To support Antimicrobial Stewardship programs (ASPs) as well as the clinical care of patients with infections, we disseminated and implemented a Videoconference Antimicrobial Stewardship Team (VAST) to connect multidisciplinary teams from rural Veterans Affairs (VA) medical centers with geographically distant ID experts. Here, we describe the clinical syndromes discussed and the response to recommendations made during VAST sessions.

Between September 2021 to February 2024, eight ID consultants established VASTs with ten rural VAMCs, holding regularly scheduled videoconference sessions to discuss clinical cases and provide recommendations. Data were collected on patient demographics, clinical syndromes, and recommendations. Acceptance of recommendations within one week of each session was assessed via chart review. Six months after the intervention began, we conducted semi-structured interviews to assess participants’ perceptions of VASTs.

VASTs reviewed 626 cases involving 527 unique patients. Among 763 clinical syndromes discussed, the most common were infections of the respiratory (29%) or urinary tract (21%). Overall, VASTs made 973 recommendations, of which 71% were accepted. Of 570 recommendations related to antibiotics, 459 (80%) were accepted. Among 403 other recommendations, 235 (58%) were accepted. Interviews with participants indicated the importance of building trust and strong interpersonal relationships.

VASTs effectively supported Antimicrobial Stewardship in rural VA medical centers (VAMCs) without local ID expertise. High acceptance rates, particularly for antibiotic-related recommendations, suggest that telehealth-enabled provider-to-provider models enhance stewardship efforts.

Describe the workload associated with using telehealth to support Antimicrobial Stewardship efforts at Veterans Affairs Medical Centers (VAMCs) without local infectious diseases (ID) expertise.

A mixed-methods process assessment to evaluate workload and workflow associated with Videoconference Antimicrobial Stewardship Teams (VASTs).

Rural VAMC champions paired with ID consultants at geographically distant VAMCs to form VASTs.

Total workload estimates were based on time that champions and ID consultants allocated to VAST activities. Clinical Procedural Terminology (CPT) codes were used to estimate the workloads for clinical encounters. Role-based process maps were developed to understand variation in implementation by VAMC.

The average workload that champions and ID consultants allocated to VAST activities was 6.7% (range 1.0%–20.0%) and 8.4% (range 2.0%–12.5%) full-time equivalents (FTEs), respectively. Clinical encounters completed by ID consultants contributed an average of 1.4% (range < 0.01%–2.5%) FTEs to the workload. The average proportion of FTEs required to sustain VASTs was 13.0% (range 3.0%–31.6%). Process maps showed four phases common to each VAST’s workflow: case identification, meeting preparation, team meeting, and documentation. The tasks associated with each phase varied between VASTs. Champions carried out most tasks related to case finding and meeting preparation; the ID consultants completed most documentation tasks.

The distribution of tasks within and among the VASTs indicated opportunities to improve workflow efficiency. Investing <12.5% of the FTE allocated to VA Antimicrobial Stewardship programs to support the time of an ID consultant from another VAMC can help rural VAMCs achieve staffing sustainability.

The implementation of the World Health Organization’s multimodal strategy is crucial for strengthening infection prevention and control (IPC) programs. This study aimed to evaluate and compare the perceptions of infection prevention and control professionals (IPCPs) and healthcare workers (HCWs) regarding the effectiveness of the multimodal strategy.

In this study, an online survey was designed that included five MM components: system change, education and training, monitoring and feedback, reminders and communication, and cultural change. The survey was completed by IPCPs and HCWs within the institutions.

A total of 102 IPCPs and 166 HCWs from 96 institutions participated in the survey. The mean career duration was 19.8 ± 9.5 years. An Infection Control Committee (ICC) was present in 98% of participants’ institutions, and 88% held meetings every three months. Only 10 (10.4%) institutions met the ratio of 1IPC nurse per 150 beds. Effective communication techniques training was provided annually to 55.9% of IPCPs and 59.6% of HCWs.

Eighty-one (79.4%) IPCPs reported using effective communication when explaining the five indications of hand hygiene, but only 77 (46.4%) HCWs agreed (P < .001). Ninety-one (89.2%) IPCPs confirmed that effective communication techniques were used in bundle training. But, only 25 (15.0%) HCWs agreed, while 112 (67.5%) HCWs reported being undecided (P < .001). Sixty-one (64.7%) IPCPs and 130 (78.3%) HCWs believed that their institutions effectively motivated HCWs towards IPC measures compliance. However, 44 (43.1%) IPCPs and 29 (17.5%) HCWs were undecided (P = .002).

The survey reveals significant inconsistencies among IPCPs and HCWs regarding communication effectiveness, and perceptions of compliance with infection-control measures. These findings highlight the need for more inclusive and transparent communication strategies, regular interdisciplinary training, and stronger organizational support to improve IPC practices.

Antimicrobial Stewardship in the intensive care unit setting is a difficult task due to diagnostic uncertainty and perceived high-risk of poor outcomes in case of delayed or inappropriate treatment. Although novel diagnostics and other strategies have been proposed to improve antimicrobial use, their clinical effectiveness in real-world settings has been suboptimal.

We designed a critical interpretative synthesis of the literature, which allows the combination of quantitative and qualitative studies to revise and critique concepts used in Antimicrobial Stewardship efforts in the ICU setting. We searched the literature in duplicate with a sensitive strategy to identify main concepts, and we developed a main theme and conceptual framework after identifying the main concepts and strategies.

After screening 41,192 titles and abstracts and reviewing 1,335 full-text manuscripts, we selected 29 main manuscripts for this synthesis. We identified that classical concepts, such as the use of broad-spectrum antibiotics followed by de-escalation and the use of biomarkers of infection and novel diagnostics, although with face validity and supported by efficacy studies, carry a high risk of being ineffective in real-world settings. We argue that this discrepancy is due to cognitive biases in antimicrobial decision-making in the ICU setting, including risk-aversion behavior, diagnostic momentum, premature closure, therapeutic momentum, hyperbolic discounting, commission bias, and anchoring bias, among others, which drive intensivists towards overdiagnosis and overtreatment of infection.

Incorporation of the cognitive theory of decision-making in future stewardship efforts is necessary in the ICU setting along with traditional stewardship interventions.

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.

Racial and ethnic disparities in healthcare-associated infections (HAIs) could have significant implications for hospital-based incentive programs. We sought to quantify racial and ethnic disparities in HAIs across inpatients in the United States.

Systematic Review and Meta-Analysis.

Medline, Embase, and Scopus were searched 2008–2022 for English language studies describing reportable HAIs among inpatients and references to race or ethnicity. Studies were excluded if they used an aggregate outcome of “infection,” described a non-acute care setting, or measured a respiratory virus. Meta-analysis was performed using random-effects model on pooled continuous outcomes with 3 or more studies.

39 studies met criteria for inclusion; 23 evaluated surgical site infections (SSI), 6 evaluated hospital-onset Clostridioides difficile infection (HO-CDI), 5 evaluated central line-associated bloodstream infections (CLABSI), and 8 examined other HAIs. There was a high degree of heterogeneity across studies. Meta-analysis was performed for 10 distinct HAI/risk factor combinations. Race and ethnicity were not associated with SSIs (Black vs White OR 1.19, 95% CI 0.95–1.48; Hispanic vs White OR 1.01, 95% CI 0.78–1.31). Hispanic patients had lower risk of HO-CDI relative to White patients (OR 0.80, 95% CI 0.64–0.98). Black (OR 1.36, 95% CI 1.07–1.70) or Hispanic (OR 1.16, 95% CI 1.073–1.249) patients had increased risk of CLABSI compared with White patients.

Racial and ethnic disparities were observed in rates of HAIs, specifically for CLABSI and HO-CDI. However, conclusions are limited by the substantial heterogeneity present. Further research characterizing social determinants of health driving these disparities is needed.

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.

Develop and implement a framework for reviewing hospital-onset bacteremia and fungemia (HOB) events to identify quality improvement opportunities.

Prospective, observational study.

Six hospitals (3 academic and 3 community) associated with two health systems.

Clinical team members involved in the care of patients with eligible HOB events.

The Joint Commission Patient Safety Event Taxonomy was adapted to develop a framework for HOB, eliciting potential contributing factors in four categories: general infection prevention, infection-site-specific, system, and human factors. From 9/2022 to 4/2024, HOB events identified from hospital blood culture databases were selected using convenience sampling to a target enrollment of 75 events, limiting inclusion of events associated with neutropenia or central line infections. Multidisciplinary case reviews were performed, and the framework was refined using feedback from the first 17 reviews. We summarized HOB sources, contributing factors, and improvement opportunities using descriptive analyses.

Seventy-four HOB case reviews were analyzed, evenly distributed between academic and community hospitals, and 42% occurred in an ICU setting. A median of 3 care team members (IQR 2–4) participated in each review. At least one infection-site-specific factor was identified in 48% of case reviews, representing 63% of HOB events with a known source of infection. 50% of cases identified a general infection prevention factor, 35% a system factor, and 34% a human factor. 57% of the cases had an actionable finding.

HOB events have diverse causes and elements of preventability, and multidisciplinary review can generate actionable findings beyond the current guideline-defined measures.

Surgical site infection (SSI) surveillance can be time consuming and resource intensive. This study investigates the potential of generative artificial intelligence (GenAI) to augment the detection and classification of SSIs.

A case control study of patients with SSI following spine surgery at one US hospital. SSIs were classified into superficial, deep, and organ space. All SSIs were confirmed by infection prevention (IP) experts as they occurred from October, 2023 to September, 2025 and matched 1:1 by year to surgeries deemed non-SSI. A secure GenAI was used to determine if patients had an SSI based on standardized prompts and clinical data. IP nurses used GenAI output to review cases with the ability to ask GenAI questions within the data provided or independently open the medical record. We compared GenAI determinations to initial IP nurses’ determinations.

A total of 555 patients had spine surgeries. All 16 SSIs were matched by year to 16 non-SSI. All SSIs were correctly identified by GenAI (sensitivity 100%, 16/16) and only 1 non-SSI was incorrectly identified as SSI (specificity 93.7%, 15/16). Although GenAI accurately identified all SSI cases, it was discordant with original review at classifying the level of infection in 37.5% (6/16) of cases. Upon final IP physician review, GenAI was correct in 66.7% (4/6) of discordant cases (often determining “organ space infections” rather than “deep”). Median time to complete GenAI assisted SSI reviews was 9 minutes (IQR 7–21).

GenAI is a promising tool to assist in SSI surveillance following spinal surgery that could improve efficiency.

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.

To evaluate factors associated with positive LTBI screening among HCWs and predictors of treatment initiation and completion across hospital sites in Ireland.

Multicentre retrospective cohort study.

Five hospital sites in Ireland.

N = 755 healthcare workers (HCWs).

Evaluation of latent tuberculosis infection (LTBI) by interferon gamma release assay in HCWs from high-incidence countries during 2023, identified via occupational health records. IGRA positivity rates, linkage to treatment and treatment outcomes were recorded. Demographic and occupational factors associated with these outcomes were investigated.

There were n = 755 HCWs from high-incidence TB countries identified via occupational health records eligible for LTBI screening. 719 underwent IGRA testing, of whom 93 (13%) were positive. Age > 50 was associated with IGRA positivity (OR 5.71; 95% CI 1.79–18.17; P = .003). In addition to these n = 93 HCWs, two additional sites provided treatment outcomes for n = 164 HCWs, and a further n = 58 IGRA-positive HCWs were referred to Site 1. Among these 313 IGRA-positive HCWs, 50% initiated therapy, with substantial variation across sites (27%–88%). Multivariable analysis showed study site, but not demographic factors, predicted treatment initiation (P < .001). Common reasons for non-initiation included treatment refusal and non-attendance. Treatment completion was high (82%) and was not associated with study site.

LTBI prevalence among HCWs in Ireland was lower than international estimates. While treatment initiation was low, completion was high. Treatment initiation varied by site, driven by institutional rather than individual factors. A standardised national programmatic approach is needed for HCWs within the LTBI cascade of care.

To define the incidence of donor-derived infection (DDI) in recipients of solid organ transplant (SOT) from donors with positive blood cultures and to assess the impact of shorter versus longer duration of targeted preemptive antibiotic therapy (PAT).

Retrospective, single-center, cohort study.

Mayo Clinic Arizona.

Recipients transplanted between 1/1/2019 and 7/1/2024 who received an organ from a donor with positive blood cultures.

The primary outcome was incidence of DDI. Secondary outcomes included duration of PAT and incidence of donor blood culture contamination.

Among 199 SOT recipients from 167 unique donors with positive blood cultures, two recipients developed confirmed DDI within 30 days of SOT. Both cases were gram negative bacillary bacteremia not treated in donors and occurred immediately posttransplant prior to adequate recipient PAT. Six-month graft survival and recipient survival were 96.5% and 97.5% respectively. 139 recipients (69.8%) received PAT for a median duration of 7 days. There was no difference in rate of infections between recipients provided with ≤7 days versus 8–14 days of PAT for donor blood cultures; however, recipients who received 8–14 days had more Clostridioides difficile infections (CDIs) within 60 days of SOT (7.7% vs 1.5% ≤ 7 days, P = .040) and were more often discharged on intravenous antibiotics (32.3% vs 11.3%, P < .001).

We observed a low rate of DDI following receipt of organs from donors with positive blood cultures. DDI occurred in cases without adequate donor/recipient treatment. Longer durations of targeted PAT resulted in more CDI and intravenous antibiotics on discharge.

To estimate the incidence of healthcare-associated infections (HAIs) in Italian long-term care facilities (LTCFs) and to evaluate whether an artificial intelligence (AI) approach, through unsupervised machine learning (ML), could stratify residents into clinically distinct groups with differing susceptibility to HAIs.

Prospective cohort study with 12-month follow-up.

24 LTCFs in Italy, participating in the European Centre for Disease Prevention and Control 12-month longitudinal study on HAIs in LTCFs, 2022–2023.

395 residents enrolled across the participating LTCFs.

Incidence measures of HAIs (rate and ratio) were estimated, using generalized estimating equations. A hierarchical cluster analysis based on residents’ clinical and demographic characteristics was implemented as an unsupervised ML approach.

Overall, 75 HAIs per 100 residents (95% CI, 70.3–78.3) and 0.23 HAIs per 1,000 resident-days (95% CI, 0.11–0.76) were estimated. Respiratory tract infections (29.5%, 95% CI 24.2–31.1), COVID-19 (26.3%, 95% CI 22.1–28.4), and urinary tract infections (15%, 95% CI 11.0–35.4) were the most frequent. Clustering identified two reproducible resident groups: Group 1 (39%), more independent and cognitively preserved, with fewer comorbidities and lower infection incidence; and Group 2 (61%), more dependent and clinically complex, with higher incidence of HAIs. Cluster stability was high (mean ARI = 0.83).

This study confirms the high burden of HAIs in Italian LTCFs and provides exploratory evidence that AI-based clustering can identify reproducible HAI susceptibility profiles in a setting where such approaches have been scarcely applied.