Prediction of Fluoroquinolone Resistance in Gram-Negative Bacteria Causing Bloodstream Infections

Survey of Bacterial Isolates and Their Antimicrobial Susceptibility Patterns from Dogs with Infective Endocarditis

Infective endocarditis (IE) is a potentially fatal disease in dogs. Limited information exists regarding the characterization of bacterial isolates from dogs with IE. The objective of this study was to describe bacterial isolates associated with IE and their antimicrobial susceptibility patterns. A retrospective analysis of dogs with IE and bacterial isolates was performed, and antimicrobial susceptibility was interpreted using current veterinary cut points where available. The susceptibility rate was assessed for association with survival and previous antimicrobial administration. Fifty-one bacterial isolates were identified from 45 dogs, and 33 had antimicrobial susceptibility performed. Staphylococcus spp. (14/51; 27.5%) was the most common organism. Antimicrobials with the lowest susceptibility rate were ampicillin (19/26; 73%), doxycycline (16/22; 73%), and enrofloxacin (22/29; 76%) with 12/33 (36%) of isolates exhibiting multidrug resistance (MDR). Individual antimicrobial resistances and the MDR rate were not associated with a difference in survival rate. Bacterial isolates from dogs that had received fluoroquinolone antimicrobials in the month before diagnosis had a higher rate of non-intrinsic fluoroquinolones resistance (5/8;62.5%) compared to those that did not receive fluoroquinolones (2/21; 9.5%) (p = 0.03). Antimicrobial resistance and MDR phenotype were common in this study. Culture and antimicrobial susceptibility testing should be pursued in dogs with IE to help guide antimicrobial therapy.

Risk for primary cephalosporin resistance in Gram-negative bacteremia

Objective

This study aimed to examine the clinical risk factors for cephalosporin resistance in patients with Gram-negative bacteremia caused by Escherichia coli (EC), Klebsiella pneumoniae (KP), Enterobacter cloacae (ENC), and Pseudomonas aeruginosa (PS).

Methods

This retrospective cohort study included 400 adults with Gram-negative bacteremia. The goal was to review 100 cases involving each species and approximately half resistant and half susceptible to first-line cephalosporins, ceftriaxone (EC or KP), or cefepime (ENC or PS). Logistic regression was used to identify factors predictive of resistance.

Results

A total of 378 cases of Gram-negative bacteremia were included in the analysis. Multivariate analysis identified significant risk factors for resistance, including admission from a chronic care hospital, skilled nursing facility, or having a history of infection within the prior 6 months (OR 3.00, P < .0001), requirement for mechanical ventilation (OR 3.76, P < .0001), presence of hemiplegia (OR 3.54, P = .0304), and presence of a connective tissue disease (OR 3.77, P = .0291).

Conclusions

Patients without the identified risk factors should be strongly considered for receiving ceftriaxone or cefepime rather than carbapenems and newer broad-spectrum agents.

Prediction of ciprofloxacin resistance in hospitalized patients using machine learning

BACKGROUND

Ciprofloxacin is a widely used antibiotic that has lost efficiency due to extensive resistance. We developed machine learning (ML) models that predict the probability of ciprofloxacin resistance in hospitalized patients.

METHODS

Data were collected from electronic records of hospitalized patients with positive bacterial cultures, during 2016-2019. Susceptibility results to ciprofloxacin (n = 10,053 cultures) were obtained for Escherichia coli, Klebsiella pneumoniae, Morganella morganii, Pseudomonas aeruginosa, Proteus mirabilis and Staphylococcus aureus. An ensemble model, combining several base models, was developed to predict ciprofloxacin resistant cultures, either with (gnostic) or without (agnostic) information on the infecting bacterial species.

RESULTS

The ensemble models' predictions are well-calibrated, and yield ROC-AUCs (area under the receiver operating characteristic curve) of 0.737 (95%CI 0.715-0.758) and 0.837 (95%CI 0.821-0.854) on independent test-sets for the agnostic and gnostic datasets, respectively. Shapley additive explanations analysis identifies that influential variables are related to resistance of previous infections, where patients arrived from (hospital, nursing home, etc.), and recent resistance frequencies in the hospital. A decision curve analysis reveals that implementing our models can be beneficial in a wide range of cost-benefits considerations of ciprofloxacin administration.

CONCLUSIONS

This study develops ML models to predict ciprofloxacin resistance in hospitalized patients. The models achieve high predictive ability, are well calibrated, have substantial net-benefit across a wide range of conditions, and rely on predictors consistent with the literature. This is a further step on the way to inclusion of ML decision support systems into clinical practice.

The time-varying association between previous antibiotic use and antibiotic resistance

OBJECTIVES

The objective of the study was to estimate how the time elapsed from previous antibiotic use is associated with antibiotic resistance.

METHODS

Data comprised electronic medical records of all patients in an Israeli hospital who had a positive bacterial culture from 2016 to 2019. These included susceptibility testing results and clinical and demographic data. Mixed-effects time-varying logistic models were fitted to estimate the association between the time elapsed since the last use of aminoglycosides and gentamicin resistance (n = 13 095), cephalosporins and ceftazidime resistance (n = 13 051), and fluoroquinolones and ciprofloxacin resistance (n = 15 364) while adjusting for multiple covariates.

RESULTS

For all examined antibiotics, previous antibiotic use had a statistically significant association with resistance (p < 0.001). These associations exhibited a clear decreasing pattern over time, which we present as a flexible function of time. Nonetheless, previous antibiotic use remained a significant risk factor for resistance for at least 180 days when the adjusted ORs were 1.94 (95% CI, 1.40-2.69), 1.33 (95% CI, 1.10-1.61), and 2.25 (95% CI, 1.49-3.41) for gentamicin, ceftazidime, and ciprofloxacin, respectively.

DISCUSSION

The association between prior antibiotic use and resistance decreases over time. Commonly used cut-offs for prior antibiotic use can either misclassify patients still at higher risk when too recent or provide a diluted estimate of the effects of antibiotic use on future resistance when too distant. Hence, prior antibiotic use should be considered a time-dependent risk factor for resistance in both epidemiological research and clinical practice.

Machine learning model for predicting ciprofloxacin resistance and presence of ESBL in patients with UTI in the ED

Increasing antimicrobial resistance in uropathogens is a clinical challenge to emergency physicians as antibiotics should be selected before an infecting pathogen or its antibiotic resistance profile is confirmed. We created a predictive model for antibiotic resistance of uropathogens, using machine learning (ML) algorithms. This single-center retrospective study evaluated patients diagnosed with urinary tract infection (UTI) in the emergency department (ED) between January 2020 and June 2021. Thirty-nine variables were used to train the model to predict resistance to ciprofloxacin and the presence of urinary pathogens' extended-spectrum beta-lactamases. The model was built with Gradient-Boosted Decision Tree (GBDT) with performance evaluation. Also, we visualized feature importance using SHapely Additive exPlanations. After two-step customization of threshold adjustment and feature selection, the final model was compared with that of the original prescribers in the emergency department (ED) according to the ineffectiveness of the antibiotic selected. The probability of using ineffective antibiotics in the ED was significantly lowered by 20% in our GBDT model through customization of the decision threshold. Moreover, we could narrow the number of predictors down to twenty and five variables with high importance while maintaining similar model performance. An ML model is potentially useful for predicting antibiotic resistance improving the effectiveness of empirical antimicrobial treatment in patients with UTI in the ED. The model could be a point-of-care decision support tool to guide clinicians toward individualized antibiotic prescriptions.

Advances in novel antibiotics to treat multidrug-resistant gram-negative bacterial infections

Antimicrobial resistance is a growing threat to public health and an increasingly common problem for acute care physicians to confront. Several novel antibiotics have been approved in the past decade to combat these infections; however, physicians may be unfamiliar with how to appropriately utilize them. The purpose of this review is to evaluate novel antibiotics active against resistant gram-negative bacteria and highlight clinical information regarding their use in the acute care setting. This review focuses on novel antibiotics useful in the treatment of infections caused by resistant gram-negative organisms that may be seen in the acute care setting. These novel antibiotics include ceftolozane/tazobactam, ceftazidime/avibactam, meropenem/vaborbactam, imipenem/cilistatin/relebactam, cefiderocol, plazomicin, eravacycline, and omadacycline. Acute care physicians should be familiar with these novel antibiotics so they can utilize them appropriately.

Type 2 diabetes mellitus and antibiotic-resistant infections: a systematic review and meta-analysis

BACKGROUND

Type 2 diabetes mellitus (T2DM) has been associated with infectious diseases; however, whether T2DM is associated with bacterial-resistant infections has not been thoroughly studied. We ascertained whether people with T2DM were more likely to experience resistant infections in comparison to T2DM-free individuals.

METHODS

Systematic review and random-effects meta-analysis. The search was conducted in Medline, Embase and Global Health. We selected observational studies in which the outcome was resistant infections (any site), and the exposure was T2DM. We studied adult subjects who could have been selected from population-based or hospital-based studies. I² was the metric of heterogeneity. We used the Newcastle-Ottawa risk of bias scale.

RESULTS

The search retrieved 3370 reports, 97 were studied in detail and 61 (449 247 subjects) were selected. Studies were mostly cross-sectional or case-control; several infection sites were studied, but mostly urinary tract and respiratory infections. The random-effects meta-analysis revealed that people with T2DM were twofold more likely to have urinary tract (OR=2.42; 95% CI 1.83 to 3.20; I² 19.1%) or respiratory (OR=2.35; 95% CI 1.49 to 3.69; I² 58.1%) resistant infections. Although evidence for other infection sites was heterogeneous, they consistently suggested that T2DM was associated with resistant infections.

CONCLUSIONS

Compelling evidence suggests that people with T2DM are more likely to experience antibiotic-resistant urinary tract and respiratory infections. The evidence for other infection sites was less conclusive but pointed to the same overall conclusion. These results could guide empirical treatment for patients with T2DM and infections.

Investigation of Antimicrobial Susceptibilities Among Bacteria Isolated from Blood Cultures in Hospitalized Patients, Tehran, Iran

Background: Bacteremia is the status, which is detected via a positive blood culture test with no contamination. Centers for Disease Control and Prevention (CDC) indicates that direct medical procedures and total costs are significantly high. Antibiotic resistance can play a major role in the costs, which are related to the long duration of treatment. Objectives: The aim of this study was to investigate the rate and profiles of antimicrobial susceptibility of blood culture isolates from Tehran, Iran. Methods: In the current cross-sectional study, a total of 5,000 blood culture samples were collected from patients hospitalized in the Loghman General Hospital, Tehran, Iran, with positive blood culture results from 2012 to 2013. Susceptibility to antimicrobial agents was analyzed using National Committee for Clinical Laboratory Standards guidelines. Results: Coagulase-negative staphylococci (38.8%), Staphylococcus aureus (20.5%), Acinetobacter (11.9%), and Escherichia coli (11.7%) were the most frequent bacteria isolated from the blood cultures, collectively accounting for > 80% of the isolates. Of isolated microorganisms, 63.75% and 36.24% belonged to Gram-positive and Gram-negative bacteria, respectively. Moreover, 88% of the isolates were MRSA (oxacillin-/methicillin-resistant), and 7% were VRE (vancomycin-resistant). Conclusions: The most frequent isolated organisms were Gram-positive bacteria, and the rate of MDR (multi-drug resistance) was high. The results of the current study obviously indicate the misuse of antibiotic in society. National surveillance studies in Iran will be useful for clinicians to choose the right empirical treatment and will help control and prevent infections caused by resistant organisms.

Predicting antibiotic resistance in hospitalized patients by applying machine learning to electronic medical records

BACKGROUND

Computerized decision support systems are becoming increasingly prevalent with advances in data collection and machine learning algorithms. However, they are scarcely used for empiric antibiotic therapy. Here we accurately predict the antibiotic resistance profiles of bacterial infections of hospitalized patients using machine learning algorithms applied to patients' electronic medical records (EMR).

METHODS

The data included antibiotic resistance results of bacterial cultures from hospitalized patients, alongside their electronic medical records. Five antibiotics were examined: Ceftazidime (n=2942), Gentamicin (n=4360), Imipenem (n=2235), Ofloxacin (n=3117) and Sulfamethoxazole-Trimethoprim (n=3544). We applied lasso logistic regression, neural networks, gradient boosted trees, and an ensemble combining all three algorithms, to predict antibiotic resistance. Variable influence was gauged by permutation tests and Shapely Additive Explanations analysis.

RESULTS

The ensemble model outperformed the separate models and produced accurate predictions on a test set data. When no knowledge regarding the infecting bacterial species was assumed, the ensemble model yielded area under the receiver-operating-characteristic (auROC) scores of 0.73-0.79, for different antibiotics. Including information regarding the bacterial species improved the auROCs to 0.8-0.88. The effects of different variables on the predictions were assessed and found consistent with previously identified risk factors for antibiotic resistance.

CONCLUSIONS

Our study demonstrates the potential of machine learning models to accurately predict antibiotic resistance of bacterial infections of hospitalized patients. Moreover, we show that rapid information regarding the infecting bacterial species can improve predictions substantially. The implementation of such systems should be seriously considered by clinicians to aid correct empiric therapy and to potentially reduce antibiotic misuse.

Pharmacist-Driven Culture and Sexually Transmitted Infection Testing Follow-Up Program in the Emergency Department

Expanding pharmacist-driven antimicrobial stewardship efforts in the emergency department (ED) can improve antibiotic management for both admitted and discharged patients. We piloted a pharmacist-driven culture and rapid diagnostic technology (RDT) follow-up program in patients discharged from the ED. This was a single-center, pre- and post-implementation, cohort study examining the impact of a pharmacist-driven culture/RDT follow-up program in the ED. Adult patients discharged from the ED with subsequent positive cultures and/or RDT during the pre- (21 August 2018–18 November 2018) and post-implementation (19 November 2018–15 February 2019) periods were screened for inclusion. The primary endpoints were time from ED discharge to culture/RDT review and completion of follow-up. Secondary endpoints included antimicrobial agent prescribed during outpatient follow-up, repeat ED encounters within 30 days, and hospital admissions within 30 days. Baseline characteristics were analyzed using descriptive statistics. Time-to-event data were analyzed using the Wilcoxon signed-rank test. One-hundred-and-twenty-seven patients were included, 64 in the pre-implementation group and 63 in the post-implementation group. There was a 36.3% reduction in the meantime to culture/RDT data review in the post-implementation group (75.2 h vs. 47.9 h, p < 0.001). There was a significant reduction in fluoroquinolone prescribing in the post-implementation group (18.1% vs. 5.4%, p = 0.036). The proportion of patients who had a repeat ED encounter or hospital admission within 30 days was not significantly different between the pre- and post-implementation groups (15.6 vs. 19.1%, p = 0.78 and 9.4% vs. 7.9%, p = 1.0, respectively). Introduction of a pharmacist culture and RDT follow-up program in the ED reduced time to data review, time to outpatient intervention and outpatient follow-up of fluoroquinolone prescribing.

Impact of Reappraisal of Fluoroquinolone Minimum Inhibitory Concentration Susceptibility Breakpoints in Gram-Negative Bloodstream Isolates

The Clinical Laboratory Standards Institute lowered the fluoroquinolone minimum inhibitory concentration (MIC) susceptibility breakpoints for Enterobacteriaceae and glucose non-fermenting Gram-negative bacilli in January 2019. This retrospective cohort study describes the impact of this reappraisal on ciprofloxacin susceptibility overall and in patients with risk factors for antimicrobial resistance. Gram-negative bloodstream isolates collected from hospitalized adults at Prisma Health-Midlands hospitals in South Carolina, USA, from January 2010 to December 2014 were included. Matched pairs mean difference (MD) with 95% confidence intervals (CI) were calculated to examine the change in ciprofloxacin susceptibility after MIC breakpoint reappraisal. Susceptibility of Enterobacteriaceae to ciprofloxacin declined by 5.2% (95% CI: −6.6, −3.8; p < 0.001) after reappraisal. The largest impact was demonstrated among Pseudomonas aeruginosa bloodstream isolates (MD −7.8, 95% CI: −14.6, −1.1; p = 0.02) despite more conservative revision in ciprofloxacin MIC breakpoints. Among antimicrobial resistance risk factors, fluoroquinolone exposure within the previous 90 days was associated with the largest change in ciprofloxacin susceptibility (MD −9.3, 95% CI: −16.1, −2.6; p = 0.007). Reappraisal of fluoroquinolone MIC breakpoints has a variable impact on the susceptibility of bloodstream isolates by microbiology and patient population. Healthcare systems should be vigilant to systematically adopt this updated recommendation in order to optimize antimicrobial therapy in patients with bloodstream and other serious infections.

Empirical fluoroquinolones versus broad-spectrum beta-lactams for Gram-negative bloodstream infections in the absence of antimicrobial resistance risk factors

OBJECTIVES

Increasing antimicrobial resistance rates limit empirical antimicrobial treatment options for Gram-negative bloodstream infections (GN-BSI). However, antimicrobial resistance may be predicted based on patient-specific risk factors using precision medicine concepts. This retrospective, 1:2 matched cohort examined clinical outcomes in hospitalized adults without major risk factors for antimicrobial resistance receiving empirical fluoroquinolones or broad-spectrum beta-lactams (BSBL) for GN-BSI at Prisma Health-Midlands hospitals in Columbia, SC, USA from January 2010 through June 2015.

METHODS

Multivariable logistic regression was used to examine early treatment failure at 72-96 h from GN-BSI. Cox proportional hazards regression was used to examine 28-day mortality and hospital length of stay (HLOS).

RESULTS

Among 74 and 148 patients receiving empirical fluoroquinolones and BSBL for GN-BSI, respectively, median age was 68 years, 159 (72%) were women, and 152 (68%) had a urinary source of infection. Early treatment failure rates were comparable in fluoroquinolone and BSBL groups (27% vs. 30%, respectively, odds ratio 0.82, 95% confidence intervals [CI] 0.43-1.54, P = 0.53), as well as 28-day mortality (8.9% vs. 9.7%, respectively, hazards ratio [HR] 0.74, 95% CI 0.26-1.90, P = 0.54). Median HLOS was 6.1 days in the fluoroquinolone group and 7.1 days in the BSBL group (HR 0.73, 95% CI 0.54-0.99, P = 0.04). Transition from intravenous to oral therapy occurred sooner in the fluoroquinolone group than in the BSBL group (3.0 vs. 4.9 days, P < 0.001).

CONCLUSIONS

In the absence of antimicrobial resistance risk factors, fluoroquinolones provide an additional empirical treatment option to BSBL for GN-BSI. Shorter HLOS in the fluoroquinolone group may be due to earlier transition from intravenous to oral antimicrobial therapy.

Prediction of trimethoprim/sulfamethoxazole resistance in community-onset urinary tract infections

OBJECTIVES

This study aimed to predict trimethoprim/sulfamethoxazole (SXT) resistance in patients with community-onset urinary tract infection (UTI) due to Enterobacteriaceae based on patient-specific risk factors.

METHODS

This was a retrospective case-control study in Prisma Health facilities in central South Carolina, USA, including three community hospitals, affiliated emergency departments and ambulatory clinics, including adult patients with community-onset UTI due to Enterobacteriaceae (1 April 2015 to 29 February 2016). Multivariate logistic regression was used to examine risk factors for SXT resistance.

RESULTS

Among 351 unique patients with community-onset UTI, 71 (20.2%) had SXT-resistant Enterobacteriaceae urinary isolates. Overall, median age was 64 years and 252 (71.8%) were female. A multivariate model identified prior urinary infection/colonisation with SXT-resistant Enterobacteriaceae (OR=8.58, 95% CI 3.92-18.81; P<0.001) and SXT use within past 12 months (OR=2.58, 95% CI 1.13-5.89; P=0.02) as predictors of SXT resistance among urinary isolates. Most patients with UTI (285; 81.2%) had no risk factors for SXT resistance. SXT resistance rates increased from 13% in the absence of risk factors to 31% in patients with prior SXT use, 66% in those with prior urinary infection/colonisation with SXT-resistant Enterobacteriaceae and 73% in the presence of both risk factors.

CONCLUSION

SXT resistance in Enterobacteriaceae urinary isolates may be predicted based on prior urine culture results and SXT use within the previous year. Utilisation of a patient-specific antibiogram may allow empirical SXT use in patients with community-onset UTI in the absence of risk factors for resistance.

Transition from intravenous to oral antimicrobial therapy in patients with uncomplicated and complicated bloodstream infections

BACKGROUND

The role of oral antimicrobial agents in the management of bloodstream infections (BSI) is currently evolving.

OBJECTIVES

This narrative review summarizes and appraises clinical studies that examined transition from intravenous to oral antimicrobials or compared effectiveness of various oral agents for definitive therapy of uncomplicated and complicated BSI in adults.

SOURCES

Relevant English-language studies from MEDLINE (since inception) and presented abstracts at international scientific meetings (since 2017).

CONTENT

Emerging data suggest potential utility of oral switch strategy, particularly to oxazolidinones, as an alternative to standard intravenous therapy in low-risk patients with uncomplicated Staphylococcus aureus BSI. Moreover, results of recent randomized clinical trials are promising that combination oral regimens may play a role in antimicrobial management of complicated Gram-positive BSI, including infective endocarditis, septic arthritis and osteomyelitis. Whereas oral fluoroquinolones have been used successfully for decades in both uncomplicated and complicated Gram-negative BSI, recent studies suggest that trimethoprim-sulfamethoxazole and aminopenicillins represent alternative oral options in uncomplicated Enterobacteriaceae BSI. Oral azoles have been used for definitive therapy of Candida species BSI and are currently recommended by the international management guidelines.

IMPLICATIONS

Recent studies demonstrate that early transition from intravenous to oral therapy is a feasible and effective strategy in most patients with BSI due to Gram-negative bacteria, obligate anaerobic bacteria and Candida species. Oral antimicrobial combinations may be considered in select patients with complicated Gram-positive BSI after 10-14 days of intravenous therapy. Future studies will determine the role of oral agents for switch therapy in uncomplicated Gram-positive BSI.

Agrichemicals and antibiotics in combination increase antibiotic resistance evolution

Antibiotic resistance in our pathogens is medicine's climate change: caused by human activity, and resulting in more extreme outcomes. Resistance emerges in microbial populations when antibiotics act on phenotypic variance within the population. This can arise from either genotypic diversity (resulting from a mutation or horizontal gene transfer), or from differences in gene expression due to environmental variation, referred to as adaptive resistance. Adaptive changes can increase fitness allowing bacteria to survive at higher concentrations of antibiotics. They can also decrease fitness, potentially leading to selection for antibiotic resistance at lower concentrations. There are opportunities for other environmental stressors to promote antibiotic resistance in ways that are hard to predict using conventional assays. Exploiting our previous observation that commonly used herbicides can increase or decrease the minimum inhibitory concentration (MIC) of different antibiotics, we provide the first comprehensive test of the hypothesis that the rate of antibiotic resistance evolution under specified conditions can increase, regardless of whether a herbicide increases or decreases the antibiotic MIC. Short term evolution experiments were used for various herbicide and antibiotic combinations. We found conditions where acquired resistance arises more frequently regardless of whether the exogenous non-antibiotic agent increased or decreased antibiotic effectiveness. This is attributed to the effect of the herbicide on either MIC or the minimum selective concentration (MSC) of a paired antibiotic. The MSC is the lowest concentration of antibiotic at which the fitness of individuals varies because of the antibiotic, and is lower than MIC. Our results suggest that additional environmental factors influencing competition between bacteria could enhance the ability of antibiotics to select antibiotic resistance. Our work demonstrates that bacteria may acquire antibiotic resistance in the environment at rates substantially faster than predicted from laboratory conditions.

Antibiotic consumption and resistance of gram-negative pathogens (collateral damage)

Antibiotics are commonly prescribed in community and hospital care. Overuse and misuse favors emergence and spread of resistant bacteria. The ATC/DDD methodology is commonly used for presenting the drug utilization data. In primary care, the consumption is usually expressed in DDD per 1,000 inhabitants per day, in hospital, preferably in DDD per 100 bed days and DDD per 100 admissions. The alternative metric is days of therapy (DOT), which needs IT support. Antibiotics have ecological adverse effects at individual and population level. Antibiotics select resistant bacteria among pathogens and normal flora. Broad-spectrum antibiotics, low dosage and prolonged antibiotic therapy favor the development of resistance. Although total use of antibiotics in hospital is much less than in the community, the intensity of use magnified by cross infection ensures a multitude of resistant bacteria in today's hospitals. Reversal of resistance is complex and might persist for many years despite the introduction of antimicrobial containment and stewardship programs.

Predicting Resistance to Piperacillin-Tazobactam, Cefepime and Meropenem in Septic Patients With Bloodstream Infection Due to Gram-Negative Bacteria

Background

Predicting antimicrobial resistance in gram-negative bacteria (GNB) could balance the need for administering appropriate empiric antibiotics while also minimizing the use of clinically unwarranted broad-spectrum agents. Our objective was to develop a practical prediction rule able to identify patients with GNB infection at low risk for resistance to piperacillin-tazobactam (PT), cefepime (CE), and meropenem (ME).

Methods

The study included adult patients with sepsis or septic shock due to bloodstream infections caused by GNB admitted between 2008 and 2015 from Barnes-Jewish Hospital. We used multivariable logistic regression analyses to describe risk factors associated with resistance to the antibiotics of interest (PT, CE, and ME). Clinical decision trees were developed using the recursive partitioning algorithm CHAID (χ2 Automatic Interaction Detection).

Results

The study included 1618 consecutive patients. Prevalence rates for resistance to PT, CE, and ME were 28.6%, 21.8%, and 8.5%, respectively. Prior antibiotic use, nursing home residence, and transfer from an outside hospital were associated with resistance to all 3 antibiotics. Resistance to ME was specifically linked with infection attributed to Pseudomonas or Acinetobacter spp. Discrimination was similar for the multivariable logistic regression and CHAID tree models, with both being better for ME than for PT and CE. Recursive partitioning algorithms separated out 2 clusters with a low probability of ME resistance and 4 with a high probability of PT, CE, and ME resistance.

Conclusions

With simple variables, clinical decision trees can be used to distinguish patients at low, intermediate, or high risk of resistance to PT, CE, and ME.

The Rapid Prediction of Carbapenem Resistance in Patients With Klebsiella pneumoniae Bacteremia Using Electronic Medical Record Data

BACKGROUND

The administration of active antibiotics is often delayed in cases of carbapenem-resistant gram-negative bacteremia. Using electronic medical record (EMR) data to rapidly predict carbapenem resistance in patients with Klebsiella pneumoniae bacteremia could help reduce the time to active therapy.

METHODS

All cases of Klebsiella pneumoniae bacteremia at Mount Sinai Hospital from September 2012 through September 2016 were included. Cases were randomly divided into a "training set" and a "testing set." EMR data from the training set cases were reviewed, and significant risk factors for carbapenem resistance were entered into a multiple logistic regression model. Performance was assessed by repeated K-fold cross-validation and by applying the training set model to the testing set. All cases were also reviewed to determine the time to effective antibiotic therapy.

RESULTS

A total of 613 cases of Klebsiella pneumoniae bacteremia were included, 61 (10%) of which were carbapenem-resistant. The training and testing sets consisted of 460 and 153 cases, respectively. The regression model derived from the training set correctly predicted 73% of carbapenem-resistant cases and 59% of carbapenem-susceptible cases in the testing set (sensitivity, 73%; specificity, 59%; positive predictive value, 16%; negative predictive value, 95%). The mean area under the receiver operator characteristic curve of the K-fold cross-validation repeats was 0.731. Patients with carbapenem-resistant infections received active antibiotics significantly later than those with susceptible infections (40.4 hours vs 9.6 hours, P < .0001).

CONCLUSIONS

A multiple logistic regression model using EMR data can generate rapid, sensitive predictions of carbapenem resistance in patients with Klebsiella pneumoniae bacteremia, which could help shorten the time to effective therapy in these cases.

Application of Fluoroquinolone Resistance Score in Management of Complicated Urinary Tract Infections

The fluoroquinolone resistance score (FQRS) predicts the probability of fluoroquinolone resistance with good discrimination. The score has been derived from patients with bloodstream infections caused by Gram-negative bacteria and is based on fluoroquinolone use within the past 6 months, among other clinical and health care exposure criteria. This study aims to examine the utility of the FQRS in patients with complicated urinary tract infections (cUTI) and determine whether extension of prior fluoroquinolone use to 12 months improves model discrimination. Adults with cUTI at Palmetto Health in central South Carolina, USA, from 1 April 2015 through 31 July 2015 were prospectively identified. Multivariate logistic regression was used to examine the association between prior fluoroquinolone use and resistance. Among 238 patients, 54 (23%) had cUTI due to fluoroquinolone-resistant bacteria. Overall, the median age was 66 years, 162 (68%) patients were women, and 137 (58%) patients had cUTI due to Escherichia coli Prior exposure to fluoroquinolones within 3 months (adjusted odds ratio [aOR], 23.4; 95% confidence interval [CI], 8.2 to 76.8; P < 0.001) and within 3 to 12 months (aOR, 13.2; 95% CI, 3.1 to 68.4; P < 0.001) was independently associated with fluoroquinolone resistance compared to no prior use. The area under the receiver operating characteristic curve for the FQRS increased from 0.73 to 0.80 when prior fluoroquinolone use was extended from 6 to 12 months. FQRSs of ≥2 and ≥3 had negative predictive values of 91% and 90%, respectively. The modified FQRS stratifies patients with cUTI on the basis of the predicted probability of fluoroquinolone resistance with very good discrimination. Application of the modified FQRS may improve antimicrobial utilization in patients with acute pyelonephritis.

Clinical Risk Score for Prediction of Extended-Spectrum β-Lactamase–Producing Enterobacteriaceae in Bloodstream Isolates

OBJECTIVE

To develop a risk score to predict probability of bloodstream infections (BSIs) due to extended-spectrum β-lactamase–producing Enterobacteriaceae (ESBLE).

DESIGN

Retrospective case-control study.

SETTING

Two large community hospitals.

PATIENTS

Hospitalized adults with Enterobacteriaceae BSI between January 1, 2010, and June 30, 2015.

METHODS

Multivariate logistic regression was used to identify independent risk factors for ESBLE BSI. Point allocation in extended-spectrum β-lactamase prediction score (ESBL-PS) was based on regression coefficients.

RESULTS

Among 910 patients with Enterobacteriaceae BSI, 42 (4.6%) had ESBLE bloodstream isolates. Most ESBLE BSIs were community onset (33 of 42; 79%), and 25 (60%) were due to Escherichia coli. Independent risk factors for ESBLE BSI and point allocation in ESBL-PS included outpatient procedures within 1 month (adjusted odds ratio [aOR], 8.7; 95% confidence interval [CI], 3.1–22.9; 1 point), prior infections or colonization with ESBLE within 12 months (aOR, 26.8; 95% CI, 7.0–108.2; 4 points), and number of prior courses of β-lactams and/or fluoroquinolones used within 3 months of BSI: 1 course (aOR, 6.3; 95% CI, 2.7–14.7; 1 point), ≥2 courses (aOR, 22.0; 95% CI, 8.6–57.1; 3 points). The area under the receiver operating characteristic curve for the ESBL-PS model was 0.86. Patients with ESBL-PSs of 0, 1, 3, and 4 had estimated probabilities of ESBLE BSI of 0.7%, 5%, 24%, and 44%, respectively. Using ESBL-PS ≥3 to indicate high risk provided a negative predictive value of 97%.

CONCLUSIONS

ESBL-PS estimated patient-specific risk of ESBLE BSI with high discrimination. Incorporation of ESBL-PS with acute severity of illness may improve adequacy of empirical antimicrobial therapy and reduce carbapenem utilization.

Infect Control Hosp Epidemiol 2017;38:266–272

Association between inappropriate empirical antimicrobial therapy and hospital length of stay in Gram-negative bloodstream infections: stratification by prognosis

OBJECTIVES

The potential benefit from appropriate empirical antimicrobial therapy in patients with favourable prognosis at initial presentation with Gram-negative bloodstream infection (BSI) remains unclear. This retrospective cohort study examined the impact of inappropriate empirical antimicrobial therapy on hospital length of stay (HLOS) following Gram-negative BSI after stratification by predicted prognosis using the BSI mortality risk score (BSIMRS).

METHODS

Hospitalized adults with first episodes of Gram-negative BSI from 1 January 2010 to 31 December 2013 at Palmetto Health Hospitals in Columbia, SC, USA were identified. Multivariate Cox proportional hazards regression was used to examine the association between inappropriate empirical antimicrobial therapy and HLOS overall and within each predefined BSIMRS category (<5 and ≥5).

RESULTS

Among 830 unique patients with Gram-negative BSI, 469 and 361 had BSIMRS <5 and ≥5, respectively. Overall, the median age was 65 years, 448 (54%) were women, Escherichia coli (444; 53%) was the most common bloodstream isolate and 444 (53%) had a urinary source of infection. After adjustments in the multivariate model, BSIMRS (HR = 1.14 per point, 95% CI = 1.11-1.17, P < 0.001) and inappropriate empirical antimicrobial therapy (HR = 1.41, 95% CI = 1.07-1.91, P = 0.01) were independently associated with increased risk of remaining hospitalized following Gram-negative BSI. Median HLOS with appropriate and inappropriate empirical antimicrobial therapy was 7 and 10 days, respectively, in patients with BSIMRS <5 (P = 0.03) and 13 and 17 days, respectively, in those with BSIMRS ≥5 (P = 0.02).

CONCLUSIONS

Inappropriate empirical antimicrobial therapy is associated with prolonged HLOS following Gram-negative BSI in patients with both good and guarded prognosis.

Risk factors for resistance to ciprofloxacin in community-acquired urinary tract infections due to Escherichia coli in an elderly population

BACKGROUND

Antimicrobial resistance to ciprofloxacin is rising worldwide, especially in bacteria causing urinary tract infections (UTIs). Prudent use of current antibiotic drugs is therefore necessary.

OBJECTIVES

We analysed (modifiable) risk factors for ciprofloxacin-resistant Escherichia coli.

METHODS

Urinary cultures of UTIs caused by E. coli were collected from participants in the Rotterdam Study, a prospective cohort study in an elderly population, and analysed for susceptibility to ciprofloxacin. Multivariate logistic regression was performed to investigate several possible risk factors for resistance.

RESULTS

Ciprofloxacin resistance in 1080 E. coli isolates was 10.2%. Multivariate analysis showed that higher age (OR 1.03; 95% CI 1.00-1.05) and use of two (OR 5.89; 95% CI 3.45-10.03) and three or more (OR 3.38; 95% CI 1.92-5.97) prescriptions of fluoroquinolones were associated with ciprofloxacin resistance, while no association between fluoroquinolone use more than 1 year before culture and ciprofloxacin resistance could be demonstrated. Furthermore, a high intake of pork (OR 3.68; 95% CI 1.36-9.99) and chicken (OR 2.72; 95% CI 1.08-6.85) and concomitant prescription of calcium supplements (OR 2.51; 95% CI 1.20-5.22) and proton pump inhibitors (OR 2.04; 95% CI 1.18-3.51) were associated with ciprofloxacin resistance.

CONCLUSIONS

Ciprofloxacin resistance in community-acquired UTI was associated with a high intake of pork and chicken and with concomitant prescription of calcium supplements and proton pump inhibitors. Modification of antibiotic use in animals as well as temporarily stopping the prescription of concomitant calcium and proton pump inhibitors need further evaluation as strategies to prevent ciprofloxacin resistance.