Citations

Abstract Objective: Electronic Health Record (EHR) data are critical for advancing translational research and AI technologies. The ENACT network offers access to structured EHR data across 57 CTSA hubs. However, substantial information is contained in clinical narratives, requiring natural language processing (NLP) for research. The ENACT NLP Working Group was formed to make NLP-derived clinical information accessible and queryable across the network. Methods: We established the ENACT NLP Working Group with 13 sites selected based on criteria including clinical notes access, IT infrastructure, NLP expertise, and institutional support. We divided sites into five focus groups targeting clinical tasks within disease contexts. Each focus group consisted of two development sites and two validation sites. We extended the ENACT ontology to standardize NLP-derived data and conducted multisite evaluations using the Open Health Natural Language Processing (OHNLP) Toolkit. Results: The working group achieved 100% site retention and deployed NLP infrastructure across all sites. We developed and validated NLP algorithms for rare disease phenotyping, social determinants of health, opioid use disorder, sleep phenotyping, and delirium phenotyping. Performance varied across sites (F1 scores 0.53–0.96), highlighting data heterogeneity impacts. We extended the ENACT common data model and ontology to incorporate NLP-derived data while maintaining Shared Health Research Informatics NEtwork (SHRINE) compatibility. Conclusion: This demonstrates feasibility of deploying NLP infrastructure across large, federated networks. The focus group approach proved more practical than general-purpose approaches. Key lessons include the challenge of data heterogeneity and importance of collaborative governance. This work also provides a foundation that other networks can build on to implement NLP capabilities for translational research.

Abstract With global aging, assessing functional status is vital for precision medicine. Electronic Health Records (EHRs), particularly unstructured data, hold abundant information on patient mobility. This study explores using Large Language Models (LLMs) to extract and standardize mobility status from unstructured EHR data (i.e., clinical notes). We annotated 600 clinical notes from three health care institutions located in southeastern Minnesota and west-central Wisconsin, focusing on expressions of mobility and associated impairment. Leveraging the open-source Llama 3 model, we tested various prompting strategies—including zero-shot, few-shot, and task decomposition—and evaluated their performance. Error analysis showed that while the model sometimes inferred impairments without explicit evidence, most errors were clinically reasonable, often reflecting borderline or ambiguous cases. While considering reasonable inference as correct, at the patient-level, Mobility Extraction achieves a micro-average accuracy of 0.952 with an F1-score of 0.962, and Impairment Classification produces a micro-average accuracy of 0.912 and an F1-score of 0.948. A local, deterministic setup improved trustworthiness by ensuring consistent outputs, safeguarding privacy, and demonstrating cross-institution generalizability. These findings highlight the feasibility of LLM-based solutions for extracting mobility functional status from unstructured EHR data, supporting both clinical applications and research.

Abstract Epcoritamab and glofitamab are CD20-directed bispecific antibodies (BsAbs) approved in the United States for relapsed or refractory (R/R) diffuse large B-cell lymphoma (DLBCL). Limited data exist for patients treated outside of trials. Patients with R/R DLBCL receiving commercial epcoritamab or glofitamab between 1 January 2023 and 15 October 2024 were collected from 21 United States institutions. Among 245 patients, 156 received epcoritamab and 89 received glofitamab, 113 were refractory to front-line therapy, 40 had MYC and BCL2 and/or BCL6 rearrangements, 147 received prior chimeric antigen receptor T-cell therapy, and 174 patients would have been ineligible for registrational trials. The overall response rate (ORR) for epcoritamab and glofitamab was 51% (23% complete response, [CR]) and 53% (30% CR), respectively. Median progression-free survival (PFS) was 2.6 months (95% confidence interval [CI], 2.0-3.8 months), and median overall survival (OS) was 7.8 months (95% CI, 6.2-11.0 months). The 6-month PFS was 36% (95% CI, 30-44) and the 6-month OS was 60% (95% CI, 54-67). Both trial ineligibility and undetectable CD20 pre-BsAbs portended shorter PFS and OS. Of 17 individuals with paired biopsies, 15 (88.2%) lost CD20 expression after BsAbs with a median time to progression of 3.7 months. This analysis including patients with R/R DLBCL shows the ORR to CD3/CD20 BsAbs was comparable to pivotal trials, although PFS and OS were lower. Baseline undetectable levels of CD20 were associated with poor outcomes. These results demonstrate the activity of BsAbs in R/R DLBCL, and underscore the importance of target antigen expression.

ABSTRACTIntroductionDurable left ventricular assist devices (LVADs) are susceptible to infection. The presence of a transvenous cardiac implantable electronic device (CIED) may further elevate this risk, making it necessary to remove the CIED system. This report presents the procedural characteristics and outcomes of CIED system extraction in patients with LVADs.MethodsThis is a single‐center retrospective study that included patients with durable LVADs who underwent transvenous CIED extraction at our institution from 2017 to 2023.ResultsTwenty patients with LVAD (HeartMate II = 14; HeartMate III = 6) underwent CIED removal during the study period. The median age was 52.5 (IQR = 42.5–56.5) years, and 7 patients had ischemic cardiomyopathy. Ninety percent (n = 18) of the extractions were due to infection, while 10% (n = 2) were for lead management. Complete removal of all leads was accomplished in 19 patients (95%). Overall, 40 out of 45 leads were removed during 20 procedures. Simple traction was used in 8 patients and specialized extraction tools were used in 12 (laser sheaths = 10; mechanical rotational sheaths = 7; both = 5). No leads were removed in one patient (5 leads; 2 abandoned leads with 42.5 years intravascular dwell time). There were no major procedure‐related complications. Two minor procedure‐related complications included a femoral arterial pseudoaneurysm and one pocket revision due to persistent surgical site bleeding. During a median follow‐up of 17 (IQR = 8.0–22.3) months, 8 patients underwent pump exchange due to persistent infection, and 15 remained on chronic suppressive antibiotics. Nine patients died during the follow‐up.ConclusionThe extraction of CIED systems in patients with LVAD is safe and feasible, exhibiting low procedure‐related morbidity and mortality. However, the long‐term prognosis for these patients remains poor despite the removal of CIEDs.

ABSTRACTPortopulmonary hypertension (PoPH), a type of pulmonary arterial hypertension (PAH) in patients with liver disease, is associated with high morbidity and mortality. The relationship between cardiopulmonary hemodynamics, PAH therapy, and survival in PoPH remains unclear. We performed a retrospective cohort study of PoPH patients from the international pulmonary hypertension (PH) meta‐registry, PVRI GoDeep. PAH was defined by a mean pulmonary arterial pressure > 20 mmHg, pulmonary arterial wedge pressure ≤ 15 mmHg, and a pulmonary vascular resistance (PVR) > 2 Wood Units. PoPH diagnoses were assigned by each center's PH specialist based on international guidelines at the time of enrollment. 246 incident PoPH patients met eligibility criteria and were included in the analysis, equally split between males (51%) and females (49%), with a median age of 54 years. When compared to both patients with IPAH and those with other subtypes of PAH (not classified as PoPH or IPAH), those with PoPH had significantly lower 5‐year survival rates (46% vs. 68% vs. 65%, log‐rank p < 0.001). Amongst the PoPH patients, however, there was no significant difference in 5‐year survival when dichotomized by disease severity, either by a PVR of 5 Wood Units or a CI of 2.5 L/min/m2. Treatment of the PoPH patients with PAH‐targeted therapies was associated with significantly higher 5‐year survival rates compared to those not receiving such treatments, as shown by Kaplan–Meier analysis. This survival benefit was observed for PDE5i (50% vs. 34%, log‐rank p = 0.029), ERA (58% vs. 34%, log‐rank p < 0.001), and the combination of PDE5i and/or ERA (51% vs. 22%, log‐rank p < 0.001), as well as any PAH‐targeting treatment (50% vs. 26%, log‐rank p = 0.007). Corresponding survival advantage was noted when including only PoPH patients with MELD Score ≥ 13. PoPH is a disease with significantly worse long‐term survival than other PAH subtypes, but targeted PAH therapy is associated with a robust survival benefit. Survival did not differ across high‐risk PVR and cardiac index thresholds, suggesting the factors that influence prognosis and survival in PoPH may be unique as compared to other PAH subtypes, and warrant further investigation.

In 2023, the VAS international working group on Buerger’s Disease (BD) recommended two diagnostic criteria based on a prior Delphi study: “definitive” and “suspected”. The “definitive” criteria are history of smoking, typical angiography, and typical histopathological features. All three features are mandatory to confirm a “definitive” diagnosis of BD. The conundrum is—what features should be considered typical of BD angiography? According to this review, segmental occlusion of infrapopliteal arteries, corkscrew collaterals that appear to continue the occluded arterial segment (Martorell’s sign) or bypass the segmental occlusion, absence of atherosclerotic plaque or aneurysm could differentiate BD from ASO. Hence, for “typical” BD angiography, these manifestations should certainly be considered. However, data for differentiating angiography patterns of BD from the small- and medium-sized vasculitis including Behcet’s disease, scleroderma, hepatitis associated vasculitis, and anti-phospholipid syndrome are limited. Further studies for investigating the angiography pattern in BD patients in early and late presentation of BD, particularly in the patients with long-term follow up, are highly recommended.

INTRODUCTION A retrospective cohort study on patients aged &lt; 18 years included in the Society of Critical Care Medicine: Viral Infection and Respiratory Illness Universal Study registry from March 2020 to April 2024 with an objective of calculating the prevalence of sepsis as defined by the Phoenix Sepsis Score (PSS) and to validate the PSS with respect to outcomes in children with COVID-19. METHODS Linear mixed-effects regression was used to examine the relationship between the PSS and hospital length of stay after controlling for confounding factors. The performance of the PSS was assessed using the receiver operating characteristic (ROC) and the precision-recall curve (PRC). Cross-validation was performed using leave-one-out cross-validation. RESULTS Out of 1731 patients (58 hospitals), 326 (18.8%) met criteria for sepsis and 167 (9.7%) for septic shock. The overall mortality was 1.4% (25/1731), with significant differences between nonseptic (10/1405, 0.7%) and both sepsis (15/326, 4.6%) and septic shock (9/167, 5.4%) groups. After adjusting for confounders, the septic group was associated with a longer hospital length of stay than the nonseptic group. One unit increase in the numeric PSS led to a 70% increase in risk of mortality (odds ratio 1.70; P &lt; .001). The area under the ROC curve was 0.80 and the area under the PRC curve was 0.13. The threshold of ≥ 2 for detection of mortality had a sensitivity of 0.63, specificity of 0.82, and positive predictive value of 0.05. CONCLUSION Phoenix Sepsis Criteria retain its validity in identifying sepsis in children with COVID-19 and can be used in further epidemiological studies in this population

Importance The role and safety of transcranial electrical stimulation (tES) for treating depressive disorders remain under evaluation. Objective To evaluate tES treatment in patients with major depressive disorder (MDD) and comorbid depressive conditions. Data Sources A search of MEDLINE, Embase, Cochrane, APA PsycINFO, and Scopus databases was conducted from inception to September 17, 2024. Study Selection Randomized clinical trials (RCTs) of adults with MDD, depression with psychiatric comorbidities (DPC), or depression with medical comorbidities (DMC), treated with transcranial direct current stimulation (tDCS), transcranial alternating current stimulation (tACS), or transcranial random noise stimulation (tRNS), compared with sham or other treatments were included. Data Extraction and Synthesis Independent reviewers extracted data in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2020 guidelines, with random-effects meta-analysis used for pooling. Main Outcomes and Measures Primary outcomes were depression severity, response and remission rates, and adverse events. Standardized mean differences (SMDs) were reported for continuous outcomes, and odds ratios (ORs) were reported for categorical outcomes. Quality of evidence (QOE) was assessed using the Grading of Recommendations Assessment, Development, and Evaluation criteria. Results The meta-analysis included 5522 participants from 114 study groups from 88 RCTs (3198 female [58.9%]; mean [range] age, 43.1 [19.4-76.9] years). Most studies (104 study groups from 79 RCTs [91.2%]) evaluated tDCS, while 7 study groups from 6 RCTs (6.1%) evaluated tACS, and 3 study groups from 3 RCTs (2.7%) evaluated tRNS. tES was associated with reduced depressive symptoms (SMD = −0.59; 95% CI, −0.83 to −0.35; low QOE) and improvement in DMC (SMD = −1.05; 95% CI, −1.67 to −0.43; low QOE) and DPC (SMD = −0.78; 95% CI, −1.27 to −0.29; low QOE) compared with MDD (SMD = −0.22; 95% CI, −0.44 to 0.01; low QOE). tDCS was associated with reduced depression in DMC (SMD = −1.05; 95% CI, −1.70 to −0.40; very low QOE) and DPC (SMD = −0.88; 95% CI, −1.40 to −0.36; low QOE) but not MDD. tACS was associated with improved MDD symptoms (SMD = −0.58; 95% CI, −0.96 to −0.20; high QOE) and response rates (OR, 2.07; 95% CI, 1.34 to 3.19; high QOE). Combined tDCS and medication was associated with reduced symptoms (SMD = −0.51; 95% CI, −0.90 to −0.13; moderate QOE) and increased response (OR, 2.25; 95% CI, 1.08 to 4.65; high QOE) in MDD. tDCS combined with psychotherapy was not associated with improvement. Subgroup analysis showed that anodal left dorsolateral prefrontal cortex DCS was associated with improved outcomes. Mild to moderate adverse events were more frequent in tES groups. Conclusions and Relevance In this systematic review and meta-analysis, tDCS was associated with improvement in depression among patients with DMC and DPC (with smaller benefits in MDD), tACS was associated with improved MDD outcomes (while tRNS had insufficient evidence) in smaller samples, and combined tDCS and medication was associated with improvement in depression. These findings suggest that tES is well-tolerated overall, with only mild to moderate adverse events, and that future research should optimize stimulation parameters and individualize tES interventions.

Background Family participation in the delivery of nonpharmacological measures has shown in past studies to prevent 17-75% of incident delirium. A scalable and sustainable method to partner with family (i.e., care partners) and support their delivery of nonpharmacologic measures is needed. Objectives To test the feasibility, acceptability, and usability of iPREPARED in a pilot randomized controlled clinical trial (RCT). iPREPARED is a mobile health technology co-designed with patients and care partners containing resources on nonpharmacological interventions to prevent delirium. Methods Hospitalized adults ( > 65 years) with > 1 delirium risk factor and a care partner (family or friend > 18yo) willing to participate were enrolled. The primary outcomes were feasibility, acceptability, and usability. Descriptive statistics are reported. Results In total, 23 dyads completed the study (median age 72 years (IQR 64, 79), 60% male, and 100% Caucasian). Dyads reported that iPREPARED was acceptable, with moderate usability. Recruitment in the emergency department was feasible whereas recruitment once hospitalized and admitted was more difficult. Conclusion The study design was feasible. iPREPARED was acceptable and usable as an intervention.