Active surveillance of carbapenem-resistant gram-negative bacteria to guide antibiotic therapy: a single-center prospective observational study

Carbapenem-resistant gram-negative bacteria (CRGNB) have become a public health concern worldwide. The risk factors associated with CRGNB infection after colonization are unknown, nor is the optimal timing of antibiotic treatment, warranting further investigation. A 4-year single-center prospective observational study was conducted. CRGNB-colonized patients were incorporated on admission into our observation cohort for an active surveillance culture program, and analysis of risk factors associated with infections after CRGNB colonization was performed. We divided patients into empirical antibiotic therapy groups and standard antibiotic therapy groups according to whether antibiotics were used before or after cultures yielded a result to explore the relationship between the timing of antibiotics and clinical efficacy. 152 out of 451 CRGNB-colonized patients in the prospective observational cohort developed CRGNB infection. The risk factors associated with CRGNB infection after colonization included CRKP (P < 0.001, OR = 3.27) and CRPA (P < 0.001, OR = 2.97) colonization, history of carbapenems use (P < 0.001, OR = 5.48), and immunocompromise (P < 0.001, OR = 7.07). There were 88 infected patients in the empirical antibiotic therapy groups and 64 in standard antibiotic therapy groups. The mortality was lower in empirical therapy groups than standard therapy groups (17.0% vs. 37.5%, P = 0.004, OR = 0.32). CRGNB colonized patients who are prone to infection have some high-risk factors included CRKP and CRPA colonization, immunocompromise, and prior carbapenems use. Once infection occurs in CRGNB-colonized patients, early use of effective antibiotics may be associated with reduced mortality, but more studies are needed.


Background
Carbapenem-resistant gram-negative bacteria (CRGNB) are one of the most serious multidrug-resistant bacteria (MDR) globally, accounting for a significant portion of Hospital Acquired Infections (HAI) and are associated with increased mortality and prolonged hospitalizations [1,2]. Active surveillance for early recognition of CRGNB carriers and other infection control measures, including contact isolation and decontamination strategies, are essential to reduce the spread of MDR [3,4]. We also found that active surveillance combined with early or preemptive isolation could reduce the spread of Page 2 of 9 Liang et al. Antimicrobial Resistance & Infection Control (2022) 11:89 carbapenem-resistant enterobacteria (CRE) in clinical practice [5]. A meta-analysis of 10 observational studies demonstrated CRE-colonized patients had a 16.5% risk of infection. Recent studies have demonstrated that the infection rates after CRGNB colonization ranged from 11 to 30% and were influenced by CRGNB endemicity and the implementation of IPC strategies [6]. Stringent measures have been designed to block the spread of CRE infection, encompassing active surveillance, strict contact isolation and hand hygiene, selective digestive decontamination (SDD), and restricted use of carbapenem antibiotics [4,7]. Antibiotics such as polymyxin B and ceftazidime-avibactam are acknowledged for their efficiency against CRGNB, while other measures have been developed, including the combination of antibiotics (double carbapenem combinations), prolonged infusion of carbapenem, and inhalation of polymyxin B and aminoglycosides to provide more options for treatment of CRGNB infections [8][9][10][11][12]. Nowadays, guidelines recommend early empirical antibiotic treatments for severe infections but tend to be cautious for the optimal antibiotic timing in CRGNB infection because of the paucity of effective antibiotics and potential inferior efficacy of the "old" antibiotics.
Antibiotics are usually indicated after CRGNB infection with documentation by culture yields, which may delay treatment for at least 2-3 days. Some studies have pointed out that earlier use of sensitive antibiotics may reduce the mortality of CRE bloodstream infection; however, there is a risk that indiscriminate use of antibiotics makes the pathogen more resistant, leading to antibiotic resistance [13]. Based on our experience, it seems more reasonable to use susceptibility-guided antibiotic treatment in CRGNB-colonized patients, but there is little empirical evidence. The purpose of this study was to explore the risk factors of CRGNB infection after colonization by active surveillance and to analyze the clinical effects of susceptibility-guided antibiotic treatment.

Study design
It was a single-center case control prospective observational study conducted at the General Intensive Care Unit (ICU) of the Second Affiliated Hospital of Zhejiang University, a 1200-bed tertiary academic care hospital with 40 intensive care beds in Hangzhou China. The study period spanned from 2017.01.01 to 2020.12.31. Since 2017.01, our ICU has implemented robust infection prevention and control (IPC) measures against CRE, emphasizing active surveillance combined with early or preemptive isolation. We have adopted active surveillance and individual isolation for patients with a high risk of MDR carriage, with two negative tests prerequisites for release from isolation [5]. In this study, CRGNB-colonized patients in active surveillance at admission were incorporated into the observation cohort. The purpose was to study the highrisk factors of infection after colonization of CRGNB in the first phase of the study. Therefore, the observation subjects were all patients with CRGNB colonization at active surveillance by collecting the risk factors of exposure, and the outcome was the occurrence of CRGNB infection. The purpose of the second phase of the study is to explore whether active surveillance of CRGNB has a positive impact on the use of antibiotics. The subjects were all patients with secondary CRGNB infection. The timing and clinical prognosis of antibiotic use were observed, and the outcome was prognosis such as transferred out of ICU or died. This study was conducted after approval from the ethics committee of the Second Affiliated Hospital of Zhejiang University school of medicine (Batch number IRB-2016-1511).

Data definition Inclusion criteria
Patients who underwent active surveillance on admission with any positive surveillance culture from the throat, rectal, or inguinal swabs, or other specimens sent by the clinicians. Inclusion criteria for active surveillance: 1. Patients transferred from other medical institutions, including (a) Patients hospitalized for more than seven days; (b) Patients hospitalized for less than seven days but who were in contact with suspected CRE patients or symptomatic; (c) Patients with prolonged hospitalization in private rehabilitation hospitals and nursing homes; (d) CRE-infected patients who were transferred from general wards but were hospitalized for less than 3 days. 2. Symptomatic patients transferred from high-risk general wards (a high-risk ward was defined as a ward with patients suffering from CRE infection within the preceding three months according to the hospital antimicrobial resistance monitoring report); 3. Patients who had been confirmed to be CRE carriers within the past year.

Exclusion criteria
Diagnosis of CRGNB infection on admission, admission less than 48 h, and age under 16. Complicated CRGNB mixed infection. If the secondary pathogenic bacteria and colonized bacteria are different types of CRGNB, we don't included in the analysis.  [16,17]. Gastrointestinal infection included intra-abdominal infection and infectious diarrhea. Other rare infections, including skin and soft tissue infection and intracranial infection, also have guidelines for reference [18,19].

Other important definitions
Medication history was limited to 3 months before CRGNB colonization, including cephalosporins, carbapenems, oral or intravenous glucocorticoids. Invasive procedures included catheterizations of deep veins and arteries for continuous renal replacement therapy, pulse indicator continuous cardiac output, and extracorporeal membrane oxygenation. Immunocompromised patients included patients with chemotherapy and long-term use of glucocorticoids or immunosuppressants. Other department history refers to patients who were admitted to other departments for more than 3 days. Other hospital history refers to the hospitalization in other hospitals or nursing institutions in the past six months. The definition of Acute kidney injury was based on the guidelines of The Kidney Disease: Improving Global Outcomes [20].

CRGNB strains including CRE, carbapenem-resistant
Klebsiella pneumoniae (CRKP), carbapenem-resistant Pseudomonas aeruginosa (CRPA), and carbapenem-resistant Acinetobacter baumannii (CRAB), were resistant to at least one of the carbapenems, including imipenem, meropenem, and ertapenem. The Antimicrobial susceptibility tests were performed with automated microbial identification and drug susceptibility systems (VITEK2 AST-GN16 France). Minimum inhibitory concentration (MIC) determination and interpretation complied with standards established by the Clinical and Laboratory Standards Institute (CLSI). Resistance to carbapenem was defined as a MIC ≥ 2 mg/L for imipenem or meropenem according to CLSI guidelines.

Statistical methods
All statistical analyses were performed in R and RStudio.
In univariate analysis, numerical variables were tested by independent sample T-test, dichotomous variables were compared by chi-square test, and a P value less than 0.05 was considered statistically significant. We include all the variables with P < 0.15 in univariate analysis into multivariate analysis in order to avoid omitting possible variables. Odds ratio (OR) and 95% confidence interval (CI) were calculated to evaluate the strength of association.  Table 2.

Results
Patients infected after colonization were divided into empirically sensitive antibiotic therapy groups (n = 88) and standard antibiotic therapy groups (n = 64), depending on the timing of antibiotic use. No statistical difference in basic characteristics was found between the two groups (Table3). The incidence of bloodstream infection was higher in empirically therapy groups than standard therapy groups; however, no statistical differences were found for the remaining part of the infection and the bacterial types. The most common antibiotic therapy used was combination therapy based on tigecycline and colistin, followed by ceftazidimeavibactam monotherapy. The proportion of combinations with polymyxin B was similar (15.9% vs. 15.6%, P = 0.96) and ceftazidime-avibactam monotherapy was more common in the empirically sensitive antibiotic therapy groups than in the standard antibiotic therapy groups (14.7% vs. 6.2%, P = 0.1). There was no statistical difference in the distribution of antibiotics regimens for CRGNB between the two groups which means that the difference in mortality did not cause by use of specific potent antibiotics. The average interval between colonization and infection was around 13 days, and ICU hospitalization was about 16 days after CRGNB infection. The mortality was lower in the empirical therapy groups than standard therapy groups (17.0% vs. 37.5%, P = 0.004, OR = 0.32, 95%CI 0.16-0.73), which was substantiated by survival curve analysis (P = 0.002) (Fig. 2).

Discussion
Nearly one third of CRGNB colonized patients develop secondary infections who underwent active surveillance on ICU admission, which is a very high proportion. These patients may benefit from early use of empirically sensitive antibiotic therapy. Contact isolation and hand hygiene have been widely emphasized, while much controversy surrounds the implementation of active surveillance for its questionable cost-effectiveness [21]. CRGNB colonization has increased with the widespread prevalence of CRGNB in Asia. Indeed, at clinical level, the medical personnel are often overwhelmed by the concomitant increase in workload following the implementation of basic IPC measures. In some areas with limited resources, active surveillance is not even recommended for asymptomatic patients [21]. Clinical studies have consistently shown that CRGNB colonization is a high-risk factor for infection [22,23]. However, the risk factors of CRGNB infection after colonization are not necessarily the same for CRGNB colonization which may also depend on the patient's immune status. It is widely acknowledged that advanced age, serious disease status, including higher sequential organ failure assessment (SOFA), Pitt score, Charlson score, carbapenem drug exposure, and invasive catheterization are high-risk factors for MDR infection after colonization [23,24].
For CRGNB-colonized patients, a series of stringent measures were adopted, including contact isolation and strict hand hygiene to avoid cross-infection; nonetheless, CRGNB infection after colonization was unavoidable, leading to a poor patient prognosis. Addressing risk factors of infection after colonization earlier may potentially be the solution to this conundrum. Herein, CRKP and CRPA colonization were associated with greater CRGNB infection rates than CRAB and other CRE. Other identified risk factors included invasive catheterization, history of carbapenem use, and immunocompromise. It is important to note that the highrisk factors analysis does not accurately predict which patients will develop CRGNB infection. In this regard, many scoring systems based on high-risk factors have been designed to solve this problem. However, no consensus has been reached on the accuracy of these scoring systems due to the inherent differences in regional characteristics, medical habits, and types of diseases [23][24][25]. Building a more efficient and robust prediction model by artificial intelligence may potentially be the solution. SDD programs have been implemented in some European countries to curb CRGNB infection rates after CRE colonization; however, the efficiency is uncertain due to their implementation in low CRE epidemic areas [21]. Given the uncertainty of the effects of SDD, our center did not adopt any decontamination strategy.
An increasing body of evidence suggests that early empirical antibiotic use can reduce mortality of severe sepsis or severe pneumonia [26,27]. The Infectious Diseases Society of America (IDSA) recommends that empirical antibiotics use should take it into account of antibiotic sensitivity data in the past six months and antibiotic exposure in the past 30 days [2]. The European consensus recommends that the empirical treatment of CRGNB should be limited to critically ill patients, and it is necessary to know the drug sensitivity of CRGNB to carry out empirical treatment [13]. In addition, misuse of antibiotics can also be dangerous. Elena Carrara et al. pointed out that inappropriate empirical antibiotic use was closely related to mortality MDR strain infections [28].  Interestingly, in a randomized controlled study by Yael Zak-Doron et al. the mortality rate was not reduced in patients with severe CRGNB infection that received early (during the first 48 h after culture taking) empirical antibiotic treatments with mainly colistin (96%) [29]. This conclusion is surprising since 77% of CRGNB infections were related to CRAB, and all patients received empirical antibiotic treatment. If empirical treatment can prevent early death, this exclusion may bias the results to be ineffective, but unfortunately, the study does not have complete data on early death and its empirical treatment [29]. In addition, the study found that documented colonization (P = 0.044, OR = 0.76) may be a potential protective factor, which may be related to empirical antibiotic therapy [29]. In appropriate empirical treatment is dangerous and emphasizes the importance of active surveillance. Active surveillance provides more background information on bacterial colonization, although subsequent infections in CRGNB colonized patients are possibly mixed infections or another pathogenic bacterium. However, during severe infection or even septic shock management, microbiological data from CRGNB-colonized patients provides evidence of the potential etiology with drug sensitivity results to help clinicians in the decisionmaking process [13]. Herein, the empirical antibiotic treatments of tigecycline, colistin, ceftazidime-avibactam achieved good results. We recommend that secondary infected patients with CRGNB colonization use of directed empirical antibiotics when new infections are uncontrollable, but we need to avoided abuse according to the drug resistance, virulence, and invasive site of local prevalence of drug-resistant bacteria.
There were some limitations in our research. This study was a single-center prospective observational study, and the level of evidence was low. To substantiate our findings, more studies should be conducted involving multiple centers and clinical trials. Moreover, potential confounding factors such as changes in recommendations for CRGNB treatment and clinical experience of physicians were not considered.

Conclusions
CRGNB colonized patients who are prone to infection have some high-risk factors included CRKP and CRPA colonization, immunocompromise, and prior carbapenems use. For patients with new infections after CRGNB colonization, early use of effective antibiotics may be associated with reduced mortality, but more studies are needed.