ESBL Escherichia coli Isolates Have Enhanced Gut Colonization Capacity Compared to Non-ESBL Strains in Neonatal Mice.

Extended-spectrum beta-lactamase (ESBL)-producing Escherichia coli can cause invasive infections in infants and immunocompromised children with high associated morbidity and mortality. The gut is a major reservoir of these strains in the community. Current dogma dictates that antimicrobial resistance is associated with a fitness cost. However, recent data show that some contemporary ESBL E. coli strains may be more "fit" compared to nonresistant E. coli strains. Here, we use whole-genome sequencing to first characterize 15 ESBL E. coli strains isolated from infants in a Pakistani community, a clinical extraintestinal pathogenic ESBL E. coli ST131 strain, and a non-ESBL commensal E. coli strain, and then use a novel animal model of early life gut colonization to assess the ability of these strains to colonize the infant mouse gut. We determined that CTX-M-15 was present in all the ESBL strains, as well as additional beta-lactamases and genes conferring resistance to multiple antibiotic classes. In the animal model, 11/16 ESBL E. coli strains had significantly higher burden of colonization at week four of life compared to commensal strains, even in the absence of selective antibiotic pressure, suggesting that these strains may have enhanced fitness despite being highly antimicrobial resistant. IMPORTANCE Antimicrobial resistance is a global public health emergency. Infants, especially preterm infants and those in the neonatal intensive care unit, immunocompromised hosts, and those with chronic illnesses are at highest risk of adverse outcomes from invasive infections with antimicrobial-resistant strains. It has long been thought that resistance is associated with a fitness cost, i.e., antimicrobial-resistant strains are not able to colonize the gut as well as nonresistant strains, and that antibiotic exposure is a key risk factor for persistent colonization with resistant strains. Here, we use a novel infant mouse model to add to the growing body of literature that some highly-resistant contemporary Escherichia coli strains can persist in the gut with a significant burden of colonization despite absence of antibiotic exposure.

Drivers of Prolonged Outpatient Antibiotic Therapy for Urinary Tract Infections and Community-Acquired Pneumonia.

BACKGROUND: Variability exists in treatment duration for community-acquired pneumonia (CAP) and urinary tract infection (UTI) in children and may be associated with non-clinical factors. METHODS: A retrospective study was conducted of patients treated for outpatient CAP and UTI in a children's hospital network from 2016 to 2019. Multivariable logistic regression was performed to identify predictors of long antibiotic duration (≥10 days). Hospitalization within 30 days was determined. RESULTS: Overall, 2124 prescriptions for CAP and 1116 prescriptions for UTI were included. Prescriptions were ≥10 days in 59.9% and 47.6% for CAP and UTI, respectively. Long durations were more common in the emergency department (ED) than in clinics for UTI's (P = .0082), and more common in convenient care for CAP (P = .045). In UTI's, Asian and Hispanic patients received shorter durations than white patients. Younger children had greater odds of long duration for both diagnoses. Medicaid insurance was associated with long therapy for UTI (OR: 1.660, P = .0042) and CAP (OR: 1.426, P = .0169). Residents and fellows were less likely to give long durations than attending physicians (P < .0001). APNs were more likely to administer long therapies in CAP (P = .0062). Subsequent hospitalizations were uncommon for UTI (n = 10) and CAP (n = 20). CONCLUSIONS: Younger age, Medicaid insurance, ED, and convenient care visits were associated with a long duration of therapy. Residents and fellows were less likely to give long durations.

Evaluation of Physicians' Attitudes Regarding Transport Modalities.

OBJECTIVE: Hospital mergers have made interhospital transfers necessary in the consolidation of medical services. Physicians must make decisions on the level of interfacility transport modalities (ITMs). We sought to assess physician knowledge of and comfort with ITMs. METHODS: A survey was e-mailed to 2,510 physicians in a health care system. Participation was voluntary and anonymous. The mean and median Likert values were calculated overall. Similar calculations were performed for emergency medicine physicians (EMPs) and critical care physicians (CCPs). These calculations were compared with those for noncritical care physicians (NCCPs) using the t-test and Mann-Whitney test. RESULTS: Of the 181 physicians who responded, 169 physicians identified a specialty. Sixty-nine were EMPs/CCPs, whereas 100 were NCCPs. The mean and median Likert values were statistically significantly higher for EMPs/CCPs compared with NCCPs (P < .0001) in the areas of knowledge of ITMs, comfort in choosing ITMs, and knowledge in choosing ground versus air critical care transport (CCT). The most important factor for using ground or air CCT was patient stability. Sixty percent believed air CCT to be faster than ground. CONCLUSION: EMPs/CCPs seem to be more comfortable with ITMs than NCCPs. Further research should evaluate whether educational interventions lead to a more appropriate use of ITMs.

Use of the Electronic Health Record to Optimize Antimicrobial Prescribing.

PURPOSE: This review summarizes how interventions in the electronic health record (EHR) can optimize antimicrobial stewardship across the continuum of antimicrobial decision making, from diagnosis of infection to discontinuation of therapy. In addition, opportunities to optimize provider communication and patient education are identified. METHODS: A narrative review was conducted to identify how interventions in the EHR can influence antimicrobial prescribing behavior. Examples from pediatrics were specifically identified. Interventions were then categorized into high-impact/low-effort, high-impact/high-effort, and low-impact/low-effort groupings based on historical experience. FINDINGS: EHR-based interventions can be used for stratifying patients at risk for infection and are useful in identifying patients with new-onset infections. Additional tools include automatically updated antibiograms tailored to specific patient populations, timely authorization of restricted antimicrobials, and more accurate allergy labeling. Medical errors can be reduced and communication between providers can be improved by standardized data fields. Clinical decision support tools can guide appropriate selection of therapy, and visual prompts can reduce unnecessarily prolonged therapy. Benchmarking of antimicrobial use, tailored patient education, and improved communication during transitions of care are enhanced through EHR-based interventions. IMPLICATIONS: Prescribing behavior can be modified through a range of interventions in the EHR, including tailored education, alerts, prompts, and restrictions on provider behavior. Further studies are needed to compare the effectiveness of various strategies.

Measuring Drug-Induced Changes in Metabolite Populations of Live Bacteria: Real Time Analysis by Raman Spectroscopy.

Raman difference spectroscopy is shown to provide a wealth of molecular detail on changes within bacterial cells caused by infusion of antibiotics or hydrogen peroxide. Escherichia coli strains paired with chloramphenicol, dihydrofolate reductase propargyl-based inhibitors, meropenem, or hydrogen peroxide provide details of the depletion of protein and nucleic acid populations in real time. Additionally, other reproducible Raman features appear and are attributed to changes in cell metabolite populations. An initial candidate for one of the metabolites involves population increases of citrate, an intermediate within the tricarboxyclic acid cycle. This is supported by the observation that a strain of E. coli without the ability to synthesize citrate, gltA, lacks an intense feature in the Raman difference spectrum that has been ascribed to citrate. The methodology for obtaining the Raman data involves infusing the drug into live cells, then washing, freezing, and finally lyophilizing the cells. The freeze-dried cells are then examined under a Raman microscope. The difference spectra [cells treated with drug] - [cells without treatment] are time-dependent and can yield population kinetics for intracellular species in vivo. There is a strong resemblance between the Raman difference spectra of E. coli cells treated with meropenem and those treated with hydrogen peroxide.

Defining Molecular Details of the Chemistry of Biofilm Formation by Raman Microspectroscopy.

Two protocols that allow for the comparison of Raman spectra of planktonic cells and biofilm formed from these cells in their growth phase have been developed. Planktonic cells are washed and flash-frozen in <1 min to reduce the time for metabolic changes during processing, prior to freeze-drying. Biofilm is formed by standing cells in 50 µL indentations in aluminum foil in an atmosphere of saturated water vapor for 24-48 h. The results for Escherichia coli type K12 cells, which do not readily form biofilm, are compared to those for Staphylococcus epidermidis cells, which prolifically synthesize biofilm. For E. coli, the Raman spectra of the planktonic and biofilm samples are similar with the exception that the spectral signature of RNA, present in planktonic cells, could not be detected in biofilm. For S. epidermidis, major changes occur upon biofilm formation. In addition to the absence of the RNA features, new bands occur near 950 cm-1 and between 1350 and 1420 cm-1 that are associated with an increase in carbohydrate content. Unlike the case in E. coli biofilm, the intensity of G base ring modes is reduced in but A and T base ring signatures become more prominent. For S. epidermis in the biofilm's amide III region, there is evidence of an increase in the level of ß-sheet structure accompanied by a decrease in α-helical content. The presence of biofilm is confirmed by microscope-aided photography and, separately, by staining with methyl violet.