Machine Learning-Based Models Incorporating Social Determinants of Health vs Traditional Models for Predicting In-Hospital Mortality in Patients With Heart Failure

Effects of environmental phenols on eGFR: machine learning modeling methods applied to cross-sectional studies

Purpose

Limited investigation is available on the correlation between environmental phenols' exposure and estimated glomerular filtration rate (eGFR). Our target is established a robust and explainable machine learning (ML) model that associates environmental phenols' exposure with eGFR.

Methods

Our datasets for constructing the associations between environmental phenols' and eGFR were collected from the National Health and Nutrition Examination Survey (NHANES, 2013-2016). Five ML models were contained and fine-tuned to eGFR regression by phenols' exposure. Regression evaluation metrics were used to extract the limitation of the models. The most effective model was then utilized for regression, with interpretation of its features carried out using shapley additive explanations (SHAP) and the game theory python package to represent the model's regression capacity.

Results

The study identified the top-performing random forest (RF) regressor with a mean absolute error of 0.621 and a coefficient of determination of 0.998 among 3,371 participants. Six environmental phenols with eGFR in linear regression models revealed that the concentrations of triclosan (TCS) and bisphenol S (BPS) in urine were positively correlated with eGFR, and the correlation coefficients were β = 0.010 (p = 0.026) and β = 0.007 (p = 0.004) respectively. SHAP values indicate that BPS (1.38), bisphenol F (BPF) (0.97), 2,5-dichlorophenol (0.87), TCS (0.78), BP3 (0.60), bisphenol A (BPA) (0.59) and 2,4-dichlorophenol (0.47) in urinary contributed to the model.

Conclusion

The RF model was efficient in identifying a correlation between phenols' exposure and eGFR among United States NHANES 2013-2016 participants. The findings indicate that BPA, BPF, and BPS are inversely associated with eGFR.

Influence of social vulnerability index on Medicare beneficiaries' expenditures upon discharge

Medicare beneficiaries' healthcare spending varies across geographical regions, influenced by availability of medical resources and institutional efficiency. We aimed to evaluate whether social vulnerability influences healthcare costs among Medicare beneficiaries. Multivariable regression analyses were conducted to determine whether the social vulnerability index (SVI), released by the Centers for Disease Control and Prevention (CDC), was associated with average submitted covered charges, total payment amounts, or total covered days upon hospital discharge among Medicare beneficiaries. We used information from discharged Medicare beneficiaries from hospitals participating in the Inpatient Prospective Payment System. Covariate adjustment included demographic information consisting of age groups, race/ethnicity, and Hierarchical Condition Category risk score. The regressions were performed with weights proportioned to the number of discharges. Average submitted covered charges significantly correlated with SVI (β = 0.50, p < 0.001) in the unadjusted model and remained significant in the covariates-adjusted model (β = 0.25, p = 0.039). The SVI was not significantly associated with the total payment amounts (β = -0.07, p = 0.238) or the total covered days (β = 0.00, p = 0.953) in the adjusted model. Regional variations in Medicare beneficiaries' healthcare spending exist and are influenced by levels of social vulnerability. Further research is warranted to fully comprehend the impact of social determinants on healthcare costs.

A Systematic Review of Artificial Intelligence Models for Time-to-Event Outcome Applied in Cardiovascular Disease Risk Prediction

Artificial intelligence (AI) based predictive models for early detection of cardiovascular disease (CVD) risk are increasingly being utilised. However, AI based risk prediction models that account for right-censored data have been overlooked. This systematic review (PROSPERO protocol CRD42023492655) includes 33 studies that utilised machine learning (ML) and deep learning (DL) models for survival outcome in CVD prediction. We provided details on the employed ML and DL models, eXplainable AI (XAI) techniques, and type of included variables, with a focus on social determinants of health (SDoH) and gender-stratification. Approximately half of the studies were published in 2023 with the majority from the United States. Random Survival Forest (RSF), Survival Gradient Boosting models, and Penalised Cox models were the most frequently employed ML models. DeepSurv was the most frequently employed DL model. DL models were better at predicting CVD outcomes than ML models. Permutation-based feature importance and Shapley values were the most utilised XAI methods for explaining AI models. Moreover, only one in five studies performed gender-stratification analysis and very few incorporate the wide range of SDoH factors in their prediction model. In conclusion, the evidence indicates that RSF and DeepSurv models are currently the optimal models for predicting CVD outcomes. This study also highlights the better predictive ability of DL survival models, compared to ML models. Future research should ensure the appropriate interpretation of AI models, accounting for SDoH, and gender stratification, as gender plays a significant role in CVD occurrence.

Integrating Social Drivers of Health into Hospital Ratings with Application to the 100 Top Hospitals Study

The objective was to investigate the relationship between social drivers of health (SDOH) and hospital performance within the 100 Top Hospitals study, exploring methods to recognize hospitals serving marginalized communities. Publicly available data sourced from the Centers for Medicare and Medicaid Services and the 2023 100 Top Hospitals study was used. The study employed multivariable hierarchical generalized linear regression models to assess the association between an SDOH composite variable derived using principal component analysis and overall hospital performance measures within the 100 Top Hospitals study. The analysis revealed a statistically significant association between SDOH factors and study ranking results. The SDOH composite variable is a significant predictor of performance within the 100 Top Hospitals study. Accounting for SDOH is essential to recognize high-performing hospitals serving marginalized communities. The findings suggest a need for broader considerations of SDOH in hospital ranking methodologies across various industry programs.

Predicting the risk of 1-year mortality among patients hospitalized for acute heart failure in China

BACKGROUND

We aimed to develop and validate a model to predict 1-year mortality risk among patients hospitalized for acute heart failure (AHF), build a risk score and interpret its application in clinical decision making.

METHODS

By using data from China Patient-Centred Evaluative Assessment of Cardiac Events Prospective Heart Failure Study, which prospectively enrolled patients hospitalized for AHF in 52 hospitals across 20 provinces, we used multivariate Cox proportional hazard model to develop and validate a model to predict 1-year mortality.

RESULTS

There were 4,875 patients included in the study, 857 (17.58%) of them died within 1-year following discharge of index hospitalization. A total of 13 predictors were selected to establish the prediction model, including age, medical history of chronic obstructive pulmonary disease and hypertension, systolic blood pressure, Kansas City Cardiomyopathy Questionnaire-12 score, angiotensin converting enzyme inhibitor or angiotensin receptor blocker at discharge, discharge symptom, N-terminal pro-brain natriuretic peptide, high-sensitivity troponin T, serum creatine, albumin, blood urea nitrogen, and highly sensitive C-reactive protein. The model showed a high performance on discrimination (C-index was 0.759 [95% confidence interval: 0.739, 0.778] in development cohort and 0.761 [95% confidence interval: 0.731, 0.791] in validation cohort), accuracy, calibration, and outperformed than several existed risk scores. A point-based risk score was built to stratify low- (0-12), intermediate- (13-16), and high-risk group (≥17) among patients.

CONCLUSIONS

A prediction model using readily available predictors was developed and internal validated to predict 1-year mortality risk among patients hospitalized for AHF. It may serve as a useful tool for individual risk stratification and informing decision making to improve clinical care.

Best of Both Worlds: Bridging One Model for All and Group-Specific Model Approaches using Ensemble-based Subpopulation Modeling

Subpopulation models have become of increasing interest in prediction of clinical outcomes because they promise to perform better for underrepresented patient subgroups. However, the personalization benefits gained from these models tradeoff their statistical power, and can be impractical when the subpopulation's sample size is small. We hypothesize that a hierarchical model in which population information is integrated into subpopulation models would preserve the personalization benefits and offset the loss of power. In this work, we integrate ideas from ensemble modeling, personalization, and hierarchical modeling and build ensemble-based subpopulation models in which specialization relies on whole group samples. This approach significantly improves the precision of the positive class, especially for the underrepresented subgroups, with minimal cost to the recall. It consistently outperforms one model for all and one model for each subgroup approaches, especially in the presence of a high class-imbalance, for subgroups with at least 380 training samples.

StrokeClassifier: ischemic stroke etiology classification by ensemble consensus modeling using electronic health records

Determining acute ischemic stroke (AIS) etiology is fundamental to secondary stroke prevention efforts but can be diagnostically challenging. We trained and validated an automated classification tool, StrokeClassifier, using electronic health record (EHR) text from 2039 non-cryptogenic AIS patients at 2 academic hospitals to predict the 4-level outcome of stroke etiology adjudicated by agreement of at least 2 board-certified vascular neurologists' review of the EHR. StrokeClassifier is an ensemble consensus meta-model of 9 machine learning classifiers applied to features extracted from discharge summary texts by natural language processing. StrokeClassifier was externally validated in 406 discharge summaries from the MIMIC-III dataset reviewed by a vascular neurologist to ascertain stroke etiology. Compared with vascular neurologists' diagnoses, StrokeClassifier achieved the mean cross-validated accuracy of 0.74 and weighted F1 of 0.74 for multi-class classification. In MIMIC-III, its accuracy and weighted F1 were 0.70 and 0.71, respectively. In binary classification, the two metrics ranged from 0.77 to 0.96. The top 5 features contributing to stroke etiology prediction were atrial fibrillation, age, middle cerebral artery occlusion, internal carotid artery occlusion, and frontal stroke location. We designed a certainty heuristic to grade the confidence of StrokeClassifier's diagnosis as non-cryptogenic by the degree of consensus among the 9 classifiers and applied it to 788 cryptogenic patients, reducing cryptogenic diagnoses from 25.2% to 7.2%. StrokeClassifier is a validated artificial intelligence tool that rivals the performance of vascular neurologists in classifying ischemic stroke etiology. With further training, StrokeClassifier may have downstream applications including its use as a clinical decision support system.

Promoting Artificial Intelligence for Global Breast Cancer Risk Prediction and Screening in Adult Women: A Scoping Review

Background: Artificial intelligence (AI) algorithms can be applied in breast cancer risk prediction and prevention by using patient history, scans, imaging information, and analysis of specific genes for cancer classification to reduce overdiagnosis and overtreatment. This scoping review aimed to identify the barriers encountered in applying innovative AI techniques and models in developing breast cancer risk prediction scores and promoting screening behaviors among adult females. Findings may inform and guide future global recommendations for AI application in breast cancer prevention and care for female populations. Methods: The PRISMA-SCR (Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews) was used as a reference checklist throughout this study. The Arksey and O'Malley methodology was used as a framework to guide this review. The framework methodology consisted of five steps: (1) Identify research questions; (2) Search for relevant studies; (3) Selection of studies relevant to the research questions; (4) Chart the data; (5) Collate, summarize, and report the results. Results: In the field of breast cancer risk detection and prevention, the following AI techniques and models have been applied: Machine and Deep Learning Model (ML-DL model) (n = 1), Academic Algorithms (n = 2), Breast Cancer Surveillance Consortium (BCSC), Clinical 5-Year Risk Prediction Model (n = 2), deep-learning computer vision AI algorithms (n = 2), AI-based thermal imaging solution (Thermalytix) (n = 1), RealRisks (n = 2), Breast Cancer Risk NAVIgation (n = 1), MammoRisk (ML-Based Tool) (n = 1), Various MLModels (n = 1), and various machine/deep learning, decision aids, and commercial algorithms (n = 7). In the 11 included studies, a total of 39 barriers to AI applications in breast cancer risk prediction and screening efforts were identified. The most common barriers in the application of innovative AI tools for breast cancer prediction and improved screening rates included lack of external validity and limited generalizability (n = 6), as AI was used in studies with either a small sample size or datasets with missing data. Many studies (n = 5) also encountered selection bias due to exclusion of certain populations based on characteristics such as race/ethnicity, family history, or past medical history. Several recommendations for future research should be considered. AI models need to include a broader spectrum and more complete predictive variables for risk assessment. Investigating long-term outcomes with improved follow-up periods is critical to assess the impacts of AI on clinical decisions beyond just the immediate outcomes. Utilizing AI to improve communication strategies at both a local and organizational level can assist in informed decision-making and compliance, especially in populations with limited literacy levels. Conclusions: The use of AI in patient education and as an adjunctive tool for providers is still early in its incorporation, and future research should explore the implementation of AI-driven resources to enhance understanding and decision-making regarding breast cancer screening, especially in vulnerable populations with limited literacy.

Effectiveness and safety of early treatment with spironolactone for new-onset acute heart failure

BACKGROUND

The effectiveness and safety of mineralocorticoid receptor antagonists (MRA) in acute heart failure (HF) is uncertain. We sought to describe the prescription of spironolactone during acute HF and whether early treatment is effective and safe in a real-world setting.

METHODS

We performed a retrospective cohort study of adult (≥18 years) nonpregnant patients hospitalized with new-onset HF with reduced ejection fraction (HFrEF, defined by ejection fraction ≤40%) within 15 Kaiser Permanente Southern California medical centers between 2016 and 2021. Early treatment was defined by spironolactone prescription at discharge. The primary effectiveness outcome was a composite of HF readmission or all-cause mortality at 180 days. Safety outcomes were hypotension and hyperkalemia at 90 days.

RESULTS

Among 2318 HFrEF patients, 368 (15.9%) were treated with spironolactone at discharge. After 1:2 propensity score matching, 354 early treatment and 708 delayed/no treatment patients were included in the analysis. The median age was 63 (IQR: 52-74) years; 61.6% were male, and 38.6% were White. By 90 days, ~20% had crossed over in the two groups. Early treatment was not associated with the composite outcome at 180 days (HR [95% CI]: 0.81 [0.56-1.17]), but a trend towards benefit by 365 days that did not reach statistical significance (0.78 [0.58-1.06]). Early treatment was also associated with hyperkalemia (subdistribution HR [95% CI]: 2.33 [1.30-4.18]) but not hypotension (0.93 [0.51-1.72]).

CONCLUSIONS

Early treatment with spironolactone at discharge for new-onset HFrEF in a real-world setting did not reduce the risk of HF readmission or mortality in the first year after discharge. The risk of hyperkalemia was increased.

Hierarchy of demographic and social determinants of mental health: analysis of cross-sectional survey data from the Global Mind Project

OBJECTIVES

To understand the extent to which various demographic and social determinants predict mental health status and their relative hierarchy of predictive power in order to prioritise and develop population-based preventative approaches.

DESIGN

Cross-sectional analysis of survey data.

SETTING

Internet-based survey from 32 countries across North America, Europe, Latin America, Middle East and North Africa, Sub-Saharan Africa, South Asia and Australia, collected between April 2020 and December 2021.

PARTICIPANTS

270 000 adults aged 18-85+ years who participated in the Global Mind Project.

OUTCOME MEASURES

We used 120+ demographic and social determinants to predict aggregate mental health status and scores of individuals (mental health quotient (MHQ)) and determine their relative predictive influence using various machine learning models including gradient boosting and random forest classification for various demographic stratifications by age, gender, geographical region and language. Outcomes reported include model performance metrics of accuracy, precision, recall, F1 scores and importance of individual factors determined by reduction in the squared error attributable to that factor.

RESULTS

Across all demographic classification models, 80% of those with negative MHQs were correctly identified, while regression models predicted specific MHQ scores within ±15% of the position on the scale. Predictions were higher for older ages (0.9+ accuracy, 0.9+ F1 Score; 65+ years) and poorer for younger ages (0.68 accuracy, 0.68 F1 Score; 18-24 years). Across all age groups, genders, regions and language groups, lack of social interaction and sufficient sleep were several times more important than all other factors. For younger ages (18-24 years), other highly predictive factors included cyberbullying and sexual abuse while not being able to work was high for ages 45-54 years.

CONCLUSION

Social determinants of traumas, adversities and lifestyle can account for 60%-90% of mental health challenges. However, additional factors are at play, particularly for younger ages, that are not included in these data and need further investigation.

Adopting artificial intelligence in cardiovascular medicine: a scoping review

Recent years have witnessed significant transformations in cardiovascular medicine, driven by the rapid evolution of artificial intelligence (AI). This scoping review was conducted to capture the breadth of AI applications within cardiovascular science. Employing a structured approach, we sourced relevant articles from PubMed, with an emphasis on journals encompassing general cardiology and digital medicine. We applied filters to highlight cardiovascular articles published in journals focusing on general internal medicine, cardiology and digital medicine, thereby identifying the prevailing trends in the field. Following a comprehensive full-text screening, a total of 140 studies were identified. Over the preceding 5 years, cardiovascular medicine's interplay with AI has seen an over tenfold augmentation. This expansive growth encompasses multiple cardiovascular subspecialties, including but not limited to, general cardiology, ischemic heart disease, heart failure, and arrhythmia. Deep learning emerged as the predominant methodology. The majority of AI endeavors in this domain have been channeled toward enhancing diagnostic and prognostic capabilities, utilizing resources such as hospital datasets, electrocardiograms, and echocardiography. A significant uptrend was observed in AI's application for omics data analysis. However, a clear gap persists in AI's full-scale integration into the clinical decision-making framework. AI, particularly deep learning, has demonstrated robust applications across cardiovascular subspecialties, indicating its transformative potential in this field. As we continue on this trajectory, ensuring the alignment of technological progress with medical ethics becomes crucial. The abundant digital health data today further accentuates the need for meticulous systematic reviews, tailoring them to each cardiovascular subspecialty.

Comparison of predicting cardiovascular disease hospitalization using individual, ZIP code-derived, and machine learning model-predicted educational attainment in New York City

BACKGROUND

Area-level social determinants of health (SDOH) based on patients' ZIP codes or census tracts have been commonly used in research instead of individual SDOHs. To our knowledge, whether machine learning (ML) could be used to derive individual SDOH measures, specifically individual educational attainment, is unknown.

METHODS

This is a retrospective study using data from the Mount Sinai BioMe Biobank. We included participants that completed a validated questionnaire on educational attainment and had home addresses in New York City. ZIP code-level education was derived from the American Community Survey matched for the participant's gender and race/ethnicity. We tested several algorithms to predict individual educational attainment from routinely collected clinical and demographic data. To evaluate how using different measures of educational attainment will impact model performance, we developed three distinct models for predicting cardiovascular (CVD) hospitalization. Educational attainment was imputed into models as either survey-derived, ZIP code-derived, or ML-predicted educational attainment.

RESULTS

A total of 20,805 participants met inclusion criteria. Concordance between survey and ZIP code-derived education was 47%, while the concordance between survey and ML model-predicted education was 67%. A total of 13,715 patients from the cohort were included into our CVD hospitalization prediction models, of which 1,538 (11.2%) had a history of CVD hospitalization. The AUROC of the model predicting CVD hospitalization using survey-derived education was significantly higher than the model using ZIP code-level education (0.77 versus 0.72; p < 0.001) and the model using ML model-predicted education (0.77 versus 0.75; p < 0.001). The AUROC for the model using ML model-predicted education was also significantly higher than that using ZIP code-level education (p = 0.003).

CONCLUSION

The concordance of survey and ZIP code-level educational attainment in NYC was low. As expected, the model utilizing survey-derived education achieved the highest performance. The model incorporating our ML model-predicted education outperformed the model relying on ZIP code-derived education. Implementing ML techniques can improve the accuracy of SDOH data and consequently increase the predictive performance of outcome models.

Evaluation of the Predictive Value of Routinely Collected Health-Related Social Needs Measures

The objective was to assess the value of routinely collected patient-reported health-related social needs (HRSNs) measures for predicting utilization and health outcomes. The authors identified Mayo Clinic patients with cancer, diabetes, or heart failure. The HRSN measures were collected as part of patient-reported screenings from June to December 2019 and outcomes (hospitalization, 30-day readmission, and death) were ascertained in 2020. For each outcome and disease combination, 4 models were used: gradient boosting machine (GBM), random forest (RF), generalized linear model (GLM), and elastic net (EN). Other predictors included clinical factors, demographics, and area-based HRSN measures-area deprivation index (ADI) and rurality. Predictive performance for models was evaluated with and without the routinely collected HRSN measures as change in area under the curve (AUC). Variable importance was also assessed. The differences in AUC were mixed. Significant improvements existed in 3 models of death for cancer (GBM: 0.0421, RF: 0.0496, EN: 0.0428), 3 models of hospitalization (GBM: 0.0372, RF: 0.0640, EN: 0.0441), and 1 of death (RF: 0.0754) for diabetes, and 1 model of readmissions (GBM: 0.1817), and 3 models of death (GBM: 0.0333, RF: 0.0519, GLM: 0.0489) for heart failure. Age, ADI, and the Charlson comorbidity index were the top 3 in variable importance and were consistently more important than routinely collected HRSN measures. The addition of routinely collected HRSN measures resulted in mixed improvement in the predictive performance of the models. These findings suggest that existing factors and the ADI are more important for prediction in these contexts. More work is needed to identify predictors that consistently improve model performance.

2023 American Heart Association Focused Update on Adult Advanced Cardiovascular Life Support: An Update to the American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care

Cardiac arrest is common and deadly, affecting up to 700 000 people in the United States annually. Advanced cardiac life support measures are commonly used to improve outcomes. This "2023 American Heart Association Focused Update on Adult Advanced Cardiovascular Life Support" summarizes the most recent published evidence for and recommendations on the use of medications, temperature management, percutaneous coronary angiography, extracorporeal cardiopulmonary resuscitation, and seizure management in this population. We discuss the lack of data in recent cardiac arrest literature that limits our ability to evaluate diversity, equity, and inclusion in this population. Last, we consider how the cardiac arrest population may make up an important pool of organ donors for those awaiting organ transplantation.

Principles of artificial intelligence and its application in cardiovascular medicine

Artificial intelligence (AI) represents a rapidly developing field. Its use can improve diagnosis and therapy in many areas of medicine. Despite this enormous progress, many physicians perceive it as a black box and are skeptical about it. This review will present the basics of machine learning. Different classifications of artificial intelligence, such as supervised versus unsupervised and discriminative versus generative AI, are given. Analogies to human intelligence are discussed as far as algorithms are oriented toward it. In the second step, the most common models like random forest, k-means clustering, convolutional neural network, and transformers will be presented in a way that the underlying idea can be understood. Corresponding medical applications in cardiovascular medicine will be named for all models, respectively. The overview is intended to show that the term artificial intelligence covers a wide range of different concepts. It should help physicians understand the principles of AI to make up one's minds about its application in cardiology. It should also enable them to evaluate results obtained with AI's help critically.

Neighborhoods and Cardiovascular Health: A Scientific Statement From the American Heart Association

The neighborhoods where individuals reside shape environmental exposures, access to resources, and opportunities. The inequitable distribution of resources and opportunities across neighborhoods perpetuates and exacerbates cardiovascular health inequities. Thus, interventions that address the neighborhood environment could reduce the inequitable burden of cardiovascular disease in disenfranchised populations. The objective of this scientific statement is to provide a roadmap illustrating how current knowledge regarding the effects of neighborhoods on cardiovascular disease can be used to develop and implement effective interventions to improve cardiovascular health at the population, health system, community, and individual levels. PubMed/Medline, CINAHL, Cochrane Library reviews, and ClinicalTrials.gov were used to identify observational studies and interventions examining or targeting neighborhood conditions in relation to cardiovascular health. The scientific statement summarizes how neighborhoods have been incorporated into the actions of health care systems, interventions in community settings, and policies and interventions that involve modifying the neighborhood environment. This scientific statement presents promising findings that can be expanded and implemented more broadly and identifies methodological challenges in designing studies to evaluate important neighborhood-related policies and interventions. Last, this scientific statement offers recommendations for areas that merit further research to promote a deeper understanding of the contributions of neighborhoods to cardiovascular health and health inequities and to stimulate the development of more effective interventions.

StrokeClassifier: Ischemic Stroke Etiology Classification by Ensemble Consensus Modeling Using Electronic Health Records

Determining the etiology of an acute ischemic stroke (AIS) is fundamental to secondary stroke prevention efforts but can be diagnostically challenging. We trained and validated an automated classification machine intelligence tool, StrokeClassifier, using electronic health record (EHR) text data from 2,039 non-cryptogenic AIS patients at 2 academic hospitals to predict the 4-level outcome of stroke etiology determined by agreement of at least 2 board-certified vascular neurologists' review of the stroke hospitalization EHR. StrokeClassifier is an ensemble consensus meta-model of 9 machine learning classifiers applied to features extracted from discharge summary texts by natural language processing. StrokeClassifier was externally validated in 406 discharge summaries from the MIMIC-III dataset reviewed by a vascular neurologist to ascertain stroke etiology. Compared with stroke etiologies adjudicated by vascular neurologists, StrokeClassifier achieved the mean cross-validated accuracy of 0.74 (±0.01) and weighted F1 of 0.74 (±0.01). In the MIMIC-III cohort, the accuracy and weighted F1 of StrokeClassifier were 0.70 and 0.71, respectively. SHapley Additive exPlanation analysis elucidated that the top 5 features contributing to stroke etiology prediction were atrial fibrillation, age, middle cerebral artery occlusion, internal carotid artery occlusion, and frontal stroke location. We then designed a certainty heuristic to deem a StrokeClassifier diagnosis as confidently non-cryptogenic by the degree of consensus among the 9 classifiers, and applied it to 788 cryptogenic patients. This reduced the percentage of the cryptogenic strokes from 25.2% to 7.2% of all ischemic strokes. StrokeClassifier is a validated artificial intelligence tool that rivals the performance of vascular neurologists in classifying ischemic stroke etiology for individual patients. With further training, StrokeClassifier may have downstream applications including its use as a clinical decision support system.

Mitigating Racial And Ethnic Bias And Advancing Health Equity In Clinical Algorithms: A Scoping Review

In August 2022 the Department of Health and Human Services (HHS) issued a notice of proposed rulemaking prohibiting covered entities, which include health care providers and health plans, from discriminating against individuals when using clinical algorithms in decision making. However, HHS did not provide specific guidelines on how covered entities should prevent discrimination. We conducted a scoping review of literature published during the period 2011-22 to identify health care applications, frameworks, reviews and perspectives, and assessment tools that identify and mitigate bias in clinical algorithms, with a specific focus on racial and ethnic bias. Our scoping review encompassed 109 articles comprising 45 empirical health care applications that included tools tested in health care settings, 16 frameworks, and 48 reviews and perspectives. We identified a wide range of technical, operational, and systemwide bias mitigation strategies for clinical algorithms, but there was no consensus in the literature on a single best practice that covered entities could employ to meet the HHS requirements. Future research should identify optimal bias mitigation methods for various scenarios, depending on factors such as patient population, clinical setting, algorithm design, and types of bias to be addressed.

Race as a Component of Cardiovascular Disease Risk Prediction Algorithms

PURPOSE OF REVIEW

Several prediction algorithms include race as a component to account for race-associated variations in disease frequencies. This practice has been questioned recently because of the risk of perpetuating race as a biological construct and diverting attention away from the social determinants of health (SDoH) for which race might be a proxy. We evaluated the appropriateness of including race in cardiovascular disease (CVD) prediction algorithms, notably the pooled cohort equations (PCE).

RECENT FINDINGS

In a recent investigation, we reported substantial and biologically implausible differences in absolute CVD risk estimates upon using PCE for predicting CVD risk in Black and White persons with identical risk factor profiles, which might result in differential treatment decisions based solely on their race. We recommend the development of raceless CVD risk prediction algorithms that obviate race-associated risk misestimation and racializing treatment practices, and instead incorporate measures of SDoH that mediate race-associated risk differences.

Social Determinants of Health Data Improve the Prediction of Cardiac Outcomes in Females with Breast Cancer

Cardiovascular disease is the leading cause of mortality among breast cancer (BC) patients aged 50 and above. Machine Learning (ML) models are increasingly utilized as prediction tools, and recent evidence suggests that incorporating social determinants of health (SDOH) data can enhance its performance. This study included females ≥ 18 years diagnosed with BC at any stage. The outcomes were the diagnosis and time-to-event of major adverse cardiovascular events (MACEs) within two years following a cancer diagnosis. Covariates encompassed demographics, risk factors, individual and neighborhood-level SDOH, tumor characteristics, and BC treatment. Race-specific and race-agnostic Extreme Gradient Boosting ML models with and without SDOH data were developed and compared based on their C-index. Among 4309 patients, 11.4% experienced a 2-year MACE. The race-agnostic models exhibited a C-index of 0.78 (95% CI 0.76-0.79) and 0.81 (95% CI 0.80-0.82) without and with SDOH data, respectively. In non-Hispanic Black women (NHB; n = 765), models without and with SDOH data achieved a C-index of 0.74 (95% CI 0.72-0.76) and 0.75 (95% CI 0.73-0.78), respectively. Among non-Hispanic White women (n = 3321), models without and with SDOH data yielded a C-index of 0.79 (95% CI 0.77-0.80) and 0.79 (95% CI 0.77-0.80), respectively. In summary, including SDOH data improves the predictive performance of ML models in forecasting 2-year MACE among BC females, particularly within NHB.

Unraveling the Ethical Enigma: Artificial Intelligence in Healthcare

The integration of artificial intelligence (AI) into healthcare promises groundbreaking advancements in patient care, revolutionizing clinical diagnosis, predictive medicine, and decision-making. This transformative technology uses machine learning, natural language processing, and large language models (LLMs) to process and reason like human intelligence. OpenAI's ChatGPT, a sophisticated LLM, holds immense potential in medical practice, research, and education. However, as AI in healthcare gains momentum, it brings forth profound ethical challenges that demand careful consideration. This comprehensive review explores key ethical concerns in the domain, including privacy, transparency, trust, responsibility, bias, and data quality. Protecting patient privacy in data-driven healthcare is crucial, with potential implications for psychological well-being and data sharing. Strategies like homomorphic encryption (HE) and secure multiparty computation (SMPC) are vital to preserving confidentiality. Transparency and trustworthiness of AI systems are essential, particularly in high-risk decision-making scenarios. Explainable AI (XAI) emerges as a critical aspect, ensuring a clear understanding of AI-generated predictions. Cybersecurity becomes a pressing concern as AI's complexity creates vulnerabilities for potential breaches. Determining responsibility in AI-driven outcomes raises important questions, with debates on AI's moral agency and human accountability. Shifting from data ownership to data stewardship enables responsible data management in compliance with regulations. Addressing bias in healthcare data is crucial to avoid AI-driven inequities. Biases present in data collection and algorithm development can perpetuate healthcare disparities. A public-health approach is advocated to address inequalities and promote diversity in AI research and the workforce. Maintaining data quality is imperative in AI applications, with convolutional neural networks showing promise in multi-input/mixed data models, offering a comprehensive patient perspective. In this ever-evolving landscape, it is imperative to adopt a multidimensional approach involving policymakers, developers, healthcare practitioners, and patients to mitigate ethical concerns. By understanding and addressing these challenges, we can harness the full potential of AI in healthcare while ensuring ethical and equitable outcomes.

Dendrogram of transparent feature importance machine learning statistics to classify associations for heart failure: A reanalysis of a retrospective cohort study of the Medical Information Mart for Intensive Care III (MIMIC-III) database

BACKGROUND

There is a continual push for developing accurate predictors for Intensive Care Unit (ICU) admitted heart failure (HF) patients and in-hospital mortality.

OBJECTIVE

The study aimed to utilize transparent machine learning and create hierarchical clustering of key predictors based off of model importance statistics gain, cover, and frequency.

METHODS

Inclusion criteria of complete patient information for in-hospital mortality in the ICU with HF from the MIMIC-III database were randomly divided into a training (n = 941, 80%) and test (n = 235, 20%). A grid search was set to find hyperparameters. Machine Learning with XGBoost were used to predict mortality followed by feature importance with Shapely Additive Explanations (SHAP) and hierarchical clustering of model metrics with a dendrogram and heat map.

RESULTS

Of the 1,176 heart failure ICU patients that met inclusion criteria for the study, 558 (47.5%) were males. The mean age was 74.05 (SD = 12.85). XGBoost model had an area under the receiver operator curve of 0.662. The highest overall SHAP explanations were urine output, leukocytes, bicarbonate, and platelets. Average urine output was 1899.28 (SD = 1272.36) mL/day with the hospital mortality group having 1345.97 (SD = 1136.58) mL/day and the group without hospital mortality having 1986.91 (SD = 1271.16) mL/day. The average leukocyte count in the cohort was 10.72 (SD = 5.23) cells per microliter. For the hospital mortality group the leukocyte count was 13.47 (SD = 7.42) cells per microliter and for the group without hospital mortality the leukocyte count was 10.28 (SD = 4.66) cells per microliter. The average bicarbonate value was 26.91 (SD = 5.17) mEq/L. Amongst the group with hospital mortality the average bicarbonate value was 24.00 (SD = 5.42) mEq/L. Amongst the group without hospital mortality the average bicarbonate value was 27.37 (SD = 4.98) mEq/L. The average platelet value was 241.52 platelets per microliter. For the group with hospital mortality the average platelet value was 216.21 platelets per microliter. For the group without hospital mortality the average platelet value was 245.47 platelets per microliter. Cluster 1 of the dendrogram grouped the temperature, platelets, urine output, Saturation of partial pressure of Oxygen (SPO2), Leukocyte count, lymphocyte count, bicarbonate, anion gap, respiratory rate, PCO2, BMI, and age as most similar in having the highest aggregate gain, cover, and frequency metrics.

CONCLUSION

Machine Learning models that incorporate dendrograms and heat maps can offer additional summaries of model statistics in differentiating factors between in patient ICU mortality in heart failure patients.

Translating Intersectionality to Fair Machine Learning in Health Sciences

Fairness approaches in machine learning should involve more than assessment of performance metrics across groups. Shifting the focus away from model metrics, we reframe fairness through the lens of intersectionality, a Black feminist theoretical framework that contextualizes individuals in interacting systems of power and oppression.

Survival prediction of heart failure patients using motion-based analysis method

BACKGROUND AND OBJECTIVE

Survival prediction of heart failure patients is critical to improve the prognostic management of the cardiovascular disease. The existing survival prediction methods focus on the clinical information while lacking the cardiac motion information. we propose a motion-based analysis method to predict the survival risk of heart failure patients for aiding clinical diagnosis and treatment.

METHODS

We propose a motion-based analysis method for survival prediction of heart failure patients. First, our method proposes the hierarchical spatial-temporal structure to capture the myocardial border. It promotes the model discrimination on border features. Second, our method explores the dense optical flow structure to capture motion fields. It improves the tracking capability on cardiac images. The cardiac motion information is obtained by fusing boundary information and motion fields of cardiac images. Finally, our method proposes the multi-modality deep-cox structure to predict the survival risk of heart failure patients. It improves the survival probability of heart failure patients.

RESULTS

The motion-based analysis method is confirmed to be able to improve the survival prediction of heart failure patients. The precision, recall, F1-score, and C-index are 0.8519, 0.8333, 0.8425, and 0.8478, respectively, which is superior to other state-of-the-art methods.

CONCLUSIONS

The experimental results show that the proposed model can effectively predict survival risk of heart failure patients. It facilitates the application of robust clinical treatment strategies.

Achieving Equity in Hospital Performance Assessments Using Composite Race-Specific Measures of Risk-Standardized Readmission and Mortality Rates for Heart Failure

BACKGROUND

The contemporary measures of hospital performance for heart failure hospitalization and 30-day risk-standardized readmission rate (RSRR) and risk-standardized mortality rate (RSMR) are estimated using the same risk adjustment model and overall event rate for all patients. Thus, these measures are mainly driven by the care quality and outcomes for the majority racial and ethnic group, and may not adequately represent the hospital performance for patients of Black and other races.

METHODS

Fee-for-service Medicare beneficiaries from January 2014 to December 2019 hospitalized with heart failure were identified. Hospital-level 30-day RSRR and RSMR were estimated using the traditional race-agnostic models and the race-specific approach. The composite race-specific performance metric was calculated as the average of the RSRR/RMSR measures derived separately for each race and ethnicity group. Correlation and concordance in hospital performance for all patients and patients of Black and other races were assessed using the composite race-specific and race-agnostic metrics.

RESULTS

The study included 1 903 232 patients (75.7% White [n=1 439 958]; 14.5% Black [n=276 684]; and 9.8% other races [n=186 590]) with heart failure from 1860 hospitals. There was a modest correlation between hospital-level 30-day performance metrics for patients of White versus Black race (Pearson correlation coefficient: RSRR=0.42; RSMR=0.26). Compared with the race-agnostic RSRR and RSMR, composite race-specific metrics for all patients demonstrated stronger correlation with RSRR (correlation coefficient: 0.60 versus 0.74) and RSMR (correlation coefficient: 0.44 versus 0.51) for Black patients. Concordance in hospital performance for all patients and patients of Black race was also higher with race-specific (versus race-agnostic) metrics (RSRR=64% versus 53% concordantly high-performing; 61% versus 51% concordantly low-performing). Race-specific RSRR and RSMR metrics (versus race-agnostic) led to reclassification in performance ranking of 35.8% and 39.2% of hospitals, respectively, with better 30-day and 1-year outcomes for patients of all race groups at hospitals reclassified as high-performing.

CONCLUSIONS

Among patients hospitalized with heart failure, race-specific 30-day RSMR and RSRR are more equitable in representing hospital performance for patients of Black and other races.

Machine learning-based in-hospital mortality risk prediction tool for intensive care unit patients with heart failure

Objective

Risk stratification of patients with congestive heart failure (HF) is vital in clinical practice. The aim of this study was to construct a machine learning model to predict the in-hospital all-cause mortality for intensive care unit (ICU) patients with HF.

Methods

eXtreme Gradient Boosting algorithm (XGBoost) was used to construct a new prediction model (XGBoost model) from the Medical Information Mart for Intensive Care IV database (MIMIC-IV) (training set). The eICU Collaborative Research Database dataset (eICU-CRD) was used for the external validation (test set). The XGBoost model performance was compared with a logistic regression model and an existing model (Get with the guideline-Heart Failure model) for mortality in the test set. Area under the receiver operating characteristic cure and Brier score were employed to evaluate the discrimination and the calibration of the three models. The SHapley Additive exPlanations (SHAP) value was applied to explain XGBoost model and calculate the importance of its features.

Results

The total of 11,156 and 9,837 patients with congestive HF from the training set and test set, respectively, were included in the study. In-hospital all-cause mortality occurred in 13.3% (1,484/11,156) and 13.4% (1,319/9,837) of patients, respectively. In the training set, of 17 features with the highest predictive value were selected into the models with LASSO regression. Acute Physiology Score III (APS III), age and Sequential Organ Failure Assessment (SOFA) were strongest predictors in SHAP. In the external validation, the XGBoost model performance was superior to that of conventional risk predictive methods, with an area under the curve of 0.771 (95% confidence interval, 0.757–0.784) and a Brier score of 0.100. In the evaluation of clinical effectiveness, the machine learning model brought a positive net benefit in the threshold probability of 0%–90%, prompting evident competitiveness compare to the other two models. This model has been translated into an online calculator which is accessible freely to the public (https://nkuwangkai-app-for-mortality-prediction-app-a8mhkf.streamlit.app).

Conclusion

This study developed a valuable machine learning risk stratification tool to accurately assess and stratify the risk of in-hospital all-cause mortality in ICU patients with congestive HF. This model was translated into a web-based calculator which access freely.