Background: Hospital Score is a well-known and validated tool for predicting readmission risk among diverse patient populations. Integrating social risk factors using natural language processing with the Hospital Score may improve its ability to predict 30-day readmissions following an acute myocardial infarction. Methods: A retrospective cohort included patients hospitalized at Vanderbilt University Medical Center between January 1, 2007, and December 31, 2016, with a primary index diagnosis of acute myocardial infarction, who were discharged alive. To supplement ascertainment of 30-day readmissions, data were linked to Center for Medicare & Medicaid Services (CMS) administrative data. Clinical notes from the cohort were extracted, and a natural language processing model was deployed, counting mentions of eight social risk factors. A logistic regression prediction model was run using the Hospital Score composite, its component variables, and the natural language processing-derived social risk factors. ROC comparison analysis was performed. Results: The cohort included 6,165 unique patients, where 4,137 (67.1%) were male, 1,020 (16.5%) were Black or other people of color, the average age was 67 years (SD:13), and the 30-day hospital readmission rate was 15.1% (N=934). The final test-set AUROCs were between 0.635 and 0.669. The model containing the Hospital Score component variables and the natural language processing-derived social risk factors obtained the highest AUROC. Discussion: Social risk factors extracted using natural language processing improved model performance when added to the Hospital Score composite. Clinicians and health systems should consider incorporating social risk factors when using the Hospital Score composite to evaluate risk for readmission among patients hospitalized for acute myocardial infarction.

Aim: The “2022 ACC/AHA Guideline for the Diagnosis and Management of Aortic Disease” provides recommendations to guide clinicians in the diagnosis, genetic evaluation and family screening, medical therapy, endovascular and surgical treatment, and long-term surveillance of patients with aortic disease across its multiple clinical presentation subsets (ie, asymptomatic, stable symptomatic, and acute aortic syndromes). Methods: A comprehensive literature search was conducted from January 2021 to April 2021, encompassing studies, reviews, and other evidence conducted on human subjects that were published in English from PubMed, EMBASE, the Cochrane Library, CINHL Complete, and other selected databases relevant to this guideline. Additional relevant studies, published through June 2022 during the guideline writing process, were also considered by the writing committee, where appropriate. Structure: Recommendations from previously published AHA/ACC guidelines on thoracic aortic disease, peripheral artery disease, and bicuspid aortic valve disease have been updated with new evidence to guide clinicians. In addition, new recommendations addressing comprehensive care for patients with aortic disease have been developed. There is added emphasis on the role of shared decision making, especially in the management of patients with aortic disease both before and during pregnancy. The is also an increased emphasis on the importance of institutional interventional volume and multidisciplinary aortic team expertise in the care of patients with aortic disease.

Understanding the conditionally-dependent clinical variables that drive cardiovascular health outcomes is a major challenge for precision medicine. Here, we deploy a recently developed massively scalable comorbidity discovery method called Poisson Binomial based Comorbidity discovery (PBC), to analyze Electronic Health Records (EHRs) from the University of Utah and Primary Children's Hospital (over 1.6 million patients and 77 million visits) for comorbid diagnoses, procedures, and medications. Using explainable Artificial Intelligence (AI) methodologies, we then tease apart the intertwined, conditionally-dependent impacts of comorbid conditions and demography upon cardiovascular health, focusing on the key areas of heart transplant, sinoatrial node dysfunction and various forms of congenital heart disease. The resulting multimorbidity networks make possible wide-ranging explorations of the comorbid and demographic landscapes surrounding these cardiovascular outcomes, and can be distributed as web-based tools for further community-based outcomes research. The ability to transform enormous collections of EHRs into compact, portable tools devoid of Protected Health Information solves many of the legal, technological, and data-scientific challenges associated with large-scale EHR analyses.

The "2022 ACC/AHA Guideline for the Diagnosis and Management of Aortic Disease" provides recommendations to guide clinicians in the diagnosis, genetic evaluation and family screening, medical therapy, endovascular and surgical treatment, and long-term surveillance of patients with aortic disease across its multiple clinical presentation subsets (ie, asymptomatic, stable symptomatic, and acute aortic syndromes).

The exponential growth of biomedical knowledge in computable formats challenges organizations to consider mobilizing artifacts in findable, accessible, interoperable, reusable, and trustable (FAIR+T) ways 1 . There is a growing need to apply biomedical knowledge artifacts to improve health in Learning Health Systems, health delivery organizations, and other settings. However, most organizations lack the infrastructure required to consume and apply computable knowledge, and national policies and standards adoption are insufficient to ensure that it is discoverable and used safely and fairly, nor is there widespread experience in the process of knowledge implementation as clinical decision support. The Mobilizing Computable Biomedical Knowledge (MCBK) community formed in 2016 to address these needs. This report summarizes the main outputs of the Fourth Annual MCBK public meeting, which was held virtually July 20 to July 21, 2021 and convened over 100 participants spanning diverse domains to frame and address important dimensions for mobilizing CBK.

Importance:In the US, more than 600 000 adults will experience an acute myocardial infarction (AMI) each year, and up to 20% of the patients will be rehospitalized within 30 days. This study highlights the need for consideration of calibration in these risk models. Objective:To compare multiple machine learning risk prediction models using an electronic health record (EHR)-derived data set standardized to a common data model. Design, Setting, and Participants:This was a retrospective cohort study that developed risk prediction models for 30-day readmission among all inpatients discharged from Vanderbilt University Medical Center between January 1, 2007, and December 31, 2016, with a primary diagnosis of AMI who were not transferred from another facility. The model was externally validated at Dartmouth-Hitchcock Medical Center from April 2, 2011, to December 31, 2016. Data analysis occurred between January 4, 2019, and November 15, 2020. Exposures:Acute myocardial infarction that required hospital admission. Main Outcomes and Measures:The main outcome was thirty-day hospital readmission. A total of 141 candidate variables were considered from administrative codes, medication orders, and laboratory tests. Multiple risk prediction models were developed using parametric models (elastic net, least absolute shrinkage and selection operator, and ridge regression) and nonparametric models (random forest and gradient boosting). The models were assessed using holdout data with area under the receiver operating characteristic curve (AUROC), percentage of calibration, and calibration curve belts. Results:The final Vanderbilt University Medical Center cohort included 6163 unique patients, among whom the mean (SD) age was 67 (13) years, 4137 were male (67.1%), 1019 (16.5%) were Black or other race, and 933 (15.1%) were rehospitalized within 30 days. The final Dartmouth-Hitchcock Medical Center cohort included 4024 unique patients, with mean (SD) age of 68 (12) years; 2584 (64.2%) were male, 412 (10.2%) were rehospitalized within 30 days, and most of the cohort were non-Hispanic and White. The final test set AUROC performance was between 0.686 to 0.695 for the parametric models and 0.686 to 0.704 for the nonparametric models. In the validation cohort, AUROC performance was between 0.558 to 0.655 for parametric models and 0.606 to 0.608 for nonparametric models. Conclusions and Relevance:In this study, 5 machine learning models were developed and externally validated to predict 30-day readmission AMI hospitalization. These models can be deployed within an EHR using routinely collected data.

We proposed two ICD-based topic modeling methods, named ICD-1 and ICD-2, which can generate topics based on the International Classification of Diseases (ICD) codes assigned to the documents. We applied the two methods to the Pittsburgh EHR dataset. For comparison, we also ran LDA on the same dataset to generate topics. Then we experimented with the three topic models on both document retrieval and sentence retrieval. As a baseline, we performed both retrievals using a keyword-matching method named TF-IDF. We evaluated the results using three methods: precision at ten (P@10), document ranking correlation, and sentence relevance determination (in terms of precision, recall, and F-score), which were based on the review and annotation made on the retrieved documents by two medical experts. In the P@10 evaluation, ICD-2 method achieved the highest average P@10 value of 0.61. In document ranking correlation, ICD-1 method achieved the highest Pearson’s correlation coefficient of 0.709. In sentence relevance determination, ICD-1 method achieved the highest F-score of 0.655. Overall, the ICD-based methods outperformed LDA and TF-IDF in the experiment.

Social determinants of health (SDoH) are increasingly important factors for population health, healthcare outcomes, and care delivery. However, many of these factors are not reliably captured within structured electronic health record (EHR) data. In this work, we evaluated and adapted a previously published NLP tool to include additional social risk factors for deployment at Vanderbilt University Medical Center in an Acute Myocardial Infarction cohort. We developed a transformation of the SDoH outputs of the tool into the OMOP common data model (CDM) for re-use across many potential use cases, yielding performance measures across 8 SDoH classes of precision 0.83 recall 0.74 and F-measure of 0.78.

INTRODUCTION:Computable biomedical knowledge artifacts (CBKs) are digital objects conveying biomedical knowledge in machine-interpretable structures. As more CBKs are produced and their complexity increases, the value obtained from sharing CBKs grows. Mobilizing CBKs and sharing them widely can only be achieved if the CBKs are findable, accessible, interoperable, reusable, and trustable (FAIR+T). To help mobilize CBKs, we describe our efforts to outline metadata categories to make CBKs FAIR+T.METHODS:We examined the literature regarding metadata with the potential to make digital artifacts FAIR+T. We also examined metadata available online today for actual CBKs of 12 different types. With iterative refinement, we came to a consensus on key categories of metadata that, when taken together, can make CBKs FAIR+T. We use subject-predicate-object triples to more clearly differentiate metadata categories.RESULTS:We defined 13 categories of CBK metadata most relevant to making CBKs FAIR+T. Eleven of these categories (type, domain, purpose, identification, location, CBK-to-CBK relationships, technical, authorization and rights management, provenance, evidential basis, and evidence from use metadata) are evident today where CBKs are stored online. Two additional categories (preservation and integrity metadata) were not evident in our examples. We provide a research agenda to guide further study and development of these and other metadata categories.CONCLUSION:A wide variety of metadata elements in various categories is needed to make CBKs FAIR+T. More work is needed to develop a common framework for CBK metadata that can make CBKs FAIR+T for all stakeholders.

The volume of biomedical knowledge is growing exponentially and much of this knowledge is represented in computer executable formats, such as models, algorithms, and programmatic code. There is a growing need to apply this knowledge to improve health in Learning Health Systems, health delivery organizations, and other settings. However, most organizations do not yet have the infrastructure required to consume and apply computable knowledge, and national policies and standards adoption are not sufficient to ensure that it is discoverable and used safely and fairly, nor is there widespread experience in the process of knowledge implementation as clinical decision support. The Mobilizing Computable Biomedical Knowledge (MCBK) community was formed in 2016 to address these needs. This report summarizes the main outputs of the third annual MCBK public meeting, which was held virtually from June 30 to July 1, 2020 and brought together over 200 participants from various domains to frame and address important dimensions for mobilizing CBK.

The volume of biomedical knowledge is growing exponentially and much of this knowledge is represented in computer executable formats, such as models, algorithms and programmatic code. There is a growing need to apply this knowledge to improve health in Learning Health Systems, health delivery organizations, and other settings. However, most organizations do not yet have the infrastructure required to consume and apply computable knowledge, and national policies and standards adoption are not sufficient to ensure that it is discoverable and used safely and fairly, nor is there widespread experience in the process of knowledge implementation as clinical decision support. The Mobilizing Computable Biomedical Knowledge (MCBK) community formed in 2016 to address these needs. This report summarizes the main outputs of the Second Annual MCBK public meeting, which was held at the National Institutes of Health on July 18-19, 2019 and brought together over 150 participants from various domains to frame and address important dimensions for mobilizing CBK.

BACKGROUND:More than 20% of patients admitted to the intensive care unit (ICU) develop an adverse event (AE). No previous study has leveraged patients' data to extract the temporal features using their structural temporal patterns, that is, trends.OBJECTIVE:This study aimed to improve AE prediction methods by using structural temporal pattern detection that captures global and local temporal trends and to demonstrate these improvements in the detection of acute kidney injury (AKI).METHODS:Using the Medical Information Mart for Intensive Care dataset, containing 22,542 patients, we extracted both global and local trends using structural pattern detection methods to predict AKI (ie, binary prediction). Classifiers were built on 17 input features consisting of vital signs and laboratory test results using state-of-the-art models; the optimal classifier was selected for comparisons with previous approaches. The classifier with structural pattern detection features was compared with two baseline classifiers that used different temporal feature extraction approaches commonly used in the literature: (1) symbolic temporal pattern detection, which is the most common approach for multivariate time series classification; and (2) the last recorded value before the prediction point, which is the most common approach to extract temporal data in the AKI prediction literature. Moreover, we assessed the individual contribution of global and local trends. Classifier performance was measured in terms of accuracy (primary outcome), area under the curve, and F-measure. For all experiments, we employed 20-fold cross-validation.RESULTS:Random forest was the best classifier using structural temporal pattern detection. The accuracy of the classifier with local and global trend features was significantly higher than that while using symbolic temporal pattern detection and the last recorded value (81.3% vs 70.6% vs 58.1%; P<.001). Excluding local or global features reduced the accuracy to 74.4% or 78.1%, respectively (P<.001).CONCLUSIONS:Classifiers using features obtained from structural temporal pattern detection significantly improved the prediction of AKI onset in ICU patients over two baselines based on common previous approaches. The proposed method is a generalizable approach to predict AEs in critical care that may be used to help clinicians intervene in a timely manner to prevent or mitigate AEs.

A work by a group of volunteer members drawn from the Mobilizing Computable Biomedical Knowledge community's Standards Workgroup. See mobilizecbk.org for more information about this community and workgroup.

Background: In ambulatory care settings, physicians largely rely on clinical guidelines and guideline-based clinical decision support (CDS) systems to make decisions on hypertension treatment. However, current clinical evidence, which is the knowledge base of clinical guidelines, is insufficient to support definitive optimal treatment. Objective: The goal of this study is to test the feasibility of using deep learning predictive models to identify optimal hypertension treatment pathways for individual patients, based on empirical data available from an electronic health record database. Materials and Methods: This study used data on 245,499 unique patients who were initially diagnosed with essential hypertension and received anti-hypertensive treatment from January 1, 2001 to December 31, 2010 in ambulatory care settings. We used recurrent neural networks (RNN), including long short-term memory (LSTM) and bi-directional LSTM, to create risk-adapted models to predict the probability of reaching the BP control targets associated with different BP treatment regimens. The ratios for the training set, the validation set, and the test set were 6:2:2. The samples for each set were independently randomly drawn from individual years with corresponding proportions. Results: The LSTM models achieved high accuracy when predicting individual probability of reaching BP goals on different treatments: for systolic BP (< 140 mmHg), diastolic BP (< 90 mmHg), and both systolic BP and diastolic BP (< 140/90 mmHg), Fl-scores were 0.928, 0.960, and 0.913, respectively. Conclusions: The results demonstrated the potential of using predictive models to select optimal hypertension treatment pathways. Along with clinical guidelines and guideline-based CDS systems, the LSTM models could be used as a powerful decision-support tool to form risk-adapted, personalized strategies for hypertension treatment plans, especially for difficult-to-treat patients.

We sought to evaluate the context of potential implementation of an automated quality measurement system for inpatients with heart failure in the U.S. Department of Veterans Affairs (VA). The research methodology was guided by the Promoting Action on Research Implementation in Health Sciences (PARIHS) framework and the sociotechnical model of health information technology. Data sources comprised semi-structured interviews ( n = 15), archival review of internal VA documents, and literature review. The interviewees consisted of four VA key informants and 11 subject matter experts (SMEs). Interviewees were VA quality management (QM) staff, clinicians, data analysts, and quality measurement experts, among others. Our interviews identified themes, which confirmed that the automated system is aligned with current internal organizational features, hardware and software infrastructure, and workflow and communication needs. We also identified facilitators and barriers to adoption of the automated system. The themes found will be used to inform future implementation of the system.

For clinical research to take advantage of artificial intelligence techniques such as the various types of deep neuralnetworks, we need to be able to explain the deep neural network models to clinicians and researchers. While someexplanation

BACKGROUND:A Cardiac-centered Frailty Ontology can be an important foundation for using NLP to assess patient frailty. Frailty is an important consideration when making patient treatment decisions, particularly in older adults, those with a cardiac diagnosis, or when major surgery is a consideration. Clinicians often report patient's frailty in progress notes and other documentation. Frailty is recorded in many different ways in patient records and many different validated frailty-measuring instruments are available, with little consistency across instruments. We specifically explored concepts relevant to decisions regarding cardiac interventions. We based our work on text found in a large corpus of clinical notes from the Department of Veterans Affairs (VA) national Electronic Health Record (EHR) database.RESULTS:The full ontology has 156 concepts, with 246 terms. It includes 86 concepts we expect to find in clinical documents, with 12 qualifier values. The remaining 58 concepts represent hierarchical groups (e.g., physical function findings). Our top-level class is clinical finding, which has children clinical history finding, instrument finding, and physical examination finding, reflecting the OGMS definition of clinical finding. Instrument finding is any score found for the existing frailty instruments. Within our ontology, we used SNOMED-CT concepts where possible. Some of the 86 concepts we expect to find in clinical documents are associated with the properties like ability interpretation. The concept ability to walk can either be able, assisted or unable. Each concept-property level pairing gets a different frailty score. Each scored concept received three scores: a frailty score, a relevance to cardiac decisions score, and a likelihood of resolving after the recommended intervention score. The ontology includes the relationship between scores from ten frailty instruments and frailty as assessed using ontology concepts. It also included rules for mapping ontology elements to instrument items for three common frailty assessment instruments. Ontology elements are used in two clinical NLP systems.CONCLUSIONS:We developed and validated a Cardiac-centered Frailty Ontology, which is a machine-interoperable description of frailty that reflects all the areas that clinicians consider when deciding which cardiac intervention will best serve the patient as well as frailty indications generally relevant to medical decisions. The ontology owl file is available on Bioportal at http://bioportal.bioontology.org/ontologies/CCFO .

Background:The impact on workfl ow is an important component in determining whether an HIT implementation will be successful.Workfl ow is, unfortunately, a concept that is often ignored when implementing HIT and the literature about workfl ow in domains of quality improvement, system implementation, and process improvement has not been adequate.HIT is not always designed to fi t the workfl ow of a given organization, making it diffi cult to truly assess HIT impact on outcomes or processes. The context and purpose of the study:The purpose of this study was to establish an initial toolkit, which would be generalizable to assess the impact of implementation on the workfl ow of quality improvement and information technology professionals' and their team in the inpatient hospital setting.The toolkit, in the form of an implementation guide, is a compilation of resources such as checklists, forms, and planning documents that provided a template for workfl ow analysis. Results:The expert evaluators rated the tools as moderately relevant, and moderately clear.The postimplementation tools were rated highest for relevance.Comments predominantly highlighted areas of the toolkit that needed additional depth or detail in the toolkit, rather than suggesting additional tools.Main fi ndings: This study provided a methodology for identifying information needs, detecting confl icts, and implementing possible workfl ow solutions; thus, the toolkit could serve as a pragmatically useful implementation guide to assess the workfl ow pertaining to the quality improvement and information technology professionals' and their team during the implementation of eMeasures at an inpatient hospital setting. Conclusions:The toolkit provided a useful collection of tools in the form of checklists, forms, and planning documents to enhance the workfl ow during implementation of heart failure (HF) eMeasures. Brief summary:We developed an initial eMeasure Implementation Toolkit for the heart failure (HF) eMeasure to allow QI and information technology (IT) professionals and their team to assess the impact of implementation on workfl ow.During the development of the toolkit, we conducted stakeholder interviews with the VA key informants and subject matter experts, who provided valuable information about the context of understanding the workfl ow during the implementation of eMeasures.The information obtained from the stakeholder engagement highlighted areas and tools that were essential to the toolkit development phase.The fi nal phase involved the evaluation of the eMeasure Implementation Toolkit for relevance and clarity by non-VA experts.The non-VA subject matter experts predominantly evaluated the sections of the toolkit that contained the tools for evaluating the workfl ow during eMeasure implementation via a survey. Any potential implications:This eMeasure Implementation Toolkit should continue to be refi ned.The toolkit should undergo further evaluation with a variety of subject matter experts from various job categories and medical centers.Second, the toolkit should be tailored to the level of cognition of each user, while focusing on the syntax used to describe the sections of the toolkit.Finally, the toolkit could be developed for other quality measures such as stroke, diabetes, pneumonia, etc., based on the research methodology used in the development of the existing toolkit.

Death after major cardiovascular procedures (MCVP) is common among older patients, yet current prediction models do not account for frailty. We compared two machine learning models (deep neural network [DNN] and support vector machine [SVM]), incorporating frailty, to predict death after MCVP. We

There is limited guidance available in the literature for establishing clinical decision support (CDS) governance and improving CDS effectiveness in a pragmatic, resource-efficient manner. Here, we describe how University of Utah Health established enterprise CDS governance in 2015 leveraging existing resources. Key components of the governance include a multi-stakeholder CDS Committee that vets new requests and reviews existing content; a requirement that proposed CDS is actually desired by intended recipients; coordination with other governance bodies; basic data analytics to identify high-frequency, low-value CDS and monitor progress; active solicitation of user issues; the transition of alert and reminder content to other, more appropriate areas in the electronic health record; and the judicious use of experimental designs to guide decision-making regarding CDS effectiveness. In the three years since establishing this governance, new CDS has been continuously added while the overall burden of clinician-facing alerts and reminders has been reduced by 53.8%.