Unbiased identification of risk factors for invasive Escherichia coli disease using machine learning.

BACKGROUND: Invasive Escherichia coli disease (IED), also known as invasive extraintestinal pathogenic E. coli disease, is a leading cause of sepsis and bacteremia in older adults that can result in hospitalization and sometimes death and is frequently associated with antimicrobial resistance. Moreover, certain patient characteristics may increase the risk of developing IED. This study aimed to validate a machine learning approach for the unbiased identification of potential risk factors that correlate with an increased risk for IED. METHODS: Using electronic health records from 6.5 million people, an XGBoost model was trained to predict IED from 663 distinct patient features, and the most predictive features were identified as potential risk factors. Using Shapley Additive predictive values, the specific relationships between features and the outcome of developing IED were characterized. RESULTS: The model independently predicted that older age, a known risk factor for IED, increased the chance of developing IED. The model also predicted that a history of ≥ 1 urinary tract infection, as well as more frequent and/or more recent urinary tract infections, and ≥ 1 emergency department or inpatient visit increased the risk for IED. Outcomes were used to calculate risk ratios in selected subpopulations, demonstrating the impact of individual or combinations of features on the incidence of IED. CONCLUSION: This study illustrates the viability and validity of using large electronic health records datasets and machine learning to identify correlating features and potential risk factors for infectious diseases, including IED. The next step is the independent validation of potential risk factors using conventional methods.

Impact of Invasive Escherichia Coli Disease on Clinical Outcomes and Medical Resource Utilization Among Asian Patients in the United States.

INTRODUCTION: Invasive Escherichia coli disease (IED) can lead to sepsis and death and is associated with a substantial burden. Yet, there is scarce information on the burden of IED in Asian patients. METHODS: This retrospective study used US hospital data from the PINC AI™ Healthcare database (October 2015-March 2020) to identify IED cases among patients aged ≥ 60 years. IED was defined as a positive E. coli culture in blood or other normally sterile body site (group 1 IED) or positive culture of E. coli in urine with signs of sepsis (group 2 IED). Eligible patients with IED were classified into Asian and non-Asian cohorts based on their reported race. Entropy balancing was used to create cohorts with similar characteristics. Outcomes following IED were descriptively reported in the balanced cohorts. RESULTS: A total of 646 Asian and 19,127 non-Asian patients with IED were included (median age 79 years; 68% female after balancing). For both cohorts, most IED encounters had community-onset (> 95%) and required hospitalization (Asian 96%, mean duration 6.9 days; non-Asian 95%, mean duration 6.8 days), with frequent admission to intensive care (Asian 35%, mean duration 3.3 days; non-Asian 34%, mean duration 3.5 days), all standardized differences [SD] < 0.20. Compared to non-Asian patients, Asian patients were more likely to be discharged home (54% vs. 43%; SD = 0.22), and less likely to be discharged to a skilled nursing facility (24% vs. 31%; SD = 0.16). In-hospital fatality rates during the IED encounter were similar across cohorts (Asian 9%, non-Asian 10%; SD = 0.01). Most E. coli isolates showed resistance to ≥ 1 antibiotic (Asian 61%; non-Asian 64%) and 36% to ≥ 3 antibiotic classes (all SD < 0.20). CONCLUSION: IED is associated with a substantial burden, including need for intensive care and considerable mortality, in Asian patients in the USA that is consistent with that observed for non-Asian patients.

Clinical presentation and antimicrobial resistance of invasive Escherichia coli disease in hospitalized older adults: a prospective multinational observational study.

BACKGROUND: Clinical data characterizing invasive Escherichia coli disease (IED) are limited. We assessed the clinical presentation of IED and antimicrobial resistance (AMR) patterns of causative E. coli isolates in older adults. METHODS: EXPECT-2 (NCT04117113) was a prospective, observational, multinational, hospital-based study conducted in patients with IED aged ≥ 60 years. IED was determined by the microbiological confirmation of E. coli from blood; or by the microbiological confirmation of E. coli from urine or an otherwise sterile body site in the presence of requisite criteria of systemic inflammatory response syndrome (SIRS), Sequential Organ Failure Assessment (SOFA), or quick SOFA (qSOFA). The primary outcomes were the clinical presentation of IED and AMR rates of E. coli isolates to clinically relevant antibiotics. Complications and in-hospital mortality were assessed through 28 days following IED diagnosis. RESULTS: Of 240 enrolled patients, 80.4% had bacteremic and 19.6% had non-bacteremic IED. One-half of infections (50.4%) were community-acquired. The most common source of infection was the urinary tract (62.9%). Of 240 patients, 65.8% fulfilled ≥ 2 SIRS criteria, and 60.4% had a total SOFA score of ≥ 2. Investigator-diagnosed sepsis and septic shock were reported in 72.1% and 10.0% of patients, respectively. The most common complication was kidney dysfunction (12.9%). The overall in-hospital mortality was 4.6%. Of 299 E. coli isolates tested, the resistance rates were: 30.4% for trimethoprim-sulfamethoxazole, 24.1% for ciprofloxacin, 22.1% for levofloxacin, 16.4% for ceftriaxone, 5.7% for cefepime, and 4.3% for ceftazidime. CONCLUSIONS: The clinical profile of identified IED cases was characterized by high rates of sepsis. IED was associated with high rates of AMR to clinically relevant antibiotics. The identification of IED can be optimized by using a combination of clinical criteria (SIRS, SOFA, or qSOFA) and culture results.

Enhancement and external validation of algorithms using diagnosis codes to identify invasive Escherichia coli disease.

OBJECTIVE: To assess the predictive accuracy of code-based algorithms for identifying invasive Escherichia coli (E. coli) disease (IED) among inpatient encounters in US hospitals. METHODS: The PINC AI Healthcare Database (10/01/2015-03/31/2020) was used to assess the performance of six published code-based algorithms to identify IED cases among inpatient encounters. Case-confirmed IEDs were identified based on microbiological confirmation of E. coli in a normally sterile body site (Group 1) or in urine with signs of sepsis (Group 2). Code-based algorithm performance was assessed overall, and separately for Group 1 and Group 2 based on sensitivity, specificity, positive and negative predictive value (PPV and NPV) and F1 score. The improvement in performance of refinements to the best-performing algorithm was also assessed. RESULTS: Among 2,595,983 encounters, 97,453 (3.8%) were case-confirmed IED (Group 1: 60.9%; Group 2: 39.1%). Across algorithms, specificity and NPV were excellent (>97%) for all but one algorithm, but there was a trade-off between sensitivity and PPV. The algorithm with the most balanced performance characteristics included diagnosis codes for: (1) infectious disease due to E. coli OR (2) sepsis/bacteremia/organ dysfunction combined with unspecified E. coli infection and no other concomitant non-E. coli invasive disease (sensitivity: 56.9%; PPV: 56.4%). Across subgroups, the algorithms achieved lower algorithm performance for Group 2 (sensitivity: 9.9%-61.1%; PPV: 3.8%-16.0%). CONCLUSIONS: This study assessed code-based algorithms to identify IED during inpatient encounters in a large US hospital database. Such algorithms could be useful to identify IED in healthcare databases that lack information on microbiology data.

Safety, Reactogenicity, Immunogenicity, and Dose Selection of 10-Valent Extraintestinal Pathogenic Escherichia coli Bioconjugate Vaccine (VAC52416) in Adults Aged 60-85 Years in a Randomized, Multicenter, Interventional, First-in-Human, Phase 1/2a Study.

Background: ExPEC10V is a bioconjugate vaccine containing O-antigen polysaccharides of 10 extraintestinal pathogenic Escherichia coli (ExPEC) serotypes. This phase 1/2a study (NCT03819049) assessed the safety, reactogenicity, and immunogenicity of ExPEC10V (VAC52416) to prevent invasive E coli disease in elderly adults. Methods: The observer-blind, active-controlled design included a 28-day screening, vaccination, 181-day follow-up, and 1-year follow-up. Participants (60-85 years of age) were randomized to ExPEC10V low dose (antigen dose range, 4-8 µg), ExPEC10V medium dose (4-16 µg), or ExPEC10V high dose (8-16 µg); 4-valent ExPEC vaccine (ExPEC4V); or 13-valent pneumococcal conjugate vaccine (PCV13). The incidence of adverse events (AEs; solicited, day 15; unsolicited, day 30; serious AEs, day 181) and immunogenicity (electrochemiluminescent-based assay [ECL] and multiplex opsonophagocytic assay [MOPA]) were assessed. Optimal ExPEC10V dose was determined from safety data through day 30 and an immunogenicity dose selection algorithm based on day 15 ECL and MOPA results. Results: A total of 416 participants were included (median age, 64.0 years; 54.8% female). The incidences of solicited local and systemic AEs were, respectively, 44.2% and 39.4% for low-dose, 52.9% and 46.1% for medium-dose, 57.7% and 45.2% for high-dose ExPEC10V, and 74.1% and 48.1% for PCV13. Five serious AEs, not vaccine related, were reported. The ECL revealed a robust antibody response to ExPEC10V through year 1. Opsonophagocytic killing activity was detected against all but serotype O8; this lack of response against serotype O8 was linked to low assay sensitivity. Based on the totality of data, high-dose ExPEC10V was considered optimal. Conclusions: ExPEC10V was well tolerated and immunogenic in elderly adults against all but serotype O8.