Clinical prediction models for ESBL-Enterobacteriaceae colonization or infection: a systematic review

doi:10.1016/j.jhin.2019.01.012

Journal of Hospital Infection

Volume 102, Issue 1, May 2019, Pages 8-16

https://doi.org/10.1016/j.jhin.2019.01.012 Get rights and content

Summary

Background

β-Lactamase resistance among certain Gram-negative bacteria has been associated with increased mortality, length of hospitalization, and hospital costs.

Aim

To identify and critically appraise existing clinical prediction models of extended-spectrum β-lactamase-producing Enterobacteriaceae (ESBL-EKP) infection or colonization.

Methods

Electronic databases, reference lists, and citations were searched from inception to April 2018. Papers were included in any language describing the development or validation, or both, of models and scores to predict the risk of ESBL-EKP infection or colonization.

Findings

In all, 1795 references were screened, of which four articles were included in the review. The included studies were carried out in different geographical locations with differing study designs, and inclusion and exclusion criteria. Most if not all studies lacked external validation and blinding of reviewers during the evaluation of the predictor variables and outcome. All studies excluded missing data and most studies did not report the number of patients excluded due to missing data. Fifteen predictors of infection or colonization with ESBL-EKP were identified. Commonly included predictors were previous antibiotic use, previous hospitalization, transfer from another healthcare facility, and previous procedures (urinary catheterization and invasive procedures).

Conclusion

Due to limitations and variations in the study design, clinicians would have to take these differences into consideration when deciding on how to use these models in clinical practice. Due to lack of external validation, the generalizability of these models remains a question. Therefore, further external validation in local settings is needed to confirm the usefulness of these models in supporting decision-making.

Introduction

Several studies have suggested that infections caused by extended-spectrum β-lactamase-producing Enterobacteriaceae (i.e. Escherichia coli, Klebsiella spp., and Proteus mirabilis) (ESBL-EKP) have an important clinical impact, and the increasing prevalence of these organisms in hospitals has been well documented [1], [2], [3]. β-Lactamase resistance among certain Gram-negative bacteria has been associated with increased mortality, length of hospitalization, and hospital costs. A study by Schwaber et al. concluded that infection with an ESBL-producing organism was associated with an adjusted 3.6-fold increased risk of in-hospital mortality, an unadjusted 2.3-fold increased risk of infection-related mortality, an adjusted 1.6-fold increase in length of stay, an adjusted 25-fold increased risk of delay in appropriate therapy, and an unadjusted 4-fold-increased likelihood of discharge to a long-term care facility for those who survived [4]. In view of these potential complications, it is important to implement measures to combat resistance, develop treatment strategies to overcome the adverse consequences of resistance, and to identify patients at risk of resistance early and accurately, so that effective antibiotic therapy can be given.

Precise determination of risk factors for ESBL-EKP infection may assist in accurate targeting of empirical carbapenem therapy; with appropriate therapy, there should be a decline in treatment failure, infectious complications, antibiotic costs, and the risk of selecting carbapenem resistance [5], [6]. A risk discrimination tool that can identify patients likely to harbour ESBL-EKP might therefore assist with rational antibiotic prescribing and the early implementation of infection control precautions [7].

Several clinical scoring tools have been published since 2011 [7], [8], [9], [10]. Attempts to develop or validate these scoring tools have been made by different study groups in different populations and their findings have been heterogeneous. To our knowledge, no previous systematic review of clinical risk scoring systems for predicting ESBL-EKP colonization or infection in hospitalized patients has been published. Therefore, we aimed to carry out a systematic review, to describe and give an overview of these scoring systems, and to compare and contrast the scores to assist clinicians in their use in clinical practice.

Section snippets

Methods

The Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modelling Studies (CHARMS) Checklist was followed to guide the framing of the review aim, search strategy, and study inclusion and exclusion criteria [11]. The review was prospectively registered in PROSPERO [12]. Reporting of the review was consistent with PRISMA guidelines [13].

Study selection

Our search identified 1795 articles after removal of duplicates. Only four articles meeting the inclusion criteria were included in the systematic review (Figure 1) [7], [8], [9], [10]. The earliest study was published in 2011 and the most recent was published in 2017.

Study design and participants

Two types of study design were used: retrospective matched case–control studies [8], [9] and cohort studies [7], [10]. Only one study performed an external prospective validation of the prediction model derived from their

Discussion

In the last two decades, extensive use of broad-spectrum β-lactams has led to the emergence of antibiotic-resistant strains of Enterobacteriaceae, including ESBL-EKP [1], [15], [16]. β-Lactams are most commonly used for the treatment of bacterial infections. The persistent exposure of bacterial strains to a multitude of β-lactams has induced dynamic and continuous production, and mutation of the β-lactamase enzymes in these bacteria, expanding their activity to newly developed β-lactam

Conflict of interest statement

None declared.

Funding sources

None.

References (25)

M. Falagas et al.
Extended-spectrum β-lactamase-producing organisms
J Hosp Infect
(2009)
Y.-S. Yang et al.
Impact of extended-spectrum β-lactamase-producing Escherichia coli and Klebsiella pneumoniae on the outcome of community-onset bacteremic urinary tract infections
J Microbiol Immunol Infect
(2010)
T. Goulenok et al.
Risk factors for developing ESBL E. coli: can clinicians predict infection in patients with prior colonization?
J Hosp Infect
(2013)
K. Kengkla et al.
Clinical risk scoring system for predicting extended-spectrum β-lactamase-producing Escherichia coli infection in hospitalized patients
J Hosp Infect
(2016)
J.D. Pitout et al.
Extended-spectrum β-lactamase-producing Enterobacteriaceae: an emerging public-health concern
Lancet Infect Dis
(2008)
A. Daroukh et al.
Characteristics and costs of carbapenemase-producing enterobacteria carriers (2012/2013)
Medecine et maladies infectieuses
(2014)
J. Maslikowska et al.
Impact of infection with extended-spectrum β-lactamase-producing Escherichia coli or Klebsiella species on outcome and hospitalization costs
J Hosp Infect
(2016)
C.G. Giske et al.
Clinical and economic impact of common multidrug-resistant Gram-negative bacilli
Antimicrob Agents Chemother
(2008)
M. Tumbarello et al.
Costs of bloodstream infections caused by Escherichia coli and influence of extended-spectrum-β-lactamase production and inadequate initial antibiotic therapy
Antimicrob Agents Chemother
(2010)
M.J. Schwaber et al.
Clinical and economic impact of bacteremia with extended-spectrum-β-lactamase-producing Enterobacteriaceae
Antimicrob Agents Chemother
(2006)

S.W. Johnson et al.

Utility of a clinical risk factor scoring model in predicting infection with extended-spectrum β-lactamase-producing Enterobacteriaceae on hospital admission

Infect Control Hosp Epidemiol

(2013)

M. Tumbarello et al.

Identifying patients harboring extended-spectrum-β-lactamase-producing Enterobacteriaceae on hospital admission: derivation and validation of a scoring system

Antimicrob Agents Chemother

(2011)

Cited by (17)

Extended spectrum cephalosporin resistant enterobacteriaceae carriage and infection in patients admitted with newly-diagnosed acute leukemia
2023, American Journal of Infection Control
Citation Excerpt :
Risk factors for acquired carriage in our patient population have not been adequately determined. Prior studies of other populations have identified a number of predictors including older age, male sex, and previous recent contact with the health care system, mainly hospitalization and antibiotic exposure.8,9 Arnan et al.10 studied a diverse population of patients with hematologic malignancies and HSCT recipients; combining both those with carriage on admission and acquired colonization, they identified only recent prior antibiotic exposure as a risk factor for ESCRE E. coli carriage.
Current information is limited on the incidence, risk factors, and consequences of extended-spectrum cephalosporin resistant Enterobacteriaceae (ESCRE) carriage in patients undergoing therapy for newly-diagnosed acute leukemia.
We monitored 300 consecutive patients who submitted a first stool within the first week of initial hospitalization for initial and hospital acquired ESCRE carriage. Selected available isolates underwent DNA sequencing for determination of strain typing and resistance genes.
19 (6%) patients had ESCRE in their initial stool, and there was continued risk for new acquisition throughout their multiple hospitalizations. Patients with AML had more acquired carriage during their initial hospitalization. Increased hospitalizations and male sex were risk factors for detected acquired ESCRE carriage. ESCRE stool carriage was predictive for ESCRE BSI but not for overall survival. Sequencing revealed that E. coli ESCRE isolates contained primarily ESBL, while Enterobacter spp. and Citrobacter spp. showed primarily AmpC genes. The antibiotic sensitivity patterns for ESCRE BSI isolates reflected these genome findings.
ESCRE carriage is common in patients with acute leukemia undergoing repeated hospitalizations and increases the risk for ESCRE BSI. ESCRE genera express differing resistance genes which may be predictive for empiric antibiotic efficacy.
Machine learning in predicting antimicrobial resistance: a systematic review and meta-analysis
2022, International Journal of Antimicrobial Agents
Citation Excerpt :
Other performance metrics (F1-score, PPV, NPV, Precision, Recall) for ML prediction are also provided in Figure 4. The contingency table was extracted from 11 ML prediction studies [19,23,27,28,33-35,38,40,41,43] and 30 scoring system studies (supplementary material references [1-5,7,8,11-13,15,16,18,19,21,22,25-29,31-39]). The pooled estimated AUC was 0.82 (0.78–0.85) for ML prediction and 0.65 (0.23–0.92) for the scoring system.
Antimicrobial resistance (AMR) is a global health threat; rapid and timely identification of AMR improves patient prognosis and reduces inappropriate antibiotic use.
Relevant literature in PubMed, Web of Science, Embase and Institute of Electrical and Electronics Engineers prior to 28 September 2021 was searched. Any study that deployed machine learning (ML) or a risk score as a tool to predict AMR was included in the final review; there were 25 studies that employed the ML algorithm to predict AMR.
Extended spectrum β-lactamases, methicillin-resistant Staphylococcus aureus (MRSA) and carbapenem resistance were the most common outcomes in studies with a specific resistance pattern. The most common algorithms in ML prediction were logistic regression (n = 14 studies), decision tree (n = 14) and random forest (n = 7). The area under the curve (AUC) range for ML prediction was 0.48–0.93. The pooled AUC for ML prediction was 0.82 (0.78–0.85). Compared with risk score, higher specificity [87% (82–91) vs. 37% (25–51)] was indicated for ML prediction, but not sensitivity [67% (62–72) vs. 73% (67–79)].
Machine learning might be a potential technology for AMR prediction; however, retrospective methodology for model development, nonstandard data processing and scarcity of validation in a randomised controlled trial or real-world study limit the application of these models in clinical practice.
Restricted antimicrobial prescribing in an area of highly prevalent antimicrobial resistance
2021, Infectious Diseases Now
Citation Excerpt :
One ASP intervention would be to defer initiation of rABX in non-responding but stable patients. This could be implemented in association with modalities such as decision support with predictive models for MDRO infections and rapid diagnostic tools [22–25]. Recent hospitalization was the most frequent stated reason for rABX prescription (72.7%).
To assess the predictive value for infection with multidrug-resistant organisms (MDROs) of reasons for empirical prescription of restricted antibiotics (rABX), in a setting with high resistance rates.
We prospectively studied all rABX prescriptions in a 550-bed tertiary teaching hospital from April 15 to June 14, 2018 and from September 1 to October 30, 2018. Prescribing physicians had to justify their decision by choosing one or more prespecified reasons.
We reviewed 172 empirical prescriptions of rABX, which accounted for 67.2% of all rABX prescriptions. Stated reasons for empirical prescription of rABX were recent hospitalization (72.7%), escalation due to non-response to previous antimicrobials (47.7%), treatment for severe sepsis/septic shock (45.9%), escalation due to recurrence or deterioration (22.1%), prior MDRO infection (12.8%), and prior MDRO colonization (7.6%). Empirical treatment for septic shock or severe sepsis was the only significant predictor of MDRO isolation (OR = 5.26, 95% CI: 1.5–18.4, P = 0.009), while recent hospitalization had a high negative predictive value for MDRO (97.4%). Fourteen per cent of microbiologically documented infections were associated with MDROs resistant to the prescribed rABX.
Empirical treatment for severe sepsis or septic shock was the only independent predictor of MDRO isolation. Recent hospitalization had a high negative predictive value for MDRO infection. The isolation of pathogens resistant to the prescribed rABX suggests that in a setting with widespread antimicrobial resistance, it could be difficult to reduce the empirical use of rABX without risking inadequate treatment.
Third-Generation Cephalosporin Resistance and Associated Discordant Antibiotic Treatment in Emergency Department Febrile Urinary Tract Infections
2021, Annals of Emergency Medicine
Citation Excerpt :
Yet, unfortunately, these risk factors were also present in 86% of control patients. This is consistent with published clinical prediction models for ESBL infection or colonization, most of which achieve sensitivities of 80% to 94%, but much lower specificities of 54% to 74%.19,27 It seems likely that any attempt to cover most ESBL-EKP UTIs with appropriate empiric therapy, such as a carbapenem, would require a tremendous amount of “overshooting,” exposing many patients unnecessarily to very broad-spectrum empiric therapy.
Third-generation cephalosporin-resistant (3GCR) Escherichia coli, Klebsiella pneumoniae, and Proteus mirabilis (EKP) are an increasingly common cause of community-onset urinary tract infections (UTIs) in the United States. The 3GCR antimicrobial resistance pattern in these Enterobacterales species is most commonly due to production of extended-spectrum β-lactamases. We sought to provide contemporary, emergency department (ED)-focused data on 3GCR-EKP UTI regional prevalence, presentation, antibiotic susceptibility, and empiric treatment patterns, and outcomes.
We performed a retrospective cohort study of all adults admitted with a febrile UTI at 21 Kaiser Permanente Northern California EDs between January 2017 and June 2019. Inclusion criteria included fever; admitting diagnosis of UTI, pyelonephritis, or sepsis; and ED urine culture with greater than 100,000 colony-forming units/mL of an EKP species. 3GCR was defined as in vitro resistance to ceftriaxone, ceftazidime, or both. 3GCR-EKP cases were compared with non-3GCR-EKP controls for the following: demographics, comorbidities, presenting clinical features, urinary isolate antimicrobial susceptibility, treatment, and clinical outcomes. The primary outcome measure was the rate of discordant initial empiric antibiotic treatment (administered within 6 hours of ED arrival) when compared with antimicrobial susceptibility testing. Secondary outcomes included hospital length of stay and 90-day mortality, adjusted for comorbidities and severity of illness.
There were 4,107 patients (median age 73 years and 35% men) who met study inclusion criteria. Of these patients, 530 (12.9%) had a 3GCR-EKP urinary tract infection. The proportion of subjects possessing risk factors for a health care-associated or extended-spectrum β-lactamase infection was 92.8% of case patients and 86.1% of controls. When comparing 3GCR-EKP case and non-3GCR-EKP control isolates, ciprofloxacin susceptibility rates were 21% versus 88%, and piperacillin/tazobactam susceptibility rates were 89% versus 97%, respectively. Initial empiric antibiotic therapy was discordant with antimicrobial susceptibility testing results in 63% of case patients versus 7% of controls (odds ratio 21.0; 95% confidence interval 16.9 to 26.0). The hospital length of stay was longer for 3GCR-EKP case patients, with an adjusted mean difference of 29.7 hours (95% CI 19.0 to 40.4). Ninety-day mortality was 12% in case patients versus 8% in controls (adjusted odds ratio 1.56; 95% confidence interval 1.07 to 2.28).
In this large, 2017 to 2019 Northern California ED study, nearly 13% of febrile EKP UTIs requiring hospitalization were caused by 3GCR-EKP, and in these cases, initial empiric therapy was often discordant with antimicrobial susceptibility testing. 3GCR-EKP infections were associated with a longer hospital length of stay and higher 90-day mortality. Similar data from other regions and for outpatient UTIs are needed.
Relative faecal abundance of extended-spectrum β-lactamase-producing Enterobacterales and its impact on infections among intensive care unit patients: a pilot study
2021, Journal of Hospital Infection
Citation Excerpt :
This could incorporate information known to be associated with the risk of ESBL-PE infection in carriers such as the source of infection (urinary, intra-abdominal), relative faecal abundance of ESBL-PE, and severity of the infection. Recent studies have attempted to build a predictive score for the risk of ESBL-PE infection in carrier patients [9,10]. These scores most often include age, prior antibiotic therapy, and source of infection (urinary and intra-abdominal).
This pilot prospective study assessed the association between the faecal relative abundance of extended-spectrum β-lactamase-producing Enterobacterales (ESBL-PE) and the occurrence of ESBL-PE related infections. Twenty-four patients were included. The median ESBL relative abundance was 32.4%. The mean ESBL-PE relative abundance (ESBL-PE-RA) was more than five-fold higher in patients exposed during the last three months to antibiotics (P = 0.002). Furthermore, the mean ESBL relative abundance was more than two-fold higher in patients colonized with non-E. coli strains (P = 0.044). The mean ESBL-PE-RA was more than 10-fold higher for the concordant patients than for the discordant patients (59.1% vs 4.9%; P < 0.001).
Five-Hour Detection of Intestinal Colonization with Extended-Spectrum-b-Lactamase-Producing Enterobacteriaceae Using the b-Lacta Phenotypic Test: the BLESSED Study
2023, Microbiology Spectrum

View all citing articles on Scopus

View full text

ReviewClinical prediction models for ESBL-Enterobacteriaceae colonization or infection: a systematic review

Summary

Background

Aim

Methods

Findings

Conclusion

Introduction

Section snippets

Methods

Study selection

Study design and participants

Discussion

Conflict of interest statement

Funding sources

J Hosp Infect

J Microbiol Immunol Infect

J Hosp Infect

J Hosp Infect

Lancet Infect Dis

Medecine et maladies infectieuses

J Hosp Infect

Clinical and economic impact of common multidrug-resistant Gram-negative bacilli

Antimicrob Agents Chemother

Costs of bloodstream infections caused by Escherichia coli and influence of extended-spectrum-β-lactamase production and inadequate initial antibiotic therapy

Antimicrob Agents Chemother

Clinical and economic impact of bacteremia with extended-spectrum-β-lactamase-producing Enterobacteriaceae

Antimicrob Agents Chemother

Utility of a clinical risk factor scoring model in predicting infection with extended-spectrum β-lactamase-producing Enterobacteriaceae on hospital admission

Infect Control Hosp Epidemiol

Identifying patients harboring extended-spectrum-β-lactamase-producing Enterobacteriaceae on hospital admission: derivation and validation of a scoring system

Antimicrob Agents Chemother

Review
Clinical prediction models for ESBL-Enterobacteriaceae colonization or infection: a systematic review