INTRODUCTION

COVID-19 is an emerging viral disease initially reported in Wuhan-China at the end of 2019, rapidly expanding and, in months, collapsed global health systems ¹. Regardless of the speed with which the causative agent, a beta coronavirus, was identified and sequenced and what this meant for developing the first vaccines, compassionate drug use and the abuse of antibiotics were common global scenarios ^2,3.

The prescription and self-medication of antimicrobials, many of them for exclusive and restricted hospital use, seeking to reduce the mortality and morbidity associated with this disease, accelerated the selective appearance of multi-resistant bacteria ⁴.

These bacteria generated significant phenotypic and antibiotype changes in global hospital epidemiology, which propelled a revolution in the massive use of empirical and targeted (directed) antibiotic therapy, using multiple schemes and with unusual antibiotics such as polymyxins and carbapenems on a large scale ⁵. In addition, given the high rates of multi-resistance, new broad-spectrum antibiotics were required. However, they were not available in the lists of essential drugs in the public system and were particularly expensive. This kind of antibiotic therapy made the availability of the drugs the exception and not the rule in lowincome centers.

With the advent and mass use of vaccines and the use of drugs such as steroids, specific antivirals, and monoclonal antibodies that have demonstrated their efficacy in extensive clinical studies, COVID-19 is on the brink of becoming endemic, and its morbidity and mortality are minor, despite the appearance of more effective variants in its transmission ⁶.

Falcone et al. ⁷ have reported the results from a nationwide multicenter study in Italy, performed with 1276 patients with bacteremia due to Gram-negative bacilli, in which they observed a marked expansion of resistance to carbapenems and other molecules, new resistance mechanisms, and associated high mortality.

This work aims to present the epidemiological, clinical, and microbiological profiles of a series of patients with bacterial superinfections hospitalized in a center that exclusively treated COVID-19 in Quito, Ecuador.

MATERIALS AND METHODS

RESULTS

Two hundred forty patients with a median age of 55 and a predominance of men (68.75%) were analyzed. The median hospital stay was 24 days. When comparing age by the discharge condition, significant differences were observed with p= 0.002, with medians of 52.5 years in survivors vs. 58.5 years in non-survivors. The hospital stay presented significant differences by discharge condition with p<0.001, with the median stay being 29 days in survivors vs. 16 days in non-survivors (Table 1).

Table 2 reports the analytical and cytometric values, which indicate that although the elevation of biomarkers characteristic of severe forms of the disease is striking, such as the white count, C-reactive protein, D- dimer, and ferritin, there were no significant differences when compared to the condition at discharge.

Regarding associated infections, ventilator-associated pneumonia (VAP) was observed more frequently in 55% of cases, followed by catheter-associated urinary infections (CAUTI) at 43.75%, central line-associated bloodstream infection (CLBSI) 31. 25%, and colonization at 15.42%, among others. The CLBSI presented significant differences by discharge condition with a p of 0.030; the proportions were 26.03% in survivors vs. 39.36% in non-survivors (Table 3).

Among the most frequent bacteria causing superadded infections in our study, KPC-type carbapenem-producing Klebsiella pneumoniae complex was observed in 24.17% of the cases, followed by Klebsiella pneumoniae producing extended-spectrum β-lactamase (ESBL) in 17.92%, and carbapenem-resistant Pseudomonas aeruginosa in 13.75%, among others. Serratia marcescens presented significant differences by discharge condition with a p of 0.013, the proportion of this bacterium being 6.16% in survivors vs. 0% in non-survivors (Table 4).

Among the most used antibiotic schemes, meropenem + vancomycin + amikacin was observed in 56.09% of the empirical scheme and 38.55% of directed schemes. In the directed scheme, significant differences were observed by discharge condition with a p = 0.036, specifically for colistin + others, whose proportions were 16.95% in survivors vs. 41.67% in non-survivors (Table 5).

Multivariate logistic regression analysis was performed to determine the relationship between age, CLBSI, and antibiotic scheme with mortality, observing the following: age was related to mortality with p = 0.006, whereas due to the one-year increase in patients, the risk of mortality of patients was increased by 3%.

CLBSI infections were related to mortality with a p of 0.028, whereas patients with CLBSI infections were 1.94 times more likely not to survive (Table 6).

Among the directed (targeted) antibiotic schemes, it was observed that Colistin + others were related to mortality with a p = 0.028, where patients treated with these antibiotics presented a 12 times greater probability of not surviving compared to those who received aminopenicillins + others (Table 7).

The enzymatic resistance of these bacteria was studied. 40% of the K. pneumoniae complex and E. coli isolates were ESBL producers. In addition, one out of every five K. pneumoniae complex isolates was a KPC producer. The resistance observed to carbapenems by P. aeruginosa is very high: resistance to carbapenems was found in seven out of ten isolates. S. aureus complex and S. epidermidis showed 30 and 80% resistance to methicillin, respectively (Table 8).

DISCUSSION

This study, conducted on 240 hospitalized patients with COVID-19 who presented with bacterial superinfection, has shown that older male patients with bacteremia who received treatment with antibiotic regimens containing colistin were associated with lower survival.

Of these factors, age, bloodstream infection, and colistin use were independent factors for mortality when adjusted in a multivariate model. Furthermore, it was also observed that VAP was the most frequent site of superinfection; and that enzymatic resistance due to ESBL and KPC-type serine beta-lactamases was frequent in enterobacteria. The most used empirical scheme was composed of carbapenems, aminoglycosides, and glycopeptides (meropenem, amikacin, and vancomycin), and the presence of carbapenem-resistant P. aeruginosa was relevant.

Patients with severe COVID-19 admitted to a hospital are at greater risk of developing infections during their hospital stay ^8-11, and this risk increases as their hospital stay is prolonged ¹². This increased risk of superinfection is due to greater exposure of patients to immunosuppressive treatments (especially steroids) and also for reasons specific to health centers, of a structural type (opening of new ICU beds in other areas of the hospital), organizational (hiring or transferring personnel without prior training in the management of critical patients) and functional (changes in patient care standards, and prolonged use of personal protective equipment). In addition, these patients require vascular access, urinary devices, and invasive mechanical ventilation due to their condition, increasing the risk of infections. These hospital infections are more complex than usual because they are associated with bacteria with resistance to antibiotics¹³.

The documented evidence demonstrates that male patients with COVID-19 are at greater risk of presenting more severe forms of COVID-19, with prolonged hospitalizations and more extended use of invasive devices, which increases the risk of superinfection,⁵ according to the findings of our study. The mean age in the sample was 55 years, with age ranges between 50 and 60 years, similar to the data reported by other authors ^5,10,14. In addition, this data was associated with higher mortality: the increase in one year of age increases mortality risk by 3%, agreeing with what has been reported in the literature ^5,7,10.

The average hospital stay was 24 days, less in the non-survivor group, with a statistically significant difference. As a possible hypothesis, it is proposed that they were patients with more severe diseases that coursed with more torpid and abrupt evolutions, consistent with what has been reported by some authors ^10,12,15,16.

The laboratory data (Table 2) show that the two groups (survivors and deceased) generally presented similar patterns of acute inflammatory reaction, state of hyper-inflammation, and hypercoagulability, with no statistically significant differences. It is suggested that both groups were patients admitted with severe disease with systemic involvement. From the point of view of the severity of the condition caused by COVID-19, they were similar groups.

VAP was the most common infection reported in this investigation, representing 55% of infections. In other studies, it has been reported that VAP represented between 40% and 60% of documented superinfections in patients with COVID-19. The highest predisposition to these processes has been associated with prolonged ventilation, prone position, use of corticosteroids, lung damage, and episodes of cross-contamination, most likely due to the prolonged use of personal protective equipment. In addition, the same COVID-19 infection can cause inflammation of the pulmonary vascular endothelium and subsequent thrombosis, creating an environment conducive to bacterial growth ^17-19.

The incidence of CLBSI tripled during the pandemic ²⁰. In this study, patients with this type of infection were 1.94 times more likely not to survive. Other authors have reported higher mortality from CLBSI in patients with COVID-19 (between 40 and 54%) than from this type of infection in pre-pandemic or non-COVID-19 patients (between 24 and 33%) ^20-22. This higher mortality may be due to the use of these devices for a longer time, the prone position that limited the care of the line, and the breach of asepsis and antisepsis when manipulating the catheter in the harsh conditions faced during the health emergency, generating bacteremia more frequently with a more significant inoculum.

K. pneumoniae complex was the most frequently recovered bacterium in cultures, followed by P. aeruginosa and E. coli. Generally, the reports in the literature vary widely because of local microbiology. However, most of the studies at a global and regional level, carried out during the pandemic, agree with this research, showing that these bacteria are the most frequent ^5,8,15-17,23.

In our study series, a high percentage of Enterobacterales were ESBL producers, and one in five K. pneumoniae complex isolates were KPC-type carbapenemase producers, consistent with other reports in Latin America during the pandemic ²⁴. Most recovered Pseudomonas cultures reported resistance to carbapenems (70%). These findings are worrying due to the impact on public health and the limitation of the therapeutic arsenal against such microorganisms, especially in countries such as Ecuador ²⁵.

In non-randomized clinical studies, using colistin as monotherapy or combination therapy has mixed results; however, it presents higher mortality in patients who previously received carbapenems ^26-28. In the IDSA guidelines, the use of colistin as targeted therapy is discouraged, giving preference to monotherapy with new antibiotics such as ceftazidime-avibactam ²⁹. Patients who received colistin (combined with other antibiotics) were associated with higher mortality in our series. We attributed this finding to prior treatment with carbapenems and subsequent resistance to them, the side effects of colistin and associated antibiotics, and the fact that they presented a more severe disease. Moreover, this higher mortality could occur because colistin was indicated as a desperate measure since there were no better therapeutic options available in the hospital or because its susceptibility to the microorganism could not be confirmed since the sensitivity cut-off points had not been established with an antibiogram for the said antibiotic, against certain bacteria and/or in some tissues.

On the other hand, it should be considered that the hospital in this study is located almost 3000 meters above sea level. Although the effect of high altitude (≥1500m) and its potential association with mortality from COVID-19 is still controversial, there are studies in this direction ^30,31; therefore, more research is required in this regard and its impact on the high mortality of severely ill patients with bacterial superinfections.

This work has significant limitations. It is monocentric and retrospective, with the biases inherent in this design; furthermore, there was no control group. There could also be a possible underdiagnosis of fungal infections, which were not reported in this series and constitute a clinical challenge, especially in the context of patients infected with SARS-CoV-2. Finally, the health center needed molecular biology studies to determine resistance genes. Nevertheless, in the case of the K. pneumoniae complex isolates, in those cases where the automated system reported strains probably producing KPC, lateral flow tests were carried out, and the strains were sent to the national reference laboratory in Ecuador: Instituto Nacional de Investigación en Salud Pública “Dr. Leopoldo Izquieta Pérez” (INSPI). Molecular biology studies were performed at the institution for confirmation, and the hospital was notified.

In the present retrospective study, we found that in patients with COVID-19 and bacterial superinfection, age, central venous catheter-associated bacteremia, and colistin use were independent predictors of mortality. The most frequently isolated microorganisms were ESBL - and KPC-producing Enterobacterales and non-fermenting Gram-negative bacilli resistant to carbapenems.

These findings are consistent with what has been reported in world and regional series and emphasize the urgent need to establish programs for the rational use of antibi-, especially in countries with a limited therapeutic arsenal.