Carbapenems, a class of broad-spectrum beta-lactam antibiotics, are the primary treatment choice for severe infections caused by multidrug-resistant Enterobacteriaceae. Carbapenem-resistant Klebsiella pneumoniae (CRKP) was first identified in northeastern USA in the early 2000s (Tzouvelekis et al., 2012); however, with the increasing use of carbapenems in hospitals, the prevalence of CRKP is increasing worldwide and has constituted a serious threat to public health (Bonomo et al., 2018; David et al., 2019; Reyes et al., 2019). In Europe, the resistance rate to CRKP has been reported to be as high as 39.7% (David et al., 2019). In Eastern Mediterranean region, Klebsiella pneumoniae (K. pneumoniae) accounts for over 50% of infections (Prestinaci et al., 2015). According to data from the China Antimicrobial Surveillance Network (CHINET), the resistance rates of K. pneumoniae to imipenem and meropenem have increased steadily from 3 and 2.9% in 2015 to 24.4 and 25% in 2023, respectively (CHINET, 2023). In 2008, the World Health Organization (WHO) classified CRKP as a critical priority on the global priority list of antimicrobial-resistant bacteria (Tacconelli et al., 2018). In 2009, the Centers for Disease Control and Prevention (CDC) classified 18 common multidrug-resistant bacteria into three categories, designating CRKP as an urgent threat to human health (CDC., 2019). The mortality of patients with CRKP bloodstream infection (CRKP-BSI) is significantly high (Xu et al., 2017; Zhu et al., 2020), and BSI has become a major global medical burden (Markwart et al., 2020). The incidence of CRKP-BSI can increase mortality and prolong the length of hospitalization (Schwaber et al., 2008; Stewardson et al., 2019), which is an important clinical threat.

Currently, carbapenemase production is the primary mechanism of resistance in CRKP, with carbapenemases classified into class A (represented by KPC), class B (e.g., NDM, VIM, and IMP), and class D (primarily OXA-48). Overexpression of efflux pumps and loss of outer membrane porins also contribute to the development of resistance. Klebsiella pneumoniae carbapenemase (KPC) has emerged as the most prevalent carbapenemase in the United States and Europe (Grundmann et al., 2017; Temkin et al., 2014), and KPC-2 (64.6%) has become the most common carbapenemase in CRKP in China (Han et al., 2020). Plasmid-encoded carbapenemases often carry additional genes that confer resistance to other antibiotics, which can easily spread among bacteria, resulting in outbreaks and limited treatment options. Only a few antibiotics, such as tigecycline and polymyxins, remain effective against CRKP. Although novel antibiotics, such as ceftazidime-avibactam, have emerged for the treatment of CRE infections, they remain ineffective against New Delhi metallo-beta-lactamase (NDM)-producing CRKP (Abe et al., 2022; Chen Y. et al., 2022). Therefore, understanding the risk factors associated with the causes and outcomes of CRKP-BSI is required to reduce its incidence and initiate appropriate therapies.

Several studies have investigated the risk factors for CRKP infection. However, some of these studies employed a case-control design, comparing a CRKP group with a carbapenem-susceptible Klebsiella pneumoniae (CSKP) group. Such a case-control design has an intrinsic flaw as it is based on the “replacement scenario” assumption, in which each patient is assumed to be infected either by antimicrobial-resistant organisms or antimicrobial-susceptible counterparts; however, the patient may be infected by other pathogens or not infected at all. CSKP infections account for only a small proportion of hospitalized cases, making them unrepresentative (Evans and Harris, 2017; Harris et al., 2001; Kaier and Frank, 2013). Therefore, an improved design is required to further elucidate the risk factors for CRKP infection. In this study, we employed a case-case-control design to evaluate the risk factors for CRKP-BSI, providing a foundation for effective control strategies.

This study provides important insights into the characteristics and outcomes of CRKP in BSIs. For studies on risk factors for CRKP infection, the selection of an appropriate control group is critical to generate correct conclusions. Several studies selected the CSKP group as the control group, which can introduce statistical selection bias. CSKP represents only a small proportion of nosocomial infections and may not be representative of all infections. In addition, certain antibiotics may inhibit or eliminate certain strains of CSKP, while not affecting CRKP, which is another important bias. Consequently, the frequency of antibiotic use in CSKP strains may decrease, indirectly amplifying the risk factors for CRKP (Ng et al., 2014). Alternatively, a few studies have only included patients who were not infected with CRKP as controls but did not distinguish between susceptible and resistant groups (Cronin et al., 2017; Giacobbe et al., 2017; Micozzi et al., 2017). However, the risk factors for KP infection cannot be analyzed in such a scenario. Therefore, we performed a case-case-control study that included matched control from inpatients to identify the risk factors for CRKP/CSKP infection and the 30-day mortality of CRKP, thereby guiding the prevention and appropriate hospital treatment strategies.

In our study, we determined that CRKP-BSI was primarily concentrated in the ICU, followed by the hematology and emergency departments in the general medical department. Nosocomial infections are readily acquired in the ICU due to airborne and contact transmission of CRKP in this confined environment (Jiao et al., 2015). The transmission of CRKP is facilitated in the ICU by frequent use of medical equipment and invasive procedures. Patients in the ICU have serious underlying diseases and may be treated with broad-spectrum antibiotics, which also contribute to the induction of CRKP-BSI. Notably, hematology patients present unique vulnerabilities that increase the risk of CRKP-BSI, including long-term hospitalization, use of high-grade antibiotics, and an impaired immune response (Chen J. et al., 2022; de Souza et al., 2024). Patients with various infectious diseases are often treated in the emergency department, where crowdedness with patients, heavy workload of medical staff, and inadequate implementation of infection control measures are common problems that may lead to cross-transmission of CRKP in the emergency department.

The findings of this study showed that antibiotic resistance was more pronounced in the CRKP-BSI group than that in the CSKP-BSI group. However, these strains remained relatively susceptible to colistin and tigecycline. Considering the above results and individual clinical conditions, combination therapy is recommended as the optimal treatment approach for CRKP-BSI.

The statistically significant factors identified in multivariate analyses of both the CRKP-BSI and CSKP-BSI groups were risk factors for K. pneumoniae-BSI. The factors that were statistically significant in the multivariate analysis of CRKP-BSI but not of CSKP-BSI were risk factors for CRKP-BSI. In contrast, factors that were statistically significant in the multivariate analysis of CSKP-BSI but not of CRKP-BSI were risk factors for CSKP-BSI (Kaye et al., 2005). We identified that exposure to antibiotics (carbapenems and tigecycline) and glucocorticoids were risk factors for CRKP-BSI. One risk factor, liver disease, was a risk factor for K. pneumoniae-BSI. Patients with liver diseases have altered enteric flora and impaired immune systems, which increase the risk of K. pneumoniae-BSI (Hsu et al., 2021; Lameirão Gomes et al., 2019). In addition, patients with liver diseases require invasive procedures for diagnosis and treatment, which can exacerbate the occurrence of nosocomial infections.

Consistent with the results of previous studies (Cao et al., 2022; Xiao et al., 2020), our study showed that prior carbapenem use was an independent risk factor for CRKP-BSI. Carbapenemase acquisition is the primary cause of carbapenem resistance and occurs in across diverse departments (Zhang et al., 2021). Han et al. showed that the most prevalent carbapenemase gene in CRKP in China is blaKPC-2, followed by blaNDM (Han et al., 2020), and carbapenem exposure may induce the emergence of resistance-conferring genes. In contrast to other studies, our study determined that tigecycline exposure was also a risk factor for CRKP-BSI, indicating that CRKP-BSI can be induced by carbapenem as well as by other drugs. A previous study suggested that overexpression of the AcrAB efflux pump, induced by activated MarA, plays an important role in the emergence of CRKP isolates (Park et al., 2020). The use of antibiotics within 3 months before a positive blood culture indicated that the patient may have had repeated infections and antibiotic exposure, thereby increasing the risk of antibiotic resistance development. Thus, there is a need to strengthen the management of antibiotics in inpatients. Treatment with high doses for appropriate durations is a better way to limit the risk of infection. Consistent with previous studies (Huang et al., 2023), we showed that glucocorticoids use is a risk factor for CRKP-BSI, possibly because patients with CRKP-BSI have serious underlying diseases and are treated with glucocorticoids. Second, the use of glucocorticoids destroys the intestinal microenvironment, kills susceptible strains, and promotes the overgrowth of drug-resistant strains, thereby promoting a shift from opportunistic pathogens to pathogenic bacteria. Therefore, more attention should be paid to the initial appropriate antibiotic therapy for patients with CRKP-BSI. In addition, strengthening antimicrobial stewardship programs and regular resistance surveillance is essential to avoid unnecessary antibiotic exposure.

When assessing the potential impact of CRKP-BSI on 30-day mortality, we determined that patient groups, respiratory diseases, use of 3rd-generation cephalosporins, mechanical ventilation, and central venous catheter insertion were independent risk factors for 30-day mortality. However, a shorter length of hospital stays was a protective factor against 30-day mortality. Various studies have reported that mortality in patients with CRKP-BSI ranged from 24 to 65%, and some studies have shown that CRKP infection is an independent risk factor for mortality (Cao et al., 2022; Chen J. et al., 2022; Li et al., 2020; Liu et al., 2019; Schwaber et al., 2008). Similarly, in our study, patients with CRKP-BSI exhibited higher mortality and longer hospital stays.

Our study showed that patients with nosocomial CRKP-BSI were often diagnosed with respiratory diseases, which has been reported in previous studies (Strich et al., 2020). Moreover, several studies have also shown that chronic obstructive pulmonary disease (COPD) is a risk factor for CRKP infection (Liu et al., 2012; Nouvenne et al., 2014; Salsano et al., 2016). Patients with respiratory diseases are more likely to be infected with or colonized by CRKP in the airway, and patients with respiratory diseases requiring mechanical ventilation or invasive procedures are susceptible to CRKP-BSI. Under the selective pressure of 3rd-generation cephalosporin use, susceptible strains are eliminated, thereby facilitating the survival of resistant strains. Mechanical ventilation and central venous catheter insertion were independent risk factors for 30-day mortality in patients with CRKP-BSI. The most likely reason for this is that invasive procedures can damage the mucosa and increase the incidence of CRKP-BSI because the majority of bacteria can pass through the mucosal barrier to the bloodstream (Li et al., 2020). A shorter LOS is a protective factor against 30-day mortality. Patients with a reduced LOS have a lower risk of CRKP infection due to decreased contact time with the hospital environment, medical staff, and other patients, which is consistent with the findings of previous studies (Correa et al., 2013; Gallagher et al., 2014). The longer the hospital stays, the greater the risk of infection from colonizing bacteria. Therefore, judicious use of 3rd-generation cephalosporins in antibiotic stewardship programs, reduction of invasive procedures, and prevention of CRKP-BSI are effective strategies to reduce 30-day mortality.

Not surprisingly, arterial catheterization was a risk factor for in-hospital mortality in patients with nosocomial CRKP-BSI, which is consistent with previous studies emphasizing the importance of aseptic operations in patient care, and that invasive procedures can damage the mucosa and increase the chance of CRKP entering the bloodstream (Wang et al., 2018). Maintaining aseptic technique is an important strategy for CRKP-BSI prevention.

Recently, an increasing number of studies have focused on active screening for CRKP to enable early detection, isolation, and intervention in high-risk departments and patients. The WHO and the US CDC recommend screening for CRE based on local epidemiological conditions (United States Department of Health and Human Services, 2015; World Health Organization, 2017). In our hospital, nosocomial CRKP-BSIs primarily occur in the ICU and in patients with specific risk factors, suggesting that higher-risk departments and patients should be actively screened for CRKP.

Therapeutic options for CRKP-BSI in our hospital include these based on monotherapy or, in combination, tigecycline, colistin or ceftazidime/avibactam. Outcomes of patients were not altered by different types of antimicrobial therapy in univariate analysis, which was consistent with previous studies (Xiao et al., 2018). This is mainly attributes to the optimal treatment option for CRKP infections is still not well established. The causes of death in patients with CRKP-BSI are complex. Clinicians should not only consider anti-infective treatment, but also consider the patient’s own conditions, including the patient’s underlying diseases and organ functions. The prompt identification of risk factors for mortality in patients infected with CRKP-BSI is essential for early prescription of adequate empirical antibiotic therapy, which was a key factor for mortality reduction.

The present study has several limitations. First, this was a single-center study with a small sample size. Therefore, it is imperative to conduct prospective, multicenter, large-sample clinical trials to validate our findings. Second, this was a retrospective cohort, and potential information and selection biases were unavoidable. Finally, the clinical microbiology laboratory of our hospital does not perform routine detection of carbapenemase production for CRKP. We cannot exclude the possibility that some carbapenemase-producing isolates were susceptible to carbapenems and may have been classified as CSKP in this study. We also could not assess whether any outbreaks occurred during the study period and whether different drug resistance mechanisms resulted in different clinical outcomes. Future studies should include molecular epidemiological investigations.

This study aimed to elucidate the predictors of CRKP-BSI and help clinicians better identify CRKP-BSI at an early stage. In our hospital, CRKP-BSI exhibited a focal distribution across specific departments, such as the ICU, hematology department, and emergency department, highlighting the need to implement interventions such as hand hygiene protocols and environmental disinfection measures to limit CRKP transmission. The identification of risk factors for CRKP-BSI infection and mortality are critical for preventing and controlling CRKP. Our results showed that liver disease is an independent risk factor for K. pneumoniae-BSI, and that exposure to antibiotics (carbapenems and tigecycline) and glucocorticoids are risk factors for CRKP-BSI. central venous catheter insertion and mechanical ventilation use were independently associated with 30-day mortality. Therefore, patients with CRKP-BSI, respiratory diseases, use of 3rd-generation cephalosporins, mechanical ventilation, central venous catheter insertion, and arterial catheter use are more likely to have a poor prognosis. Close attention should be paid to patients who have used carbapenem and tigecycline over an extended period. Moreover, outcomes of patients were not altered by different types of antimicrobial therapy, the prompt identification of risk factors for mortality was a key factor for mortality reduction. The rational use of antibiotics, reduction in the length of hospital stays and reducing the frequency and duration of invasive procedures are effective measures to reduce CRKP-BSI and 30-day or in-hospital mortality (Xiao et al., 2018; Yuan et al., 2020). These findings provide valuable insights for the development of anti-infective therapy, infection prevention and the control strategies for CRKP-BSIs.