Determine site attributes affecting uptake of a national, multi-site antimicrobial stewardship project focused on reducing unnecessary urine culture orders and antibiotic prescriptions for asymptomatic bacteriuria (ASB).

Forty-one United States Veterans Affairs Medical Centers (VAMCs) were randomized into two arms: “Technical Assistance” (TA) or the “Virtual Learning Collaborative” (VLC). TA sites could request one-on-one guidance from the coordinating center, and VLC sites participated in semimonthly informational webinars. Otherwise, we provided TA and VLC sites with the same resources to educate healthcare professionals on proper ASB management. At the intervention’s conclusion, we conducted 45–60-minute semi-structured interviews with local site leads and focus groups with their teams. TA and VLC sites were sampled based on time spent on intervention implementation, resulting in 19 participants across varying roles (e.g., pharmacists, physicians). We iteratively coded transcripts using thematic analysis.

We propose five considerations for how site attributes informed uptake of our stewardship initiative: (1) existing connections with leadership and important contacts; (2) workplace structure and culture; (3) previous and concurrent antimicrobial stewardship initiatives; (4) availability of existing resources (e.g., time); and (5) personnel turnover (e.g., daily provider transition, job turnover).

We recommend conducting pre-intervention interviews to identify site-specific obstacles and then tailoring the intervention to site-specific needs. Additionally, we stress the importance of sites’ prior experiences with antibiotic stewardship and the availability of personnel to work on stewardship initiatives. Ultimately, a better understanding of sites’ unique environments is an essential step for improving uptake of antimicrobial stewardship projects.

Unnecessary urine cultures contribute to inappropriate antibiotic use, antimicrobial resistance, and Clostridioides difficile infection, particularly when asymptomatic bacteriuria (ASB) is misclassified as infection. We evaluated the diagnostic stewardship impact of an algorithm-based best practice alert (BPA) embedded in the electronic medical record (EMR) to guide urine testing in hospitalized adults.

This prospective quality improvement study was conducted at a 740-bed tertiary care hospital. The BPA triggered when urinalysis with reflex to culture (UACC) was ordered for patients admitted ≥48 hours, guiding clinicians through an algorithm-based ordering workflow. Monthly rates of UACC and pan-culturing were compared between a 12-month pre-intervention period and a 12-month post-intervention period using interrupted time series (ITS) and Mood’s median two-sample test. Gram-negative rod (GNR) bacteremia rates were monitored for safety.

Urine testing decreased from 6.45 to 4.41 tests per 1,000 patient-days (31.6% reduction; P < .01), and pan-culturing decreased from 3.47 to 2.70 per 1,000 patient-days (22.2% reduction; P < .01). ITS showed declining trends both before and after implementation, without significant immediate changes in level or slope following the intervention. CAUTI rates remained stable (0.91 vs 0.82 per 1,000 catheter-days; P = .68), as did rates of gram-negative rod bacteremia (0.47 vs 0.70 per 1,000 patient-days; P = .22). Algorithm adherence averaged 63.9% and increased over time (P < .01). CAUTI cases classified as potential asymptomatic bacteriuria declined from 31.8% to 25.0% (P = .68).

An EMR-integrated, algorithm-based BPA coincided with sustained lower urine testing and pan-culturing rates without adverse safety signals within the context of existing downward trends. Ongoing monitoring is needed to sustain adoption and appropriate use.

To analyze the correlation between scale of Candida auris clustering and extent of environmental contamination, exploring its implications for improving infection control measures in healthcare settings.

Retrospective observational study.

An acute hospital and a convalescent hospital in Hong Kong.

Laboratory confirmed C. auris carriers diagnosed in two hospitals between March and October 2023.

C. auris screening was conducted on patients with defined risk factors and through regular surveillance. Environmental samples were collected from high-touch surfaces and air grilles in corresponding wards.

One-hundred-seventy new C. auris patients were identified, including 65 from nine outbreaks and 105 sporadic cases. Environmental screening from eight outbreaks and 46 sporadic isolated cases was analyzed. The environmental contamination rate was significantly higher in the outbreak group (15.1% vs 2.6%, P < 0.05). Longer outbreak duration (34 vs 7 days, P < 0.05) and a higher median number of affected patients per outbreak (9 vs 5, P = 0.05) were associated with higher contamination rates. Notably, air grille samples had a significantly higher contamination rate than high-touch surfaces in both the outbreak (31.6% vs 10.2%, P < 0.05) and sporadic groups (4.6% vs 1.5%, P < 0.05).

Prolonged and sizeable C. auris clusterings were linked with more extensive environmental contamination, particularly in air grilles, which are often overlooked during decontamination

These findings underscore the need for enhanced infection control measures, including thorough environmental decontamination of air ventilation systems, to mitigate transmission risks in healthcare settings.

Pneumocystis jirovecii pneumonia is a serious opportunistic infection in immunocompromised individuals. Despite recognized person-to-person transmission and healthcare-associated outbreaks, optimal infection control strategies remain unclear. The COVID-19 pandemic led to the implementation of universal masking and physical distancing in hospitals, providing a unique setting to observe P. jirovecii transmission under stringent “droplet precaution”-like conditions. This study investigated healthcare-associated P. jirovecii transmission between June 2020 and November 2021.

Retrospective cohort study.

One tertiary-care hospital in Montréal, QC, Canada.

All patients with P. jirovecii pneumonia at our institution during that period.

Cases were identified via laboratory data and chart review. P. jirovecii-positive samples underwent genotyping using multilocus sequence typing. A transmission map was constructed based on shared genotypes and spatiotemporal overlap of hospital visits within a defined window of potential exposure.

Twenty-eight P. jirovecii pneumonia cases were identified. Genotyping succeeded at providing a distinct sequence type (ST) in 21 cases, revealing 7 patients with shared genotypes (3 with ST52, 2 with STX7, 2 with ST19). The transmission map of 12 patients with shared or unknown genotypes revealed 34 same-day and 34 within-one-day contacts, exclusively within outpatient clinics and imaging facilities. Three spatiotemporal clusters of plausible healthcare-associated transmission were identified despite universal masking.

The occurrence of plausible healthcare-associated P. jirovecii transmission despite stringent universal masking suggests that traditional “droplet precautions” alone may be insufficient to prevent spread, supporting airborne transmission. Infection prevention strategies may need to be expanded in high-risk settings and should account for airborne transmission.

Environmental contamination is a significant factor in the transmission of healthcare-associated infections. This study evaluated the effectiveness of gaseous ozone in reducing microbial contamination within hospital settings, focusing on both air and surfaces.

A cross-sectional study conducted in a single healthcare facility.

Two experimental protocols were implemented using the Sany-Plus ozonator: a three-timepoint protocol (3T), which included routine cleaning followed by ozone disinfection, and a two-timepoint protocol (2T), where ozone treatment occurred after patient activity without intermediate cleaning. We conducted active air and surface samplings using RODAC plates. Microbial loads were quantified in terms of colony-forming units (CFU) for both bacteria and molds/yeasts. Data analysis was performed using the Friedman and Wilcoxon tests, applying Bonferroni correction for multiple comparisons.

In the 3T protocol, bacterial and fungal contamination on surfaces significantly decreased between patient activity and postcleaning, with P-values of .0024 and .0060, respectively. However, no further reduction was observed after ozone disinfection. Airborne microbial loads did not show any statistically significant changes. In the 2T protocol, there was no significant overall decrease in bacterial contamination (P = .0766), although fungal contamination did experience a significant reduction (P = .0058). No changes in air contamination were detected.

These findings suggest that routine cleaning is the primary determinant of surface decontamination, while ozone treatment may provide a limited complementary antifungal effect. No significant impact on airborne microorganisms was noted. Therefore, ozone should be regarded as an adjunct to, rather than a replacement for, conventional cleaning and disinfection in healthcare environments.

To report on the investigation of a pseudo-outbreak of Mycobacterium mucogenicum and examine its clinical and operational impact.

Outbreak investigation, retrospective cohort study.

Academic medical center in Los Angeles, California.

Patients whose bronchoalveolar lavage (BAL) cultures grew M. mucogenicum from 2020–2024.

We performed an institutional outbreak investigation of M. mucogenicum, reviewed electronic medical records of a subset of affected patients (2023–2024), and assessed the operational impact.

The incidence of M. mucogenicum in BAL cultures at Hospital A increased from 6.1% (29/473) in 2020 to 18.6% (29/156) in the first quarter of 2024. Epidemiologic investigation revealed non-sterile ice baths used to cool uncapped sterile syringes during bronchoscopy procedures as the contamination source. Next generation sequencing linked clinical isolates to M. mucogenicum recovered from a perioperative ice machine. Nearly all (157/160) clinical isolates grew from nocardia media rather than acid-fast bacilli media. Among 154 patients, including 51 (33.1%) who were highly immunocompromised, no true infections were identified. Thirty-nine (25.3%) patients were referred to infectious diseases for consultation, seven (4.5%) underwent additional workup, and only one received targeted treatment. The pseudo-outbreak incurred 458 hours of microbiology technologist and infection preventionist time and cost the laboratory $88,426.

A four-year pseudo-outbreak of M. mucogenicum traced to contaminated ice baths used during bronchoscopy resulted in unnecessary infectious disease referrals and substantial operational and financial burden to the institution. Avoidance of non-sterile ice use in procedures prevents costly and burdensome pseudo-outbreaks of environmental mycobacteria in healthcare settings.

To evaluate the association between department-level implementation of broad multiplex respiratory pathogen panels and antibacterial prescribing.

Retrospective repeated cross-sectional staggered-adoption study with cohort-stacked difference-in-differences models.

US emergency and urgent-care departments contributing deidentified electronic health record data to Epic Cosmos, 2016–2025.

Adults with acute respiratory illness without a concurrent clear nonrespiratory bacterial indication.

Incident implementation was the first sustained increase in strict panel use from below 2% to at least 10%, with complete 8-quarter preimplementation and postimplementation windows. Adopters were compared with contemporaneous nonadopters and not-yet-adopters. The primary outcome was a same-day systemic oral or enteral antibacterial order after emergency department discharge.

Among 55,059,432 eligible encounters, 32 adopting departments in 9 cohorts were compared with 1,075 departments, yielding 114,832 department-quarter observations. Implementation was associated with a 23.53-percentage-point increase in panel use (95% CI, 20.32 to 26.74) and a 1.21-percentage-point reduction in same-day antibacterial prescribing (95% CI, −2.37 to −0.05; P = .041), a 5.7% relative decrease. However, alternative 5, 15, and 20% thresholds were not statistically significant, unrestricted analyses were null, and several preimplementation diagnostics rejected flat trends.

Sustained panel implementation was associated with a small reduction in antibiotic prescribing. Threshold sensitivity and mixed preimplementation diagnostics limit causal interpretation and support integrating testing into broader diagnostic-stewardship workflows.

Estimate bacterial pathogen contamination of healthcare workers’ (HCW) long-sleeved attire.

Prospective observational study.

Tertiary care hospital.

HCWs wearing long-sleeved attire providing direct inpatient care.

Sampling of both sleeves of HCWs wearing long-sleeved attire was performed using a swab and cultured for aerobic bacterial growth classified as potential pathogens or presumptive skin commensals. Potential predictors of sleeve contamination, including participant survey responses related to attire and infection prevention practices, were analyzed using univariate analyses. Whole genome sequencing compared isolates to a genomic surveillance database of patient clinical isolates.

Among 280 samples, 81.1% (n = 227) demonstrated any bacterial growth and 20.7% (n = 58) grew ≥1 potential pathogen. Speciated organisms included alpha-hemolytic streptococci (n = 28), Bacillus sp. (n = 20), and Pantoea/Mixta sp. (n = 8), gram-negative bacilli (n = 6), and Staphylococcus aureus (n = 2). Univariate analysis demonstrated that sleeves sampled on non-intensive care units (P = .038) were significantly associated with any bacterial growth, and attire type (P = .002) and sleeve material (P = .004) were associated with growth of ≥1 potential pathogen. Fleece attire and material were more likely to be contaminated than other attire and material types. Sequenced isolates from sleeve samples were not genetically related to any patient isolates.

HCW long sleeve contamination occurs frequently, including with potential pathogens. Changing trends in attire type may have an impact on bacterial transmissibility. While this study could not infer transmission events associated with clinically diagnosed patient infections, the potential benefit of a “bare below the elbows” attire policy warrants further investigation.

To understand current practices for screening, implementing and clearing contact precautions (CP) for multidrug-resistant organisms (MDRO).

A survey of US-based SHEA Research Network (SRN) members and Community Healthcare Epidemiologists and Stewards (CHES).

The SRN is a consortium of healthcare facilities collaborating on multicenter research projects. CHES is an SRN-affiliated group of community hospital epidemiologists.

A Research Electronic Data Capture (REDCap) survey was administered between January 7, 2025 and February 25, 2025. Frequencies and proportions were calculated in REDCap and Microsoft Excel. Fisher’s exact and χ2 tests were calculated in R.

Of 134 facilities, 60 (45%) responded; 4 answered only demographic questions and were excluded from analysis. Most facilities (70%) performed active surveillance for ≥1 MDRO, but only 16% employed preemptive CP. All respondents reported using CP for ≥1 MDRO. CP were employed more often for infection than colonization. Clearance protocols to discontinue CP were common for MRSA (97%), C. difficile (95%), and vancomycin-resistant enterococci (82%), but uncommon for Gram-negatives. Training and adherence of frontline staff (70%), unavailability of private rooms (41%), and lack of evidence-based strategies to eradicate reservoirs (34%) were the top 3 identified barriers to MDRO infection control and prevention.

Infection control practices varied by infection, colonization, specific MDRO, and time to clearance. Variation may indicate need for more data on transmission risk by disease state, organism, and time. Evidence-based strategies are needed to guide MDRO prioritization and CP duration and clearance policies. Guidance should acknowledge limitations in adherence, availability of private rooms, and persistent environmental reservoirs as major barriers to MDRO containment identified in this survey.

Respiratory virus transmission in healthcare settings is not well understood. To investigate the transmission dynamics of common healthcare-associated respiratory virus infections, we performed retrospective whole genome sequencing (WGS) surveillance at three teaching hospitals.

From January 2, 2018, to January 4, 2020, nasal swab specimens positive for rhinovirus, influenza virus, human metapneumovirus (HMPV), or respiratory syncytial virus (RSV) from patients hospitalized for ≥3 days were sequenced. High-quality genomes were assessed for genetic relatedness using ≤3 single nucleotide polymorphisms (SNPs) as a cutoff, except for rhinovirus (≤10 SNPs). Patient health records were reviewed for genetically related clusters to identify epidemiological connections.

We collected 436 viral specimens from 359 patients: rhinovirus (n = 291), influenza virus (n = 50), RSV (n = 48), and HMPV (n = 47). Of these, 42%% (152/359 patients) were from a pediatric hospital, and 58% were from adult hospitals. WGS was performed on 61.2% (178/291) rhinovirus, 78% (39/50) influenza virus, 90% (43/48) RSV, and all HMPV specimens. Among high-quality genomes, we identified 14 genetically related clusters involving 36 patients (range: 2–5 patients per cluster). We identified common epidemiological links for 53% (19/36) of clustered patients; 63% (12/19) of patients had same-unit stays, 26% (5/19) had overlapping hospital stays, and 11% (2/19) shared common providers. On average, genetically related clusters spanned 16 days (range: 0 − 55 days).

WGS offered new insights into respiratory virus transmission dynamics. These advancements could potentially improve infection prevention and control strategies, leading to enhanced patient safety and healthcare outcomes.

Antimicrobial Stewardship Programs (ASPs) need healthcare economic analyses to support and inform ASP strategies. This work aimed to determine whether widely available artificial intelligence (AI) platforms like Microsoft CopilotTM could facilitate healthcare economics analyses for ASP programs without dedicated healthcare economic supports.

AI (Microsoft CopilotTM) was prompted to develop a cytomegalovirus prophylaxis protocol for lung transplant recipients using only PubMed-indexed articles. CopilotTM was then prompted to produce probabilistic samples of simulated patients from aggregate statistics of a 165-patient cohort from Vanderbilt University Medical Center and to analyze cost-effectiveness across four distinct cytomegalovirus prophylaxis protocols, including its own.

Tertiary care academic medical center, including outpatient and inpatient environments.

Simulated patient data was developed via random, single-blind, probabilistic selection from pre-defined aggregate cohort statistics.

The AI-generated prophylaxis protocol was evidence-based without hallucination, but this conservative protocol relied on outdated evidence and was associated with significant increases in expected per-patient cost (mean +$4740, P < .01) compared to recent guideline-based and institutional protocols. AI independently identified and executed sensitivity analyses, which revealed that in this simplified model, letermovir use had a large impact on expected per-patient cost.

The AI-proposed protocol was less cost-effective, but data suggest that careful prompting can provide appropriate PubMed-indexed literature to support ASP protocol development. Additionally, CoPilotTM provided a thorough cost-effectiveness analysis comparing all potential and existing protocols. With appropriate oversight, AI and Microsoft CopilotTM can conduct healthcare economic analyses suitable for ASP strategic planning and implementation.

This study investigates the potential aerosol transmission of respiratory syncytial virus (RSV), a major cause of viral pneumonia and bronchiolitis in young children.

Two hundred samples were collected in a long-term environmental surveillance program from January 2022 until January 2023. Samples were collected in a pediatric emergency corridor. The analyses were performed using reverse transcription-quantitative polymerase chain reaction (RT-qPCR) targeting the RSV matrix gene. Information on the daily number of emergencies related with pediatric RSV infections was provided by the hospital.

Aerosol samples collected from a pediatric hospital corridor revealed detectable RSV RNA, particularly during peak infection seasons. RSV RNA was detected in 35 of 200 aerosol samples with a median concentration (interquartile range) of 1.8 (4.1) gc/m3. During the month of the peak season of RSV infections (November), RSV RNA was detected in 95% of the aerosol samples. Correlation analysis suggests a link between pediatric RSV cases and airborne RSV RNA concentration.

RSV RNA has been detected in aerosols in a healthcare setting, particularly during peak infection periods. This does not constitute evidence of transmission of the RSV via aerosols. However, the observed correlation with pediatric RSV cases suggests that further research on viral viability and infectivity from RSV detected in aerosols should be conducted. It also shows the potential of characterizing RSV RNA in aerosols for environmental surveillance purposes.

Some inpatient antibiotic use metrics have important limitations, including the absence of patient-level risk adjustment. To address this limitation, we previously developed two risk-standardized ratios (RSRs), one based on days of therapy (RSR-DOT) and the other on days of antibiotic spectrum coverage (RSR-DASC). Here, we evaluated the construct validity of these metrics by examining whether hospital performance was associated with the quality of antibiotic selection and duration in patients with either lower respiratory tract infections (LRIs) or urinary tract infections (UTIs).

Using Veterans Health Administration data from October 2020 through September 2021, we calculated RSR-DOT and RSR-DASC for 121 acute-care hospitals and selected 24 hospitals with relatively low or high values. At each hospital, a blinded infectious diseases physician reviewed 10 admissions each for LRI and UTI. Antibiotic selection at day 3 and total antibiotic duration were adjudicated and summarized as a 6-category ordinal composite outcome. Associations between RSR metrics and antibiotic-prescribing quality were evaluated using ordinal logistic regression; model-fit and bootstrap analyses compared RSR-DOT and RSR-DASC.

We reviewed 479 admissions (239 LRI and 240 UTI). In LRI, lower RSR-DASC tended to be associated with a higher quality of antibiotic use (coefficient, 2.906 [P = .053]), whereas RSR-DOT was not (1.418 [P = .374]). In UTI, neither metric was associated with the quality of antibiotic therapy. Model-fit and bootstrap analyses consistently favor RSR-DASC over RSR-DOT for LRI.

RSR-DASC may capture antibiotic-prescribing quality more closely than RSR-DOT for LRI, but not for UTI, potentially supporting the value of spectrum-informed benchmarking metrics.

Penicillin allergy reporting is common in U.S. healthcare systems, but true allergies and clinically significant reactions are inaccurately reported. Validated tools like PEN-FAST score allow for structured risk assessment; however, many clinicians remain unfamiliar with how to utilize the score to inform decision-making and prescribing. Integrating the PEN-FAST tool into the electronic health record (EHR) admission workflow may promote awareness and improve clinical utility.

We integrated the PEN-FAST tool into the admission navigator of our hospitals’ electronic health record to be completed by nursing staff. Over a seven-month period, completed PEN-FAST scores across our health system were analyzed to evaluate the overall opportunity for allergy assessment and delabeling. The study population consisted of patients with a documented penicillin class allergy and a completed PEN-FAST score. Patients with incomplete scores or responses marked as unknown for every item were excluded.

A total of 13,121 patients were included in the final evaluation. There were 10,309 (78.6%) patients with low-risk scores (PEN-FAST score of 0–2), indicating they were potential candidates for direct antibiotic challenges. The remaining 2,812 (21.4%) patients were categorized as high-risk (PEN-FAST score of 3+), who were likely not eligible for a challenge without prior skin testing.

A substantial opportunity remains to improve the assessment and documentation of penicillin allergies throughout the healthcare system. Utilizing the electronic health record to prompt frontline staff to use validated risk assessment tools may improve documentation of allergies and support better management of patients with penicillin allergy labels.

Fungal contamination of healthcare environments is increasingly recognized as a potential contributor to healthcare-associated infections, yet standardized environmental surveillance strategies remain poorly defined. Culture-based methods are widely used but have limitations including low sensitivity and prolonged turnaround time. Molecular approaches such as quantitative PCR (qPCR) may improve detection but have not been extensively evaluated in real-world hospital environments.

We conducted a longitudinal observational study across three inpatient units at a tertiary academic medical center from September 2023 through June 2024. Environmental samples were collected monthly from patient rooms and shared unit areas, including air, bathroom floors, HVAC components, patient bedrails, and linen storage areas. Samples were analyzed using direct-from-sample 18S qPCR and culture-based detection followed by 18S or ITS sequencing.

Among 742 samples collected, 474 (64%) were positive for fungi by qPCR compared with 213 (29%) by culture (P < .01). qPCR showed higher detection rates across most sample types, including air (35% vs 3%), bathroom floors (86% vs 42%), HVAC exhaust vents (72% vs 41%), and patient bedrails (78% vs 10%). Culture methods identified a broader diversity of fungi, including Talaromyces, Candida, and Penicillium, while qPCR detections were dominated by Malassezia.

Molecular and culture-based methods provide complementary insights into hospital fungal contamination. qPCR demonstrated greater sensitivity, while culture identified a broader range of viable fungi. Future surveillance strategies may benefit from leveraging qPCR sensitivity using targeted primers for clinically important fungi, reducing the need for broad sequencing and bioinformatic analysis.

Therapies provided in nontraditional settings or by individuals without knowledge of core principles of infection prevention and control (IPC) may place consumers at risk of adverse events. We summarize findings from our recent consultation experiences in the Division of Healthcare Quality Promotion (DHQP) at the Centers for Disease Control and Prevention (CDC) to help identify targets for prevention efforts and enhanced consumer protections.

We reviewed records of DHQP consultations with state, local, or territorial health officials conducted from 2016 to 2023 to identify investigations involving nontraditional therapies or settings in the United States. Information collected included type of procedure or therapy, type of setting, pathogens involved, number of affected individuals, and types of IPC issues identified.

Of 1500 consultations, we identified 30 (2%) involving nontraditional therapies or settings. Therapies were performed for esthetic reasons, general wellness, or as treatment for pain. Infectious pathogens were identified in the blood, joint fluid, and skin and soft tissue, and included viruses, bacteria, and fungi. Investigations revealed breaches in injection safety, medical device reprocessing, medication preparation and handling, and environmental disinfection. Over 7000 consumers received notification regarding potential exposure to an infectious pathogen.

This summary highlights the potential infectious risks to consumers posed by nontraditional therapies, which may have unproven benefits and poorly defined risks. With continued expansion of nontraditional therapies and settings, consumers will benefit from education, training, and oversight of those who provide therapies in these settings.

Expanded multiplex PCR gastrointestinal panels (GIPs) are routinely ordered to diagnose infectious diarrhea. However, recommendations for repeat GIP testing are limited, resulting in variable testing practices. We evaluated the diagnostic yield of repeat GIP testing within 14 days.

We conducted a retrospective cohort study of adults (age ≥ 18 years) tested with GIPs across 12 hospitals and outpatient centers (2019–2024). We analyzed the first diarrheal episode per patient, excluding cases with invalid/missing results and C. difficile results. Repeat testing was defined as a GIP completed within 14 days of an index GIP, excluding confirmatory testing using the same stool sample. The primary outcome was diagnostic yield (new pathogen detection), and the secondary outcome was pathogen persistence (same pathogen detection).

Among 16,502 patients, 507 (3.1%) underwent repeat GIP testing within 14 days (median interval: 6.3 days; IQR: 2.7–9.8). Only 4.6% [19/415] index-negative patients and 2.2% [2/92] index-positive patients detected new pathogens on repeat testing, with 51% [47/92] index-positive patients demonstrating persistence of at least one pathogen from their initial test. The number needed to test (NNT) to identify one new pathogen was 24 (95% CI: 16–39) tests overall, and 127 (95% CI: 50–455) tests to identify one new pathogen warranting antimicrobial treatment. Most repeat testing (86% [436/507]) was ordered by a different clinician.

Repeat GIP testing within 14 days rarely provided new diagnostic information, highlighting the limited utility of early repeat testing. Institutional policies discouraging repeat GIPs within 14 days may improve diagnostic stewardship.

Compare characteristics of community-associated (CA) versus healthcare-associated (HCA) extended-spectrum beta-lactamase-producing Enterobacterales (ESBL-E) urinary tract infection (UTI) cases and assess associations between ESBL-E UTI prevalence and census-tract social vulnerability.

Retrospective cohort study.

Monroe County, NY, January 2020–December 2023.

Population surveillance of first urinary ESBL-E isolate among county residents.

ESBL-E, defined as resistance to ≥1 third-generation cephalosporin, was identified from inpatient and outpatient urinary specimens. Demographic, clinical, and healthcare exposure data were abstracted; cases were classified as CA or HCA using standardized definitions. Addresses were geocoded to census tracts and linked to CDC Social Vulnerability Index (SVI) scores. Characteristics were compared using χ2 and Wilcoxon tests. Generalized linear models assessed associations between SVI themes and ESBL-E prevalence by epidemiologic classification.

Of 3,100 ESBL-E UTI cases, 1,621 (52.3%) were CA and 1,468 (47.4%) HCA. CA cases were younger (median 56 vs 71.5 years), more often female (89.5% vs 68.4%), and less medically complex than HCA cases. HCA prevalence increased with SVI score, while CA prevalence remained stable across SVI levels. The HCA association was mainly driven by SVI Theme 2 (household composition/disability; P = .0011).

CA and HCA ESBL-E UTIs exhibit distinct epidemiology. No association was observed between SVI and CA ESBL-E UTI prevalence. HCA prevalence increased with social vulnerability, largely driven by SVI Theme 2, potentially reflecting healthcare exposure and medical complexity. Factors driving CA ESBL-E differ, and ESBL-E reduction efforts should consider broader structural and community factors beyond those captured by the SVI.

Early identification of patients with carbapenemase-producing Enterobacterales (CPE) colonization is crucial for infection control; however, microbiological testing may delay detection and be costly. Machine learning may enhance predictive analytics for timely identification of at-risk patients.

Four machine learning models: Decision Tree, Random Forest, Gradient Boosting, and XGBoost, were used to predict CPE colonization within 48 hours of admission using microbiological and demographic data. Model performance was assessed through sensitivity, specificity, and area under the receiver operating characteristic curve (AUROC). Uniform Manifold Approximation and Projection (UMAP) evaluated topological separability of CPE-positive cases and CPE-negative controls.

From January 1, 2015 to December 31, 2024, 453,372 fecal specimens were submitted for CPE screening, with 194,917 (43.0%) collected within 48 hours of admission, comprising 3,328 CPE-positive cases (1.7%) and 191,589 CPE-negative controls. The Gradient Boosting classifier showed the best performance, achieving an AUROC of 0.598, sensitivity of 54.4%, and specificity of 59.1%. Demographic factors (age ≥ 75 and male sex), history of hospitalization, and known CPE colonization in the past year, and admission specialty (general medicine and general surgery) were consistently included in all models as top predictors. UMAP revealed significant overlap between CPE-positive and CPE-negative patients, indicating challenges in differentiating the risk profiles.

This study highlights the complexities of using machine learning to predict CPE colonization within 48 hours of admission. The low AUROC values suggest that the models may not effectively predict CPE colonization at the patient level, potentially due to inherent rarity of events and overlapping risk profiles.

Inadequate empiric therapy for sepsis has been correlated with heightened mortality; however, bloodstream infections do not uniformly manifest as severe presentations. Empiric administration of broad-spectrum antibiotics may result in adverse events and treatment failure. This study aimed to evaluate the impact of time to antibiotic optimization on outcomes of patients with bloodstream infections who received either inadequate or excessive empiric therapy.

This single-center retrospective study included patients aged ≥18 years with positive blood culture results and confirmed antimicrobial susceptibility between April 2018 and March 2024. The primary endpoint was a composite outcome of all-cause 30-day mortality, microbiological failure, and antibiotic modification due to insufficient efficacy. The secondary endpoints included adverse event–related discontinuation in the excessive group.

Of the 1943 patients screened, 405 met the inclusion criteria. In the inadequate group (n = 150), antimicrobial stewardship team intervention was associated with treatment success (OR, 0.38; 95% CI, 0.15–0.90; P = .028), whereas time to optimization was not associated with any of the outcomes. In the excessive group (n = 255), delayed optimization (≥64 h) was a risk factor for treatment failure (OR, 2.63; 95% CI, 1.14–6.07; P = .023). Later optimization (≥100 h) was also linked to higher antibiotic discontinuation due to adverse events.

Delayed optimization was independently associated with adverse outcomes in cohorts receiving excessive empiric therapy. These findings emphasize the importance of prompt de-escalation of broad-spectrum antibiotics to mitigate treatment failure and treatment discontinuation attributable to adverse events.