfection group. The mortality rate was 26.4% (23/87), 22/23 in the superinfection group. The most common microorganisms were E. coli 27 (28.4%), E. faecalis 25 (26.3%) and E. faecium 20 (21.1%). There was an increased incidence of E. faecalis and E. faecium in UTIs as well as hospital-acquired UTIs. This can be related to urethral catheterization during hospitalization, UCI admissions and the number of days of hospitalization.


Introduction
The COVID-19 pandemic emerged after the first cases were reported in China in December 2019. As of 6 December 2021, more than 265 million cases of COVID-19 caused by SARS-CoV-2 infection have been reported, with more than 5.2 million deaths, according to WHO data [1]. The risk of these patients for severe COVID-19 and hospitalization is higher in people aged ≥60 years, those living in a nursing home or long-term care facility and those with chronic medical conditions [2][3][4]. Two large cohorts of surveillance have identified that 14% of patients required hospitalization, and 2% were admitted to the intensive care unit (ICU) [5,6]. These data have led to one of the main challenges for the global healthcare systems, the large number of patients requiring hospitalization, notably in the ICU. Some studies have identified low rates of coinfections on admission [7,8], even those studies that have analysed bacterial and fungal infections of the respiratory system [9][10][11][12]. When coinfection (community-acquired, diagnosed within the first 48 h since admission) or superinfection (diagnosis ≥48 h from admission, considered as hospitalacquired) was presented, it seems to be associated with more severe COVID-19 and worse outcomes [13,14], especially in critically ill patients [15].
The aim of this study is to assess the clinical features of UTIs and the bacterial isolates from urine samples of hospitalized patients with severe COVID-19 during the first two months of the first wave of maximum incidence in our hospital.

Study Design and Patients
We conducted a retrospective observational study of all patients admitted at a tertiary hospital in Madrid, with a confirmed diagnosis of COVID-19 and significative urine culture from 25 February to 4 May 2020. Our large cohort of COVID-19 patients hospitalized was published [18], and there is one subanalysis about respiratory coinfections published [19]. The study was approved by the Clinical Research Ethics Committee of Hospital Universitario La Paz with the code HULP: PI-PI-4321.3.

Data Collection and Definitions
We included in the study all patients aged 18 years and over with positive RT-PCR for SARS-CoV-2 and significative bacterial urine culture who were hospitalized between 25 February and 4 May 2020. Electronic health records and microbiology laboratory data were used to collect information demographics, risk factors and comorbidities, antibiotic therapy and anti-COVID-19 therapy, microbiological data and outcome variables. Conventional microbiological urine cultures were requested when infection was suspected, not by protocol. We only included in the study UTIs defined as the presence of a bacterial positive urine culture with clinical signs of infection and/or worsening organ failure, according to the Centers for Disease Control (CDC), the Spanish Society of Infectious Diseases and the Clinical and Infectious Diseases Society of America clinical practice Guidelines [20][21][22]. UTI data were reviewed to determine the presence of a true clinical UTI and its source. We divided UTIs into asymptomatic bacteriuria (ASB), symptomatic urinary tract infection noncatheter and CAUTI. BSIs were defined according to the CDC [20] and were only included if patients met the rest of the inclusion criteria. Related bacteraemia was defined when the same microorganism was isolated in blood and urine samples. All UTIs were categorized as urinary tract coinfection (community-acquired UTI) or urinary tract superinfection (hospital-acquired UTI). All patients were treated according to the COVID-19 protocol of our institution and the Spanish Agency of Medicine guidelines [23]. We registered acquired resistance mechanisms to antimicrobials of isolates of urine cultures. Outcome variables were ICU admission, length of hospital stay, hospital discharge, mortality and survival until 30 September 2021.

Laboratory Technique
All patients had a nasopharyngeal swab sample positive RT-PCR for SARS-CoV-2. Urine and blood samples were processed with standard procedures to the protocols of the Hospital's Clinical Microbiology Service. We collected only bacterial urine isolates of more than >100,000 CFU/mL. Bacterial species were identified by matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS−Bruker Daltonics, Bremen, Germany). Antimicrobial susceptibilities were determined by automated broth microdilution assay system (Microscan Walkaway ® , Beckman Coulter, Brea, CA, USA) and interpreted by means of guidelines of European Committee on Antimicrobial Susceptibility Testing (EUCAST) [24].

Statistical Analysis
The median and IQR were used for the quantitative variables and absolute and relative frequencies for the qualitative variables. For the statistical analysis, first an analysis of the distribution of the variables was performed with the Kolmogorov-Smirnov test. As no variable converged to a normal distribution, the Mann-Whitney U test was used for the comparison by origin of coinfection and the Kruskal-Wallis test for the comparison by type of coinfection. The Chi Square test was also used to compare qualitative variables, and Fisher's exact test, when necessary. Statistical significance was defined as a p-value ≤ 0.05. St Statistical analysis was performed using R software (version 4.1.1, R Core Team (2020), Vienna, Austria).

Results
We determined that 87 out of 2281 (3.81%) COVID-19 patients admitted to our hospital presented fulfilled inclusion criteria for the study. Clinical characteristics of the patients are shown in Table 1, comparing urinary tract coinfection and urinary tract superinfection of these patients.

Demographic and Epidemiological Data
Forty-eight patients were women (55.2%), and 51 (58.6%) were in the age range of 70-80 years. Among the conditions identified as risks of UTI, urethral catheterization was the most important, followed by urological disease. Urethral catheterization was more frequent in urinary tract superinfections. All UTIs developed in ICU were superinfections.

Type of UTI, Laboratory and Microbiological Characteristics
Community-acquired urinary tract coinfections were uncommon, whereas hospitalacquired urinary tract superinfections were more frequent. Overall, CAUTI was the most common type, although in the coinfection group, non-CAUTI prevailed. Seventy-nine (90.8%) patients had monomicrobial infections vs. three (33.3%) polymicrobial infections, this being a statistically significant association p = 0.042. The mean length of stay was significantly longer among patients with urinary tract superinfections. The median days of hospitalization among urinary tract coinfected patients was 13 days (IQR 11-23) and 34 days (IQR 23-47) in hospital-acquired UTI patients, p = 0.006. All ICU UTIs were the superinfection type, with almost a month of ICU stay (27 days (IQR 19-37)); 35/38 of them were CAUTI. In the superinfection group, 19 (24.4%) patients presented with bacteraemia (only two outside the ICU), of which 8 (42.1%) were bacteraemia related to a UTI (Table 1).
Details of 95 isolates urine cultures are presented in Table 2. The most common microorganism was E. coli 27 (28.4%), followed by Enterococci (E. faecalis 25 (26.3%) and E. faecium 20 (21.1%)). A total of 78 (89.6%) patients were treated with targeted antibiotic therapy using their antibiogram. Microorganisms with some acquired resistance mechanisms were registered in 61 patients, the most frequent of which included quinolones resistances, other B-lactams and Gram-negative bacilli extended-spectrum beta-lactamase (ESBL). Gram-negative carbapenemase-producing bacilli type OXA-48 and Gram-negative carbapenemase-producing bacilli type VIM only appeared in UTI superinfections.  Table 1 describes the main risk factors potentially associated with the development of hospital-acquired infections in COVID-19 patients. The majority of patients on admission were treated with a broad-spectrum antibiotic such as B-lactams (n = 63, 72.4%); ceftriaxone was the most frequent (n = 49, 56.3%). Other antibiotics used were azithromycin (n = 47, 54.0%) or quinolones (n = 11, 12.8%). Thirty-two (36.8%) patients received the combination of ceftriaxone and azithromycin. Moreover, 77 (88.5%) patients were given a course of antibiotic treatment for a duration of at least 3 days. In addition, 34 (39.1%) patients received therapy with corticosteroids and/or 18 (20.7%) tocilizumab.

Risk Factors for Healthcare-Associated Infection and Outcome
The development of hospital-acquired infections was significantly associated with days of admission, ICU admission and central venous catheter (CVC). A total of 23 (26.4%) patients died during admission; of these, 19 were admitted to the ICU, and 11 presented with bloodstream infection. Four of eight patients with bacteraemia-related infections died, 22 (95.6%) patients died who belonged to the group of urinary tract superinfections and 10 (15.6%) patients died during the follow-up period; all of them belonged to the superinfection group (Table 1).

Discussion
We studied bacterial UTIs in COVID-19 patients during the first outbreak of the SARS-COV-2 pandemic in Spain. We found a rate 3.81% bacterial urinary tract infection in hospitalized patients (87/2281) where CAUTI was the most frequent type of UTI. There was an increased incidence of Enterococci UTI and superinfection type of UTI. Urethral catheter during hospitalization, ICU admissions and numbers of days of hospitalization were higher in the superinfection group.
Infection rates reported by other authors were 1.9 to 9.7% [14,15,25,26]. We have observed that it was particularly difficult to compare UTI rates (3.81% in our study) in COVID-19 patients because the large series in the SARS-COV-2 pandemic did not specifically analyse this type of infection [7,8,12,27] or the UTI definitions used were different according to the author. Thus, in the studies evaluating HAIs, CAUTI was the type of UTI studied, and its incidence was evaluated by different measures: a rate of CAUTIs per 1,000 catheter days, 0.77 [17]; a rate of CAUTIs by number of beds, 1.09 to 2.13 [16]; or a CAUTI rate of 1% by full cohort and 2.4% in the ICU subgroup [28]. UTI-type coinfection in our work is much lower than UTI-type superinfection. We attribute this to the use of urethral catheters and a prolonged stay. Indeed, many or our patients were diagnosed with UTIs in the ICU; they were carriers of a urinary catheter, with a mean stay of 27 days, which predisposes them to greater use of bladder catheters and greater manipulation of the urinary tract. This observation is in agreement with other COVID-19 cohort studies in which the superinfection rates are higher than coinfection rates, mainly in patients admitted to the ICU [14,15,27,29]. The urinary catheter, CVC, ICU admission and a prolonged stay in the ICU are risk factors for healthcare-associated infections that were analysed in our work. These risk factors explain that all UTIs of our ICU patients were nosocomial infections. In addition, 38 patients of the superinfection group stayed in the ICU, 35 were CAUTIs, 37 had CVCs and 17 had a bloodstream infection. The lower rates of coinfections vs. superinfections were in agreement with the rates in the general population [27,29] but opposite to the findings of García-Vidal et al. [14] or those of Gudiol et al. [30], who observed a coinfection rate of UTIs of 27.7% and a superinfection rate of UTIs of 12.7% (in a large cohort of onco-haematological COVID-19 patients).
Women in our cohort presented a higher incidence of bacterial UTIs probably due to the anatomical predisposition. The demographic and clinical features in COVID-19 patients coinfected with UTIs presented older ages and comorbidities as in other studies [7,30]; however, in other work, only comorbidities had been associated [14]. It is difficult to make a comparison between the studies due to the heterogeneity between the populations studied and the scarcity of research on this topic. Notably, urethral catheterization and urological disease were risk factors for hospital-acquired UTI in COVID-19 patients by univariate analysis. In a retrospective study carried out in the USA [17], no increase in CAUTI was found. They attribute this to an augment in antimicrobial use producing the suppression of bacteriuria.
Differences of hospitalized patient populations versus prepandemic hospitalizations could help explain these results as well. To compare the prevalence of UTIs and microbiological aetiology in our patients, we used data from the study on the prevalence of nosocomial infections in Spain of our hospital for 2019 (EPINE-HULP) [31]. Regarding the prevalence of UTIs described in EPINE-HULP 2019, community-acquired UTIs were 19.2% (0.39% in our work) and hospitalized-acquired UTIs were 4.4% (3.42% in our work).
The frequency of microorganisms identified shows significant differences between our study and EPINE-HULP 2019: Enterococci spp. (E. faecalis 25 (26.3%), E. faecium 20 (21.1%)), followed by E. coli 27 (28.4%) and Pseudomonas aeruginosa 7 (7.4%), are the most common microorganisms in our study, while in EPINE-HULP 2019, the most frequent was Escherichia coli 11 (39.29%) followed by Pseudomonas aeruginosa 2 (7.14%), Klebsiella spp. 2 (7.14%) and Enterococcus faecalis 1 (3.57%). The increase in UTIs due to Enterococci has been described in other works such as Bardi et al. [15], where the most frequent microorganisms were E. faecalis and E. faecium, but not in the findings of García-Vidal et al. [14], whose main causative agent were Gram-negative bacilli. The increase of Enterococcus infections in COVID-19 patients has been analysed in different publications. DeVoe et al. observed that these patients had higher incidence of bloodstream infections (BSIs) due to Enterococcus, but not BSI in general [28]. They did not demonstrate a nosocomial transmission, so the mechanism underlying the increased rates of Enterococcus in COVID-19 patients requires further investigation. The use of broad-spectrum antibiotics such as B-lactams, mainly ceftriaxone, may have played a role in changing the microbiota through the selection of less frequent microorganisms such as Enterococcus. In addition, antibiotic pressure might select resistance mechanisms such as EBSL or Gram-negative carbapenemase-producing bacilli, as it was observed in our study. Moreover, some types of microorganisms such as Enterococci [32] or Proteus mirabilis [33] or certain virulence factors as mrk genes (type3 fimbriae) in E. coli and K. pneumoniae [34] are associated biofilm infection and might play the role in the CAUTI.
Our findings show an elevated leukocyte count with neutrophilia mainly in patients with urinary tract superinfections, a common condition in the case of bacterial infection. This fact is probably related to the greater number of patients admitted to the ICU and more seriously ill. Biomarkers such as C-reactive protein and procalcitonin were elevated in all cases, although the cytokine storm associated to SARS-COV-2 makes it difficult to interpret these data. In the COVID-19 patient series, they do not appear as risk variables associated with coinfections and superinfections [15,28].
Limitations of our study include being a retrospective and observational study in a single centre. The diagnostic tests obtained and the procedures performed were conditioned by the daily care without a standardized protocol and a small number of patients limited by period of study. In addition, these results were not corrected for multiple testing, highly increasing the risk of type I errors. Our patients were treated by groups of healthcare workers sometimes not very accustomed to caring for severely ill patients. Furthermore, during the first wave of the pandemic, there were continuous changes in the protocols for the management of COVID-19 patients that may have conditioned prognosis. The impact our local epidemiology and antimicrobial resistance may limit the generalizability of our findings.

Conclusions
UTIs are infrequent in hospitalized COVID-19 patients; however, when they emerge, they can condition the prognosis. E. faecalis or E. faecium are the most frequent microorganisms identified, followed by E. coli. It is necessary to maintain and develop measures for the prevention of CAUTIs and to assess the true need to maintain the use of urinary catheters and administer antibiotic treatment to avoid infectious complications in this type of patient.

Informed Consent Statement: Not applicable.
Data Availability Statement: The datasets generated during and/or analysed during the current study are not publicly available but are available from the corresponding author on reasonable request. Service, the Infectious Diseases Unit and the Clinical Microbiology Service of the La Paz Hospital for their dedication to these patients. B.D.P. acknowledges E. Cano for his support in the preparation of the manuscript.

Conflicts of Interest:
The authors declare no conflict of interest.