Explore the vast range of titles available.

Postoperative patients are at risk of life-threatening complications such as hemodynamic decompensation or arrhythmia. Automated detection of patients with such risks via a real-time clinical decision support system may provide opportunities for early and timely interventions that can significantly improve patient outcomes. We utilize multimodal features derived from digital signal processing techniques and tensor formation, as well as the electronic health record (EHR), to create machine learning models that predict the occurrence of several life-threatening complications up to 4 hours prior to the event. In order to ensure that our models are generalizable across different surgical cohorts, we trained the models on a cardiac surgery cohort and tested them on vascular and non-cardiac acute surgery cohorts. The best performing models achieved an area under the receiver operating characteristic curve (AUROC) of 0.94 on training and 0.94 and 0.82, respectively, on testing for the 0.5-hour interval. The AUROCs only slightly dropped to 0.93, 0.92, and 0.77, respectively, for the 4-hour interval. This study serves as a proof-of-concept that EHR data and physiologic waveform data can be combined to enable the early detection of postoperative deterioration events.

To quantify preoperative heart failure (HF) diagnostic agreement and identify characteristics of patients in whom physicians agreed versus disagreed about the diagnosis. Observational cohort study. Patients undergoing major non-cardiac surgery at an academic center between 2015 and 2019. 40,659 patients undergoing major non-cardiac surgery, among which a stratified subsample of 1018 patients with and without documented HF was reviewed. Via a panel of physicians frequently managing patients with HF (cardiologists, cardiac anesthesiologists, intensivists), detailed chart reviews were performed (two per patient; median review time 32 min per reviewer per patient) to render adjudicated HF diagnoses. Adjudicated diagnostic agreement measures (percent agreement, Krippendorf's alpha) and univariate comparisons (standardized differences) between patients in whom physicians agreed versus disagreed about the preoperative HF diagnosis. Among patients with documented HF, physicians agreed about the diagnosis in 80.0% of cases (consensus positive), disagreed in 13.8% (disagreement), and refuted the diagnosis in 6.3% (consensus negative). Conversely, among patients without documented HF, physicians agreed about the diagnosis in 88.0% (consensus negative), disagreed in 8.4% (disagreement), and refuted the diagnosis in 3.6% (consensus positive). The estimated agreement for the 40,659 cases was 91.1% (95% CI 88.3%–93.9%); Krippendorff's alpha was 0.77 (0.75–0.80). Compared to patients in whom physicians agreed about a HF diagnosis, patients in whom physicians disagreed exhibited fewer guideline-defined HF diagnostic criteria. Physicians usually agree about HF diagnoses adjudicated via chart review, although disagreement is not uncommon and may be partly explained by heterogeneous clinical presentations. Our findings inform preoperative screening processes by identifying patients whose characteristics contribute to physician disagreement via chart review. Clinical Trial Number / Registry URL: Not applicable. [Display omitted] •A physician panel performed reviews of patients undergoing non-cardiac surgery.•The panel agreed about the diagnosis of heart failure (HF) in 91% of cases.•Disagreement occurred among patients with fewer guideline-defined HF criteria.•Disagreement may be partly explained by heterogenous clinical presentations of HF.•Preoperative HF screening processes may consider characteristics in this study.

IntroductionCardiac inotrope medications administered to cardiac surgical patients carry steep risk–benefit trade-offs, yet wide inter-institutional variation exists in inotrope practices. Despite known wide variation in use of any inotrope for cardiac surgery, limited multicentre data exist regarding determinants of inotrope selection and time course for use. Additionally, the reasons that underpin how clinicians decide on inotrope usage and the factors that influence inotrope practice change are not well understood.Methods and analysisThis is an investigator-initiated, multicentre mixed methods study. Quantitative data will include electronic health records from an observational cohort of adult cardiac procedures within the Multicenter Perioperative Outcomes Group (MPOG) database, comprising cardiac surgical procedures from over 30 US academic and community hospitals. Additional quantitative data will be collected via surveys of clinicians involved in inotrope decision-making, contacted through an existing multicentre research and quality improvement infrastructure with engaged clinician representatives participating across MPOG hospitals. Qualitative data will be collected from open-ended questions within surveys, as well as semi-structured interviews with surveyed clinicians, sampled across approximately six institutions selected for diversity of settings and inotrope practices. An explanatory sequential mixed methods design will merge quantitative and qualitative data to develop meta-inferences explaining inotrope practices, as guided by an existing framework for characterising clinical practice variation and levers for practice change.Ethics and disseminationThe study is approved by the institutional review board at the University of Michigan Medical School (HUM00245353). Findings will be disseminated through peer-reviewed journals, conference proceedings and quality improvement forums. The study began in February 2025 and will continue until 2028.

Challenges with model generalizability and data privacy have led to a shift in health artificial intelligence (AI) models being trained locally within individual health systems rather than relying on multicenter data. Localization carries the promise of capturing local practice patterns and patient demographics, presumably resulting in better models. Our study empirically tests this hypothesis in a national research network by comparing locally trained models predicting acute kidney injury (AKI) after cardiac surgery with two multicenter modeling approaches, pooling and a novel federated model stacking method. Trained on 43,926 cases across 23 hospitals, the study finds that multicenter models outperform single-center approaches, with higher area under the receiver operating characteristic curves (AUCs) for all AKI severity levels in both temporal and external validation sets. Hospitals with smaller case volumes benefit the most from multicenter approaches, showing the greatest AUC increase over locally trained models.

Background Prior studies support a genetic basis for postoperative acute kidney injury (AKI). We conducted a genome-wide association study (GWAS), assessed the clinical utility of a polygenic risk score (PRS), and estimated the heritable component of AKI in patients who underwent noncardiac surgery. Methods We performed a retrospective large-scale genome-wide association study followed by a meta-analysis of patients who underwent noncardiac surgery at the Vanderbilt University Medical Center (“Vanderbilt” cohort) or Michigan Medicine, the academic medical center of the University of Michigan (“Michigan” cohort). In the Vanderbilt cohort, the relationship between polygenic risk score for estimated glomerular filtration rate and postoperative AKI was also tested to explore the predictive power of aggregating multiple common genetic variants associated with AKI risk. Similarly, in the Vanderbilt cohort genome-wide complex trait analysis was used to estimate the heritable component of AKI due to common genetic variants. Results The study population included 8248 adults in the Vanderbilt cohort (mean [SD] 58.05 [15.23] years, 50.2% men) and 5998 adults in Michigan cohort (56.24 [14.76] years, 49% men). Incident postoperative AKI events occurred in 959 patients (11.6%) and in 277 patients (4.6%), respectively. No loci met genome-wide significance in the GWAS and meta-analysis. PRS for estimated glomerular filtration rate explained a very small percentage of variance in rates of postoperative AKI and was not significantly associated with AKI (odds ratio 1.050 per 1 SD increase in polygenic risk score [95% CI, 0.971–1.134]). The estimated heritability among common variants for AKI was 4.5% (SE = 4.5%) suggesting low heritability. Conclusion The findings of this study indicate that common genetic variation minimally contributes to postoperative AKI after noncardiac surgery, and likely has little clinical utility for identifying high-risk patients.

Background There are few data to guide the intraoperative management of patients with reduced left ventricular ejection fraction (LVEF). This study aimed to describe how patients with reduced LVEF are managed differently and to identify and treatments had a different risk profile in this population. Methods We performed a retrospective cohort study of adult patients who underwent general anesthesia for non-cardiac surgery. The effect of anesthesia medications and fluid balance was compared between those with and without a reduced preoperative LVEF. The primary outcome was a composite of acute kidney injury, myocardial injury, pulmonary complications, and 30-day mortality. Multivariable logistic regression was used to adjust for confounders. Treatments that affected patients with reduced LVEF differently were defined as those associated with the primary outcome that also had a significant interaction with LVEF. Results A total of 9420 patients were included. Patients with reduced LVEF tended to have a less positive fluid balance. Etomidate, calcium, and phenylephrine were use more frequently, while propofol and remifentanil were used less frequently. Remifentanil affected patients with reduced LVEF differently than those without (interaction term OR 2.71, 95% CI 1.30–5.68, p  = 0.008). While the use of remifentanil was associated with fewer complications in patients with normal systolic function (OR 0.54, 95% CI 0.42–0.68, p  < 0.001), it was associated with an increase in complications in patients with reduced LVEF (OR = 3.13, 95% CI 3.06–5.98, p  = 0.026). Conclusions Patients with a reduced preoperative LVEF are treated differently than those with a normal LVEF when undergoing non-cardiac surgery. An association was found between the use of remifentanil and an increase in postoperative adverse events that was unique to this population. Future research is needed to determine if this relationship is secondary to the medication itself or reflects a difference in how remifentanil is used in patients with reduced LVEF.

Background Heart failure (HF) is a major global cause of death. Early risk prediction and intervention could mitigate disease progression. We aimed to improve HF prediction by integrating genome-wide association studies (GWAS)- and electronic health records (EHR)-derived risk scores. Methods We previously performed a large HF GWAS within the Global Biobank Meta-analysis Initiative to create a polygenic risk score (PRS). Three Michigan Medicine (MM) cohorts were used to develop the clinical risk score (ClinRS): 1) Primary Care Provider cohort (MM-PCP; N = 61,849), 2) Heart Failure cohort (MM-HF; N = 53,272), and 3) Michigan Genomics Initiative cohort (MM-MGI; N = 60,215). To extract information from high-dimensional EHR data, we leveraged natural language processing to generate 350 latent phenotypes representing EHR codes and used coefficients from LASSO regression on these phenotypes in a training set as weights to calculate ClinRS in a validation set. Using logistic regression, model performances were compared between baseline model and models with risk scores added: 1) PRS, 2) ClinRS, and 3) ClinRS+PRS. We further compared the proposed models with Atherosclerosis Risk in Communities (ARIC) HF risk score. Results PRS and ClinRS each predict HF outcomes significantly better than the baseline model, up to eight years prior to HF diagnosis. Including both PRS and ClinRS further improves prediction performance up to ten years prior to diagnosis, two years earlier than either score alone. Additionally, ClinRS significantly outperforms the ARIC model one year prior. Conclusions We demonstrate the additive power of integrating GWAS- and EHR-derived risk scores to predict HF cases prior to diagnosis. This standardizable and scalable risk predictor may enable physicians to provide earlier interventions to improve patient outcomes. Plain language summary Heart failure (HF) is a leading cause of death worldwide. Early identification of individuals at high risk could facilitate interventions to slow disease progression. In this study, we develop an approach to improve HF risk prediction by combining patient genetic information and clinical information from electronic health records (EHR). We create two risk scores: a polygenic risk score (PRS) based on genetic information, and a clinical risk score (ClinRS) based on patient EHR. We test how well these scores predict HF before diagnosis. Both PRS and ClinRS improve predictions individually and identify high-risk individuals up to eight years in advance. When used together, they provide greater accuracy, predicting HF up to ten years before diagnosis. We suggest that combining genetic and clinical information could help doctors detect HF earlier for better treatment and prevention strategies in the future. Kuan-Han et al. examine if integrating patient genetic and clinical information from electronic health records can better predict heart failure in patients. Their findings show improvement in heart failure prediction up to ten years prior to diagnosis, which is two years earlier than using a single risk score alone.

Spontaneous coronary artery dissection (SCAD) is a non-atherosclerotic cause of myocardial infarction (MI), typically in young women. We undertook a genome-wide association study of SCAD (N cases  = 270/N controls  = 5,263) and identified and replicated an association of rs12740679 at chromosome 1q21.2 ( P discovery+replication  = 2.19 × 10 −12 , OR = 1.8) influencing ADAMTSL4 expression. Meta-analysis of discovery and replication samples identified associations with P  < 5 × 10 −8 at chromosome 6p24.1 in PHACTR1 , chromosome 12q13.3 in LRP1 , and in females-only, at chromosome 21q22.11 near LINC00310 . A polygenic risk score for SCAD was associated with (1) higher risk of SCAD in individuals with fibromuscular dysplasia ( P  = 0.021, OR = 1.82 [95% CI: 1.09–3.02]) and (2) lower risk of atherosclerotic coronary artery disease and MI in the UK Biobank ( P  = 1.28 × 10 −17 , HR = 0.91 [95% CI :0.89–0.93], for MI) and Million Veteran Program ( P  = 9.33 × 10 −36 , OR = 0.95 [95% CI: 0.94–0.96], for CAD; P  = 3.35 × 10 −6 , OR = 0.96 [95% CI: 0.95–0.98] for MI). Here we report that SCAD-related MI and atherosclerotic MI exist at opposite ends of a genetic risk spectrum, inciting MI with disparate underlying vascular biology. Spontaneous coronary artery dissection (SCAD) is a cause of myocardial infarction Here, the authors present a genome-wide association study of SCAD, finding an association at 1q21.2 which potentially affects expression of ADAMTSL4.

Background:The integration of artificial intelligence (AI) into clinical practice is transforming both clinical practice and medical education. AI-based systems aim to improve the efficacy of clinical tasks, enhancing diagnostic accuracy and tailoring treatment delivery. As it becomes increasingly prevalent in health care for high-quality patient care, it is critical for health care providers to use the systems responsibly to mitigate bias, ensure effective outcomes, and provide safe clinical practices. In this study, the clinical task is the identification of heart failure (HF) prior to surgery with the intention of enhancing clinical decision-making skills. HF is a common and severe disease, but detection remains challenging due to its subtle manifestation, often concurrent with other medical conditions, and the absence of a simple and effective diagnostic test. While advanced HF algorithms have been developed, the use of these AI-based systems to enhance clinical decision-making in medical education remains understudied.Objective:This research protocol is to demonstrate our study design, systematic procedures for selecting surgical cases from electronic health records, and interventions. The primary objective of this study is to measure the effectiveness of interventions aimed at improving HF recognition before surgery, the second objective is to evaluate the impact of inaccurate AI recommendations, and the third objective is to explore the relationship between the inclination to accept AI recommendations and their accuracy.Methods:Our study used a 3 × 2 factorial design (intervention type × order of prepost sets) for this randomized trial with medical students. The student participants are asked to complete a 30-minute e-learning module that includes key information about the intervention and a 5-question quiz, and a 60-minute review of 20 surgical cases to determine the presence of HF. To mitigate selection bias in the pre- and posttests, we adopted a feature-based systematic sampling procedure. From a pool of 703 expert-reviewed surgical cases, 20 were selected based on features such as case complexity, model performance, and positive and negative labels. This study comprises three interventions: (1) a direct AI-based recommendation with a predicted HF score, (2) an indirect AI-based recommendation gauged through the area under the curve metric, and (3) an HF guideline-based intervention.Results:As of July 2023, 62 of the enrolled medical students have fulfilled this study’s participation, including the completion of a short quiz and the review of 20 surgical cases. The subject enrollment commenced in August 2022 and will end in December 2023, with the goal of recruiting 75 medical students in years 3 and 4 with clinical experience.Conclusions:We demonstrated a study protocol for the randomized trial, measuring the effectiveness of interventions using AI and HF guidelines among medical students to enhance HF recognition in preoperative care with electronic health record data.International Registered Report Identifier (IRRID):DERR1-10.2196/49842