Epidemiology of nosocomial carbapenem-resistant Klebsiella pneumonia bloodstream infections in south China: factors related to the patient mortality

Background: Carbapenem-resistant Klebsiella pneumoniae (CRKP) causing Bloodstream infection (BSI) are associated with high rates of mortality . Nevertheless, only a few studies regarding the epidemiology of CRKP BSI in south China. The purpose of this study was to describe the epidemiology, clinical characteristics, and the mortality of risk factors associated with CRKP causing bloodstream infection. Methods: A retrospective study of patients with CRKP BSI was recruited from teaching hospital in south China from January 2016 to December 2018. Clinical data were collected from medical records. Results: In total, 90 patients with CRKP BSI were enrolled in the study, while 57% (51/90) of the CRKP BSI were obtained from ICU. Most CRKP BSIs originated from hospitals (81; 85%), while the rest (9; 10%) were healthcare-associated . In univariate analysis, gastrointestinal hemorrhage (p=0.029), Pitt bacteremia score (P=0.045), Charlson comorbidity index (p=0.018) and Corticosteroids use (p=0.036) and Septic shock (p=0.001) were associated with the risk factors for mortality. In a multivariate analysis, septic shock (adjusted odds ratio [aOR] 5.591, 95% confidence interval [CI] 1.405-22.246, P=0.015) and Corticosteroids use (aOR 4.148, 95% CI 1.331-12.928, P=0.014) were independently predictors of mortality. Conclusion: Our data showed that the morbidity and mortality of CRKP BSIs patient from ICU and non-ICU was no significant difference. Standardizing operation and improving nurse quality may play an important role in CRKP BSI patient in intensive care unit. Septic shock and Corticosteroids use were the independent factors of CRKP BSI patient mortality. However, the study did not show an association between invasive procedures and the development of CRKP BSI.


Introduction
Klebsiella pneumoniae isolates are among the most common bacteria causing hospital acquired infections, including bloodstream infections (BSIs) [1,2]. Carbapenem-resistant K. pneumoniae (CRKP) was originally identified in 1996 and quickly became a serious health care problem in many countries, including China [3,4]. CRKP can cause many serious infections, including, soft tissue infections, pneumonia, urinary tract infections, intraabdominal and bloodstream infections. Bloodstream infection (BSI) with carbapenemresistant Klebsiella pneumoniae (CRKP) is related to particularly high mortality rates [5].
The mortality rate of carbapenem-resistant Klebsiella pneumoniae (CRKP) infections were reported to be 50.06%, 46.71%, 33.24%, and 44.82% in Europe, South America, North America, and Asia respectively [6]. Indeed, all-cause mortality from critically ill patients with severe bloodstream CRKP infections is nearly 70% [7][8][9]. Emerging resistance to antibiotics of last resort (i.e. tigecycline or colistin) leaving very few therapeutic options have been shown in recent studies [10,11]. Increasing antimicrobial drug resistance, especially in carbapenem-resistant Klebsiella pneumonia (CRKP), can limit the choice of antibiotics used for the treatment of infectious diseases and further poses a negative impact on patient outcome. The high occurrence of cases and mortality, economic burden and dearth of alternative drugs makes CRKP infections an important threat to public health worldwide.
There are several risk factors associated with the development of CRKP BSI. Risk factors related to CRKP BSI include comorbidities, ICU stay, use of invasive devices, and exposure to carbapenem prior to culture [12]. Due to the extensive use of invasive surgery and the frequent use of antimicrobial agents, ICUs are described as a factory for generate, spread, and expansion of antimicrobial resistance. While regional and institutional infection control programs remain the cornerstone for reducing the risk of patient acquisition and death from MDR pathogens, additional tools are needed to facilitate patient screening, early diagnosis and timely use of appropriate antibiotic treatment regimens. The physician may consider this risk assessment when deciding whether a particular patient should receive early appropriate therapy for CRKP BSI until culture and sensitivity results are available.
To the best of our knowledge, few studies have assessed the hazard factors for mortality of CRKP bloodstream infection (BSI) in south China. In this study, we sought to evaluate the epidemiological, clinical characteristics, the risk factors for mortality of CRKP BSIs in patients admitted to a teaching hospital in Wenzhou, China.

Study design and patients
From January 2016 to December 2018, we retrospectively analyzed clinical data in the First Affiliated Hospital of Wenzhou Medical University, a tertiary university hospital with 4,100 beds in Zhejiang, southern China. We included all adult patients (age of day≥18 years) hospitalized in the First Affiliated Hospital of Wenzhou Medical University in south China. BSI onset was defined as the collection date of a positive blood culture. Only the first CRKP BSI was included, and patients with polymicrobial BSI were excluded. All criteria are graded within 48 hours before or on the day of first positive blood culture. Laboratory examination included lymphocytes (10*9/L). The study was approved by the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University. Because the retrospective study did not cause harm to the patient, informed consent was waived.

Data collection and definitions
The data of the clinical microbiology laboratory collected included information regarding demographic characteristics, comorbidities, intravascular catheter use, recent (30 days) surgical procedures, the use of Corticosteroids (use Corticosteroids during bloodstream infections), carbapenem exposure (exposure to carbapenems during the 30 days preceding admission to the ICU) and patient outcomes were also collected(Table1). BSIs patients were further classified into death group (n=60) and survivor group (n=30) according to clinical outcome of antimicrobial treatment. BSIs were defined as hospitalacquired if the positive blood culture results more than 48h after the admission of patient to the hospital and no symptoms of infection had been noted at admission. If blood culture samples had been collected within 48h after hospital admission, the isolate was defined as Community-acquired or healthcare-associated [13]. Septic shock was defined as sepsis associated with organ dysfunction and persistent hypotension despite volume replacement [14]. Comorbidities were calculated according to the Charlson's score [15], and the Pitt bacteremia score (PBS) was used to assess disease severity at the time of positive blood culture [16].

Microbiological methods
We processed bacterial cultures in the clinical microbiology laboratory. Initially, Vitek2 system (BioMèrieux, France) was used to determine bacterial identification and the antimicrobial susceptibility. Then, the isolates were stored in frozen condition at −80•C with 30% glycerol. The minimum inhibitory concentration (MIC) values for tested antimicrobial agents were determined by an automated broth dilution method and interpreted by the recommendation of the European Committee on Antimicrobial Susceptibility Testing (EUCAST, 2015).

Statistical analysis
Statistical analysis was carried out by SPSS for Windows, version 21.0 (IBM, Armonk, NY, USA).
The mean and standard deviation (SD), median and interquartile range (IQR) were used to compare the continuous variables of normally and non-normally distributed, respectively. Students't test (normal distribution) and Mann-Whitney U-test (non-normal distribution) were used to analyze and compare continuous variables. Categorical variables are represented by numbers and percentages. The categorical variables were analyzed by Fisher's exact test. Logistic regression analysis was used to determine the risk factors of death in each variable. In univariate analysis, the significant variables with P < 0.05 were included in multivariate analysis, in which we evaluated the risk factors for CRKP BSIrelated mortality. A p value <0.05 was considered to statistically significant.

Characteristics of patients
Overall, 90 patients with CRKP BSI met the inclusion criteria. Ultimately, according to the inclusion and exclusion the unique episodes were included in the analysis (Fig 1). The demographic and clinical characteristics of patients with CRKP-BSIs are presented in Table   1. CRKP-BSI cases mainly occurred in subjects aged over 60 years (72%). The median patient age was 68 years (interquartile range 21-93 years), and 64 of 90 patients (67 %) were males. All episodes of CPKP BSIs were either hospital acquired (85%) or healthcare associated (10%), while none was community acquired.

Increased rates of CRKP-BSIs, including ICU and Non-ICU patients
In our study, 90 unique episodes of CRKP-BSI were included, 57 % (51/90) of the CRKP isolates were came from ICU and 43% (39/90) of the CRKP BSI patients were came from non-ICU. We discovered that the occurrence of bloodstream infection causing nosocomial carbapenem-resistant Klebsiella pneumonia is increasing since 2016, and the rates of

Risk factors for mortality of CRKP BSI patients
Overall in-hospital crude mortality was 65% (59/90), similar to the previous studies[17].
Our results indicate that the invasive Procedures was indwelled Urinary catheter (69%), indwelled central venous catheter (72%), mechanically ventilated (45%), gastrointestinal decompression (63%), renal replacement therapy (14%) ( Table 1). Previous studies have evaluated several risk factors for the mortality associated with CRKP BSI, including use of indwelled Urinary catheter, indwelled central venous catheter, mechanically ventilated and exposure to carbapenems. However, in our analysis, gastrointestinal hemorrhage (p=0.029), Pitt bacteremia score (P=0.045), Charlton comorbidity index (p=0.018) and Corticosteroids therapy (p=0.036) and Septic shock (p=0.001) were associated with risk factors for mortality (Fig 4). In univariate analyses, variables significant at p < 0.  CRKP-BSI cases mainly occurred in aged over 60 years (72%) and more frequently in males (67%) than in females. Confirmation of this finding is also available in the European Centers for Disease Control and Prevention database, where the higher proportion of carbapenem-resistant K. pneumoniae in males is particularly evident in countries at high prevalence of carbapenem resistance such as Italy and Romania. The reason for the higher incidence of CRKP-BSI in males than in females is not clear, but it is worth noting that the higher incidence of sepsis in males is associated with genetic susceptibility [19]. In our study, we observed an increase in CRKP-BSI with age. This increase, especially for patients 60 years of age and older, may be associated with the high vulnerability in the elderly[20], which may determine higher frequencies and longer hospital stays than younger patients.  Fig-4 shows that CRKP BSI was related to gastrointestinal hemorrhage, Pitt bacteremia score, Charlton comorbidity index and Corticosteroids use and Septic shock. Septic shock and Corticosteroids use were independent risk factors for CRKP BSI, which is different with previous reported. The reason for this might provide an explanation that the regional differences and effective Hospital Infection-Control of CRKP BSI in our study.
Corticosteroid should be added to septic shock patients according to the Surviving Sepsis Campaign guidelines [30]. In our study, an alarming finding is that the use of corticosteroids as adjunctive therapy in CRKP infections has deleterious effects. The most important risk factor for mortality in our study was septic shock at CRKP BSI onset; not surprisingly, this variable had been found to be associated with mortality in several previous studies performed on general population patients with CRKP BSI[8, 31].
There were several limitations associated with this study. First, the current study was a retrospective, single-center study including 90 patients with clinical analysis. Second, Clinical data were obtained retrospectively from medical records, and there may be some differences in physician practices or accuracy of information. Finally, prospective and further multi-centric studies are needed to confirm our findings.

Conclusion
There was no significant difference in CRKP BSIs patient morbidity and mortality between ICU and non-ICU. Standardizing operation and improving nurse quality may play an important role in CRKP BSI patient in intensive care unit. Septic shock and Corticosteroids therapy were independent risk factor related to the patient mortality. However, the present study did not show an association between invasive procedures and the development of CRKP BSI, which is different with previous studies reported. The