Machine Learning to Predict the Risk of Incident Heart Failure Hospitalization Among Patients With Diabetes: The WATCH-DM Risk Score

DCRM 2.0: Multispecialty practice recommendations for the management of diabetes, cardiorenal, and metabolic diseases

The spectrum of cardiorenal and metabolic diseases comprises many disorders, including obesity, type 2 diabetes (T2D), chronic kidney disease (CKD), atherosclerotic cardiovascular disease (ASCVD), heart failure (HF), dyslipidemias, hypertension, and associated comorbidities such as pulmonary diseases and metabolism dysfunction-associated steatotic liver disease and metabolism dysfunction-associated steatohepatitis (MASLD and MASH, respectively, formerly known as nonalcoholic fatty liver disease and nonalcoholic steatohepatitis [NAFLD and NASH]). Because cardiorenal and metabolic diseases share pathophysiologic pathways, two or more are often present in the same individual. Findings from recent outcome trials have demonstrated benefits of various treatments across a range of conditions, suggesting a need for practice recommendations that will guide clinicians to better manage complex conditions involving diabetes, cardiorenal, and/or metabolic (DCRM) diseases. To meet this need, we formed an international volunteer task force comprising leading cardiologists, nephrologists, endocrinologists, and primary care physicians to develop the DCRM 2.0 Practice Recommendations, an updated and expanded revision of a previously published multispecialty consensus on the comprehensive management of persons living with DCRM. The recommendations are presented as 22 separate graphics covering the essentials of management to improve general health, control cardiorenal risk factors, and manage cardiorenal and metabolic comorbidities, leading to improved patient outcomes.

Advancements in Incident Heart Failure Risk Prediction and Screening Tools

Heart failure (HF) is a major cause of mortality and morbidity in the United States that carries substantial healthcare costs. Multiple risk prediction models and strategies have been developed over the past 30 years with the aim of identifying those at high risk of developing HF and of implementing preventive therapies effectively. This review highlights recent developments in HF risk prediction tools including emerging risk factors, innovative risk prediction models, and novel screening strategies from artificial intelligence to biomarkers. These developments allow more accurate prediction, but their impact on clinical outcomes remains to be investigated. Implementation of these risk models in clinical practice is a considerable challenge, but HF risk prediction tools offer a promising opportunity to improve outcomes while maintaining value.

Electronic Health Record Alert With Heart Failure Risk and Sodium Glucose Cotransporter 2 Inhibitor Prescriptions in Diabetes: A Randomized Clinical Trial

BACKGROUND

Sodium glucose cotransporter 2 inhibitors (SGLT2i) prevent heart failure (HF) in patients with type 2 diabetes mellitus (T2DM) but prescription rates are low. The effect of an electronic health record (EHR) alert notifying providers of patients' estimated risk of developing HF on SGTL2i prescriptions is unknown.

METHODS

This was a pragmatic, randomized clinical trial that compared an EHR alert and usual care among patients with T2DM and no history of HF or SGLT2i use at a single center. The EHR alert notified providers of their patient's HF risk and recommended HF prevention strategies. Randomization was performed at the provider level across general and subspecialty internal medicine as well as family medicine outpatient clinics. The primary outcome was proportion of SGLT2i prescriptions within 30 days. Proportion of natriuretic peptide (NP) tests within 90 days was also assessed.

RESULTS

A total of 1524 patients (median age 75 years, 45% women, 23% Black) were enrolled between September 28, 2021, and April 29, 2022 from 189 outpatient clinics. SGLT2i were prescribed to 1.2% (9/780) of patients in the EHR alert group and 0% (0/744) of those in the usual care group (P value = 0.009). Natriuretic peptide testing was performed within 90 days among 10.8% (84/780) of patients in the EHR alert group and 7.3% (54/744) of patients in the usual care group (P value = 0.02).

CONCLUSIONS

In a single-center trial with low overall SGLT2i use, an EHR alert incorporating HF risk information significantly increased SGLT2i prescriptions and NP testing although the absolute rates were low.

Review of machine learning solutions for eating disorders

BACKGROUND

Eating Disorders (EDs) are one of the most complex psychiatric disorders, with significant impairment of psychological and physical health, and psychosocial functioning, and are associated with low rates of early detection, low recovery and high relapse rates. This underscores the need for better diagnostic and treatment methods.

OBJECTIVE

This narrative review explores current Machine Learning (ML) and Artificial Intelligence (AI) applications in the domain of EDs, with a specific emphasis on clinical management in treatment settings. The primary objective are to (i) decrease the knowledge gap between ED researchers and AI-practitioners, by presenting the current state-of-the-art AI applications (including models for causality) in different ED use-cases; (ii) identify limitations of these existing AI interventions and how to address them.

RESULTS

AI/ML methods have been applied in different ED use-cases, including ED risk factor identification and incidence prediction (including the analysis of social media content in the general population), diagnosis, monitoring patients and treatment response and prognosis in clinical populations. A comparative analysis of AI-techniques deployed in these use-cases have been performed, considering factors such as complexity, flexibility, functionality, explainability and adaptability to healthcare constraints.

CONCLUSION

Multiple restrictions have been identified in the existing methods in ML and Causality in terms of achieving actionable healthcare for ED, like lack of good quality and quantity of data for models to train on, while requiring models to be flexible, high-performing, yet being explainable and producing counterfactual explanations, for ensuring the fairness and trustworthiness of its decisions. We conclude that to overcome these limitations and for future AI research and application in clinical management of ED, (i) careful considerations are required with regards to AI-model selection, and (ii) joint efforts from ED researcher and patient community are essential in building better quality and quantity of dedicated ED datasets and secure AI-solution framework.

Digital tools in heart failure: addressing unmet needs

This Series paper provides an overview of digital tools in heart failure care, encompassing screening, early diagnosis, treatment initiation and optimisation, and monitoring, and the implications these tools could have for research. The current medical environment favours the implementation of digital tools in heart failure due to rapid advancements in technology and computing power, unprecedented global connectivity, and the paradigm shift towards digitisation. Despite available effective therapies for heart failure, substantial inadequacies in managing the condition have hindered improvements in patient outcomes, particularly in low-income and middle-income countries. As digital health tools continue to evolve and exert a growing influence on both clinical care and research, establishing clinical frameworks and supportive ecosystems that enable their effective use on a global scale is crucial.

DNA methylation in cardiovascular disease and heart failure: novel prediction models?

BACKGROUND

Cardiovascular diseases (CVD) affect over half a billion people worldwide and are the leading cause of global deaths. In particular, due to population aging and worldwide spreading of risk factors, the prevalence of heart failure (HF) is also increasing. HF accounts for approximately 36% of all CVD-related deaths and stands as the foremost cause of hospitalization. Patients affected by CVD or HF experience a substantial decrease in health-related quality of life compared to healthy subjects or affected by other diffused chronic diseases.

MAIN BODY

For both CVD and HF, prediction models have been developed, which utilize patient data, routine laboratory and further diagnostic tests. While some of these scores are currently used in clinical practice, there still is a need for innovative approaches to optimize CVD and HF prediction and to reduce the impact of these conditions on the global population. Epigenetic biomarkers, particularly DNA methylation (DNAm) changes, offer valuable insight for predicting risk, disease diagnosis and prognosis, and for monitoring treatment. The present work reviews current information relating DNAm, CVD and HF and discusses the use of DNAm in improving clinical risk prediction of CVD and HF as well as that of DNAm age as a proxy for cardiac aging.

CONCLUSION

DNAm biomarkers offer a valuable contribution to improving the accuracy of CV risk models. Many CpG sites have been adopted to develop specific prediction scores for CVD and HF with similar or enhanced performance on the top of existing risk measures. In the near future, integrating data from DNA methylome and other sources and advancements in new machine learning algorithms will help develop more precise and personalized risk prediction methods for CVD and HF.

Artificial Intelligence-Based Digital Biomarkers for Type 2 Diabetes: A Review

Type 2 diabetes mellitus (T2DM), a complex metabolic disorder that burdens the health care system, requires early detection and treatment. Recent strides in digital health technologies, coupled with artificial intelligence (AI), may have the potential to revolutionize T2DM screening, diagnosis of complications, and management through the development of digital biomarkers. This review provides an overview of the potential applications of AI-driven biomarkers in the context of screening, diagnosing complications, and managing patients with T2DM. The benefits of using multisensor devices to develop digital biomarkers are discussed. The summary of these findings and patterns between model architecture and sensor type are presented. In addition, we highlight the pivotal role of AI techniques in clinical intervention and implementation, encompassing clinical decision support systems, telemedicine interventions, and population health initiatives. Challenges such as data privacy, algorithm interpretability, and regulatory considerations are also highlighted, alongside future research directions to explore the use of AI-driven digital biomarkers in T2DM screening and management.

Artificial intelligence for diabetes care: current and future prospects

Artificial intelligence (AI) use in diabetes care is increasingly being explored to personalise care for people with diabetes and adapt treatments for complex presentations. However, the rapid advancement of AI also introduces challenges such as potential biases, ethical considerations, and implementation challenges in ensuring that its deployment is equitable. Ensuring inclusive and ethical developments of AI technology can empower both health-care providers and people with diabetes in managing the condition. In this Review, we explore and summarise the current and future prospects of AI across the diabetes care continuum, from enhancing screening and diagnosis to optimising treatment and predicting and managing complications.

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.

Impact of Diabetes and Glycemia on Cardiac Improvement and Adverse Events Following Mechanical Circulatory Support

BACKGROUND

Type 2 diabetes is prevalent in cardiovascular disease and contributes to excess morbidity and mortality. We sought to investigate the effect of glycemia on functional cardiac improvement, morbidity, and mortality in durable left ventricular assist device (LVAD) recipients.

METHODS AND RESULTS

Consecutive patients with an LVAD were prospectively evaluated (n=531). After excluding patients missing pre-LVAD glycated hemoglobin (HbA1c) measurements or having inadequate post-LVAD follow-up, 375 patients were studied. To assess functional cardiac improvement, we used absolute left ventricular ejection fraction change (ΔLVEF: LVEF post-LVAD-LVEF pre-LVAD). We quantified the association of pre-LVAD HbA1c with ΔLVEF as the primary outcome, and all-cause mortality and LVAD-related adverse event rates (ischemic stroke/transient ischemic attack, intracerebral hemorrhage, gastrointestinal bleeding, LVAD-related infection, device thrombosis) as secondary outcomes. Last, we assessed HbA1c differences pre- and post-LVAD. Patients with type 2 diabetes were older, more likely men suffering ischemic cardiomyopathy, and had longer heart failure duration. Pre-LVAD HbA1c was inversely associated with ΔLVEF in patients with nonischemic cardiomyopathy but not in those with ischemic cardiomyopathy, after adjusting for age, sex, heart failure duration, and left ventricular end-diastolic diameter. Pre-LVAD HbA1c was not associated with all-cause mortality, but higher pre-LVAD HbA1c was shown to increase the risk of intracerebral hemorrhage, LVAD-related infection, and device thrombosis by 3 years on LVAD support (P<0.05 for all). HbA1c decreased from 6.68±1.52% pre-LVAD to 6.11±1.33% post-LVAD (P<0.001).

CONCLUSIONS

Type 2 diabetes and pre-LVAD glycemia modify the potential for functional cardiac improvement and the risk for adverse events on LVAD support. The degree and duration of pre-LVAD glycemic control optimization to favorably affect these outcomes warrants further investigation.

Prediction of new-onset heart failure in patients with type 2 diabetes derived from ALTITUDE and CANVAS

AIM

To create and validate a prediction model to identify patients with type 2 diabetes (T2D) at high risk of new-onset heart failure (HF), including those treated with a sodium-glucose cotransporter-2 (SGLT2) inhibitor.

METHODS

A prediction model was developed from the Aliskiren Trial in Type 2 Diabetes Using Cardiorenal Endpoints (ALTITUDE), a trial in T2D patients with albuminuria or cardiovascular disease. We included 5081 patients with baseline N-terminal pro B-type natriuretic peptide (NT-proBNP) measurement and no history of HF. The model was developed using Cox regression and validated externally in the placebo arm of the Canagliflozin Cardiovascular Assessment Study (CANVAS), which included 996 participants with T2D and established cardiovascular disease or high cardiovascular risk, and in patients treated with canagliflozin.

RESULTS

ALTITUDE participants (mean age 64 ± 9.8 years) had a median serum NT-proBNP level of 157 (25th-75th percentile 70-359) pg/mL. Higher NT-proBNP level, troponin T (TnT) level and body mass index (BMI) emerged as significant and independent predictors of new-onset HF in both cohorts. The model further contained urinary albumin-to-creatinine ratio, glycated haemoglobin, age, haematocrit, and use of calcium channel blockers. A prediction model including these variables had a C-statistic of 0.828 (95% confidence interval [CI] 0.801-0.855) in ALTITUDE and 0.800 (95% CI 0.720-0.880) in CANVAS. The C-statistic of this model increased to 0.847 (95% CI 0.792-0.902) in patients after 1 year of canagliflozin treatment.

CONCLUSION

In patients with T2D, higher NT-proBNP level, TnT level and BMI are independent and externally validated predictors of new-onset HF, including patients using an SGLT2 inhibitor. This newly developed model may identify patients at high risk of new-onset HF, contributing to early recognition and possibly prevention.

Artificial Intelligence Predicts Hospitalization for Acute Heart Failure Exacerbation in Patients Undergoing Myocardial Perfusion Imaging

Heart failure (HF) is a leading cause of morbidity and mortality in the United States and worldwide, with a high associated economic burden. This study aimed to assess whether artificial intelligence models incorporating clinical, stress test, and imaging parameters could predict hospitalization for acute HF exacerbation in patients undergoing SPECT/CT myocardial perfusion imaging. Methods: The HF risk prediction model was developed using data from 4,766 patients who underwent SPECT/CT at a single center (internal cohort). The algorithm used clinical risk factors, stress variables, SPECT imaging parameters, and fully automated deep learning-generated calcium scores from attenuation CT scans. The model was trained and validated using repeated hold-out (10-fold cross-validation). External validation was conducted on a separate cohort of 2,912 patients. During a median follow-up of 1.9 y, 297 patients (6%) in the internal cohort were admitted for HF exacerbation. Results: The final model demonstrated a higher area under the receiver-operating-characteristic curve (0.87 ± 0.03) for predicting HF admissions than did stress left ventricular ejection fraction (0.73 ± 0.05, P < 0.0001) or a model developed using only clinical parameters (0.81 ± 0.04, P < 0.0001). These findings were confirmed in the external validation cohort (area under the receiver-operating-characteristic curve: 0.80 ± 0.04 for final model, 0.70 ± 0.06 for stress left ventricular ejection fraction, 0.72 ± 0.05 for clinical model; P < 0.001 for all). Conclusion: Integrating SPECT myocardial perfusion imaging into an artificial intelligence-based risk assessment algorithm improves the prediction of HF hospitalization. The proposed method could enable early interventions to prevent HF hospitalizations, leading to improved patient care and better outcomes.

Emerging Biomarkers in the Laboratory and in Practice: A Novel Approach to Diagnosing Heart Failure in Diabetes

The Biomarkers for the Diagnosis of Heart Failure in Diabetes webinar was hosted by Diabetes Technology Society on September 20, 2023, with the objective to review current evidence and management practices of biomarker screening for heart failure in people with diabetes. The webinar discussed (1) the four stages of heart failure, (2) diabetes and heart failure, (3) natriuretic peptide and troponin for diagnosing heart failure in diabetes, (4) emerging composite and investigational biomarkers for diagnosing heart failure, and (5) prevention of heart failure progression. Experts in heart failure from the fields of clinical chemistry, cardiology, and diabetology presented data about the importance of screening for heart failure as an often-unnoticed complication of people with type 1 and type 2 diabetes.

Methods to predict heart failure in diabetes patients

INTRODUCTION

Type 2 diabetes mellitus (T2DM) is one of the leading causes of cardiovascular disease and powerful predictor for new-onset heart failure (HF).

AREAS COVERED

We focus on the relevant literature covering evidence of risk stratification based on imaging predictors and circulating biomarkers to optimize approaches to preventing HF in DM patients.

EXPERT OPINION

Multiple diagnostic algorithms based on echocardiographic parameters of cardiac remodeling including global longitudinal strain/strain rate are likely to be promising approach to justify individuals at higher risk of incident HF. Signature of cardiometabolic status may justify HF risk among T2DM individuals with low levels of natriuretic peptides, which preserve their significance in HF with clinical presentation. However, diagnostic and predictive values of conventional guideline-directed biomarker HF strategy may be non-optimal in patients with obesity and T2DM. Alternative biomarkers affecting cardiac fibrosis, inflammation, myopathy, and adipose tissue dysfunction are plausible tools for improving accuracy natriuretic peptides among T2DM patients at higher HF risk. In summary, risk identification and management of the patients with T2DM with established HF require conventional biomarkers monitoring, while the role of alternative biomarker approach among patients with multiple CV and metabolic risk factors appears to be plausible tool for improving clinical outcomes.

Machine learning approach for predicting cardiovascular disease in Bangladesh: evidence from a cross-sectional study in 2023

BACKGROUND

Cardiovascular disorders (CVDs) are the leading cause of death worldwide. Lower- and middle-income countries (LMICs), such as Bangladesh, are also affected by several types of CVDs, such as heart failure and stroke. The leading cause of death in Bangladesh has recently switched from severe infections and parasitic illnesses to CVDs.

MATERIALS AND METHODS

The study dataset comprised a random sample of 391 CVD patients' medical records collected between August 2022 and April 2023 using simple random sampling. Moreover, 260 data points were collected from individuals with no CVD problems for comparison purposes. Crosstabs and chi-square tests were used to determine the association between CVD and the explanatory variables. Logistic regression, Naïve Bayes classifier, Decision Tree, AdaBoost classifier, Random Forest, Bagging Tree, and Ensemble learning classifiers were used to predict CVD. The performance evaluations encompassed accuracy, sensitivity, specificity, and area under the receiver operator characteristic (AU-ROC) curve.

RESULTS

Random Forest had the highest precision among the five techniques considered. The precision rates for the mentioned classifiers are as follows: Logistic Regression (93.67%), Naïve Bayes (94.87%), Decision Tree (96.1%), AdaBoost (94.94%), Random Forest (96.15%), and Bagging Tree (94.87%). The Random Forest classifier maintains the highest balance between correct and incorrect predictions. With 98.04% accuracy, the Random Forest classifier achieved the best precision (96.15%), robust recall (100%), and high F1 score (97.7%). In contrast, the Logistic Regression model achieved the lowest accuracy of 95.42%. Remarkably, the Random Forest classifier achieved the highest AUC value (0.989).

CONCLUSION

This research mainly focused on identifying factors that are critical in impacting patients with CVD and predicting CVD risk. It is strongly advised that the Random Forest technique be implemented in a system for predicting cardiac diseases. This research may change clinical practice by providing doctors with a new instrument to determine a patient's CVD prognosis.

Cardiovascular Disease Management With Sodium-Glucose Cotransporter-2 Inhibitors in Patients With Type 2 Diabetes: A Cardiology Primer

Patients with type 2 diabetes face an elevated risk of cardiovascular disease. This review centers on sodium-glucose cotransporter-2 (SGLT2) inhibitors, a class of drugs that, according to a growing body of evidence, may have major potential for managing cardiovascular disease in patients with type 2 diabetes. This review presents findings from multiple clinical trials suggesting that SGLT2 inhibitors can not only serve as preventive therapeutic agents but also play a role in the active management of heart failure. The discussion includes the mechanism of action of SGLT2 inhibitors, emphasizing that they enhance urinary glucose excretion, which could lead to improved glycemic control and contribute to metabolic shifts beneficial to cardiac function. Alongside these cardiometabolic effects, safety concerns and practical considerations for prescribing these agents are addressed, taking into account potential adverse effects such as genitourinary infections and diabetic ketoacidosis as well as the financial implications for patients. Despite these drawbacks, therapeutic indications for SGLT2 inhibitors continue to expand, including for kidney protection, although further research is necessary to fully understand the mechanisms driving the cardioprotective and kidney-protective effects of SGLT2 inhibitors. By synthesizing current knowledge, this review intends to inform and guide clinical decision-making, thereby enhancing cardiovascular disease outcomes in patients with type 2 diabetes.

Impact of Residential Social Deprivation on Prediction of Heart Failure in Patients With Type 2 Diabetes: External Validation and Recalibration of the WATCH-DM Score Using Real World Data

BACKGROUND

Patients with type 2 diabetes are at risk of heart failure hospitalization. As social determinants of health are rarely included in risk models, we validated and recalibrated the WATCH-DM score in a diverse patient-group using their social deprivation index (SDI).

METHODS

We identified US Veterans with type 2 diabetes without heart failure that received outpatient care during 2010 at Veterans Affairs medical centers nationwide, linked them to their SDI using residential ZIP codes and grouped them as SDI <20%, 21% to 40%, 41% to 60%, 61% to 80%, and >80% (higher values represent increased deprivation). Accounting for all-cause mortality, we obtained the incidence for heart failure hospitalization at 5 years follow-up; overall and in each SDI group. We evaluated the WATCH-DM score using the C statistic, the Greenwood Nam D'Agostino test χ2 test and calibration plots and further recalibrated the WATCH-DM score for each SDI group using a statistical correction factor.

RESULTS

In 1 065 691 studied patients (mean age 67 years, 25% Black and 6% Hispanic patients), the 5-year incidence of heart failure hospitalization was 5.39%. In SDI group 1 (least deprived) and 5 (most deprived), the 5-year heart failure hospitalization was 3.18% and 11%, respectively. The score C statistic was 0.62; WATCH-DM systematically overestimated heart failure risk in SDI groups 1 to 2 (expected/observed ratios, 1.38 and 1.36, respectively) and underestimated the heart failure risk in groups 4 to 5 (expected/observed ratios, 0.95 and 0.80, respectively). Graphical evaluation demonstrated that the recalibration of WATCH-DM using an SDI group-based correction factor improved predictive capabilities as supported by reduction in the χ2 test results (801-27 in SDI groups I; 623-23 in SDI group V).

CONCLUSIONS

Including social determinants of health to recalibrate the WATCH-DM score improved risk prediction highlighting the importance of including social determinants in future clinical risk prediction models.

Association Between Age at Diagnosis of Type 2 Diabetes and Hospitalization for Heart Failure: A Population-Based Study

BACKGROUND

The relation between age at diagnosis of type 2 diabetes (T2D) and hospitalization for heart failure (HHF) is unclear. We assessed the association between age at diagnosis of T2D and HHF.

METHODS AND RESULTS

We conducted a population-based cohort study using administrative health databases from the Canadian province of Ontario, including participants without prior heart failure. We identified people with new-onset T2D between April 1, 2005 and March 31, 2015, and matched each person with 3 diabetes-free adults, according to birth year and sex. We estimated adjusted hazard ratios (HRs) and rate ratios (RRs) for the association between age at T2D diagnosis and incident HHF, which was assessed until March 31, 2020. Among 743 053 individuals with T2D and 2 199 539 matched individuals without T2D, 126 241 incident HHF events occurred over 8.9 years. T2D was associated with a greater adjusted hazard of HHF at younger ages (eg, HR at age 30 years: 6.94 [95% CI, 6.54-7.36]) than at older ages (eg, HR at age 60 years: 2.50 [95% CI, 2.45-2.56]) relative to matched individuals. Additional adjustment for mediators (hypertension, coronary artery disease, and chronic kidney disease) marginally attenuated this relationship. Age at T2D diagnosis was associated with a greater number of HHF events relative to matched individuals at younger ages (eg, RR at age 30 years: 6.39 [95% CI, 5.76-7.08]) than at older ages (eg, RR at age 60 years: 2.65 [95% CI, 2.54-2.76]).

CONCLUSIONS

Younger age at T2D diagnosis is associated with a disproportionately elevated HHF risk relative to age-matched individuals without T2D.

Development and validation of machine learning-augmented algorithm for insulin sensitivity assessment in the community and primary care settings: a population-based study in China

Objective

Insulin plays a central role in the regulation of energy and glucose homeostasis, and insulin resistance (IR) is widely considered as the "common soil" of a cluster of cardiometabolic disorders. Assessment of insulin sensitivity is very important in preventing and treating IR-related disease. This study aims to develop and validate machine learning (ML)-augmented algorithms for insulin sensitivity assessment in the community and primary care settings.

Methods

We analyzed the data of 9358 participants over 40 years old who participated in the population-based cohort of the Hubei center of the REACTION study (Risk Evaluation of Cancers in Chinese Diabetic Individuals). Three non-ensemble algorithms and four ensemble algorithms were used to develop the models with 70 non-laboratory variables for the community and 87 (70 non-laboratory and 17 laboratory) variables for the primary care settings to screen the classifier of the state-of-the-art. The models with the best performance were further streamlined using top-ranked 5, 8, 10, 13, 15, and 20 features. Performances of these ML models were evaluated using the area under the receiver operating characteristic curve (AUROC), the area under the precision-recall curve (AUPR), and the Brier score. The Shapley additive explanation (SHAP) analysis was employed to evaluate the importance of features and interpret the models.

Results

The LightGBM models developed for the community (AUROC 0.794, AUPR 0.575, Brier score 0.145) and primary care settings (AUROC 0.867, AUPR 0.705, Brier score 0.119) achieved higher performance than the models constructed by the other six algorithms. The streamlined LightGBM models for the community (AUROC 0.791, AUPR 0.563, Brier score 0.146) and primary care settings (AUROC 0.863, AUPR 0.692, Brier score 0.124) using the 20 top-ranked variables also showed excellent performance. SHAP analysis indicated that the top-ranked features included fasting plasma glucose (FPG), waist circumference (WC), body mass index (BMI), triglycerides (TG), gender, waist-to-height ratio (WHtR), the number of daughters born, resting pulse rate (RPR), etc.

Conclusion

The ML models using the LightGBM algorithm are efficient to predict insulin sensitivity in the community and primary care settings accurately and might potentially become an efficient and practical tool for insulin sensitivity assessment in these settings.

Optimal Screening for Predicting and Preventing the Risk of Heart Failure Among Adults With Diabetes Without Atherosclerotic Cardiovascular Disease: A Pooled Cohort Analysis

BACKGROUND

The optimal approach to identify individuals with diabetes who are at a high risk for developing heart failure (HF) to inform implementation of preventive therapies is unknown, especially in those without atherosclerotic cardiovascular disease (ASCVD).

METHODS

Adults with diabetes and no HF at baseline from 7 community-based cohorts were included. Participants without ASCVD who were at high risk for developing HF were identified using 1-step screening strategies: risk score (WATCH-DM [Weight, Age, Hypertension, Creatinine, HDL-C, Diabetes Control, QRS Duration, MI, and CABG] ≥12), NT-proBNP (N-terminal pro-B-type natriuretic peptide ≥125 pg/mL), hs-cTn (high-sensitivity cardiac troponin T ≥14 ng/L; hs-cTnI ≥31 ng/L), and echocardiography-based diabetic cardiomyopathy (echo-DbCM; left atrial enlargement, left ventricular hypertrophy, or diastolic dysfunction). High-risk participants were also identified using 2-step screening strategies with a second test to identify residual risk among those deemed low risk by the first test: WATCH-DM/NT-proBNP, NT-proBNP/hs-cTn, NT-proBNP/echo-DbCM. Across screening strategies, the proportion of HF events identified, 5-year number needed to treat and number needed to screen to prevent 1 HF event with an SGLT2i (sodium-glucose cotransporter 2 inhibitor) among high-risk participants, and cost of screening were estimated.

RESULTS

The initial study cohort included 6293 participants (48.2% women), of whom 77.7% without prevalent ASCVD were evaluated with different HF screening strategies. At 5-year follow-up, 6.2% of participants without ASCVD developed incident HF. The 5-year number needed to treat to prevent 1 HF event with an SGLT2i among participants without ASCVD was 43 (95% CI, 29-72). In the cohort without ASCVD, high-risk participants identified using 1-step screening strategies had a low 5-year number needed to treat (22 for NT-proBNP to 37 for echo-DbCM). However, a substantial proportion of HF events occurred among participants identified as low risk using 1-step screening approaches (29% for echo-DbCM to 47% for hs-cTn). Two-step screening strategies captured most HF events (75-89%) in the high-risk subgroup with a comparable 5-year number needed to treat as the 1-step screening approaches (30-32). The 5-year number needed to screen to prevent 1 HF event was similar across 2-step screening strategies (45-61). However, the number of tests and associated costs were lowest for WATCH-DM/NT-proBNP ($1061) compared with other 2-step screening strategies (NT-proBNP/hs-cTn: $2894; NT-proBNP/echo-DbCM: $16 358).

CONCLUSIONS

Selective NT-proBNP testing based on the WATCH-DM score efficiently identified a high-risk primary prevention population with diabetes expected to derive marked absolute benefits from SGLT2i to prevent HF.

The WATCH-DM risk score estimates clinical outcomes in type 2 diabetic patients with heart failure with preserved ejection fraction

The coexistence of heart failure is frequent and associated with higher mortality in patients with type 2 diabetes (T2DM), and its management is a critical issue. The WATCH-DM risk score is a tool to predict heart failure in patients with type 2 diabetes mellitus (T2DM). We investigated whether it could estimate outcomes in T2DM patients with heart failure with preserved ejection fraction (HFpEF). The WATCH-DM risk score was calculated in 418 patients with T2DM hospitalized for HFpEF (male 49.5%, age 80 ± 9 years, HbA1c 6.8 ± 1.0%), and they were divided into the "average or lower" (≤ 10 points), "high" (11-13 points) and "very high" (≥ 14 points) risk groups. We followed patients to observe all-cause death for 386 days (median). We compared the area under the curve (AUC) of the WATCH-DM score for predicting 1-year mortality with that of the Meta-Analysis Global Group in Chronic Heart Failure (MAGGIC) score and of the Barcelona Bio-Heart Failure Risk (BCN Bio-HF). Among the study patients, 108 patients (25.8%) had average or lower risk scores, 147 patients (35.2%) had high risk scores, and 163 patients (39.0%) had very high risk scores. The Cox proportional hazard model selected the WATCH-DM score as an independent predictor of all-cause death (HR per unit 1.10, 95% CI 1.03 to 1.19), and the "average or lower" risk group had lower mortality than the other groups (p = 0.047 by log-rank test). The AUC of the WATCH-DM for 1-year mortality was 0.64 (95% CI 0.45 to 0.74), which was not different from that of the MAGGIC score (0.72, 95% CI 0.63 to 0.80, p = 0.08) or that of BCN Bio-HF (0.70, 0.61 to 0.80, p = 0.25). The WATCH-DM risk score can estimate prognosis in T2DM patients with HFpEF and can identify patients at higher risk of mortality.

Machine learning identification of risk factors for heart failure in patients with diabetes mellitus with metabolic dysfunction associated steatotic liver disease (MASLD): the Silesia Diabetes-Heart Project

BACKGROUND

Diabetes mellitus (DM), heart failure (HF) and metabolic dysfunction associated steatotic liver disease (MASLD) are overlapping diseases of increasing prevalence. Because there are still high numbers of patients with HF who are undiagnosed and untreated, there is a need for improving efforts to better identify HF in patients with DM with or without MASLD. This study aims to develop machine learning (ML) models for assessing the risk of the HF occurrence in patients with DM with and without MASLD.

RESEARCH DESIGN AND METHODS

In the Silesia Diabetes-Heart Project (NCT05626413), patients with DM with and without MASLD were analyzed to identify the most important HF risk factors with the use of a ML approach. The multiple logistic regression (MLR) classifier exploiting the most discriminative patient's parameters selected by the χ2 test following the Monte Carlo strategy was implemented. The classification capabilities of the ML models were quantified using sensitivity, specificity, and the percentage of correctly classified (CC) high- and low-risk patients.

RESULTS

We studied 2000 patients with DM (mean age 58.85 ± SD 17.37 years; 48% women). In the feature selection process, we identified 5 parameters: age, type of DM, atrial fibrillation (AF), hyperuricemia and estimated glomerular filtration rate (eGFR). In the case of MASLD( +) patients, the same criterion was met by 3 features: AF, hyperuricemia and eGFR, and for MASLD(-) patients, by 2 features: age and eGFR. Amongst all patients, sensitivity and specificity were 0.81 and 0.70, respectively, with the area under the receiver operating curve (AUC) of 0.84 (95% CI 0.82-0.86).

CONCLUSION

A ML approach demonstrated high performance in identifying HF in patients with DM independently of their MASLD status, as well as both in patients with and without MASLD based on easy-to-obtain patient parameters.

Deep learning-based prognostic model using non-enhanced cardiac cine MRI for outcome prediction in patients with heart failure

OBJECTIVES

To evaluate the performance of a deep learning-based multi-source model for survival prediction and risk stratification in patients with heart failure.

METHODS

Patients with heart failure with reduced ejection fraction (HFrEF) who underwent cardiac magnetic resonance between January 2015 and April 2020 were retrospectively included in this study. Baseline electronic health record data, including clinical demographic information, laboratory data, and electrocardiographic information, were collected. Short-axis non-contrast cine images of the whole heart were acquired to estimate the cardiac function parameters and the motion features of the left ventricle. Model accuracy was evaluated using the Harrell's concordance index. All patients were followed up for major adverse cardiac events (MACEs), and survival prediction was assessed using Kaplan-Meier curves.

RESULTS

A total of 329 patients were evaluated (age 54 ± 14 years; men, 254) in this study. During a median follow-up period of 1041 days, 62 patients experienced MACEs and their median survival time was 495 days. When compared with conventional Cox hazard prediction models, deep learning models showed better survival prediction performance. Multi-data denoising autoencoder (DAE) model reached the concordance index of 0.8546 (95% CI: 0.7902-0.8883). Furthermore, when divided into phenogroups, the multi-data DAE model could significantly discriminate between the survival outcomes of the high-risk and low-risk groups compared with other models (p < 0.001).

CONCLUSIONS

The proposed deep learning (DL) model based on non-contrast cardiac cine magnetic resonance imaging could independently predict the outcome of patients with HFrEF and showed better prediction efficiency than conventional methods.

CLINICAL RELEVANCE STATEMENT

The proposed multi-source deep learning model based on cardiac magnetic resonance enables survival prediction in patients with heart failure.

KEY POINTS

• A multi-source deep learning model based on non-contrast cardiovascular magnetic resonance (CMR) cine images was built to make robust survival prediction in patients with heart failure. • The ground truth definition contains electronic health record data as well as DL-based motion data, and cardiac motion information is extracted by optical flow method from non-contrast CMR cine images. • The DL-based model exhibits better prognostic value and stratification performance when compared with conventional prediction models and could aid in the risk stratification in patients with HF.

Cardiovascular risk in diabetes mellitus: epidemiology, assessment and prevention

Cardiovascular diseases (CVDs) are the leading causes of morbidity and mortality in individuals with type 2 diabetes mellitus (T2DM). Secular changes in CVD outcomes have occurred over the past few decades, mainly due to a decline in the incidence of ischaemic heart disease. The onset of T2DM at a young age (<40 years), leading to a greater number of life-years lost, has also become increasingly common. Researchers are now looking beyond established risk factors in patients with T2DM towards the role of ectopic fat and, potentially, haemodynamic abnormalities in mediating important outcomes (such as heart failure). T2DM confers a wide spectrum of risk and is not necessarily a CVD risk equivalent, indicating the importance of risk assessment strategies (such as global risk scoring, consideration of risk-enhancing factors and assessment of subclinical atherosclerosis) to inform treatment. Data from epidemiological studies and clinical trials demonstrate that successful control of multiple risk factors can reduce the risk of CVD events by ≥50%; however, only ≤20% of patients achieve targets for risk factor reduction (plasma lipid levels, blood pressure, glycaemic control, body weight and non-smoking status). Improvements in composite risk factor control with lifestyle management (including a greater emphasis on weight loss interventions) and evidence-based generic and novel pharmacological therapies are therefore needed when the risk of CVD is high.

Machine learning in precision diabetes care and cardiovascular risk prediction

Artificial intelligence and machine learning are driving a paradigm shift in medicine, promising data-driven, personalized solutions for managing diabetes and the excess cardiovascular risk it poses. In this comprehensive review of machine learning applications in the care of patients with diabetes at increased cardiovascular risk, we offer a broad overview of various data-driven methods and how they may be leveraged in developing predictive models for personalized care. We review existing as well as expected artificial intelligence solutions in the context of diagnosis, prognostication, phenotyping, and treatment of diabetes and its cardiovascular complications. In addition to discussing the key properties of such models that enable their successful application in complex risk prediction, we define challenges that arise from their misuse and the role of methodological standards in overcoming these limitations. We also identify key issues in equity and bias mitigation in healthcare and discuss how the current regulatory framework should ensure the efficacy and safety of medical artificial intelligence products in transforming cardiovascular care and outcomes in diabetes.

Prognostic models for heart failure in patients with type 2 diabetes: a systematic review and meta-analysis

OBJECTIVE

To provide a systematic review, critical appraisal, assessment of performance and generalisability of all the reported prognostic models for heart failure (HF) in patients with type 2 diabetes (T2D).

METHODS

We performed a literature search in Medline, Embase, Central Register of Controlled Trials, Cochrane Database of Systematic Reviews and Scopus (from inception to July 2022) and grey literature to identify any study developing and/or validating models predicting HF applicable to patients with T2D. We extracted data on study characteristics, modelling methods and measures of performance, and we performed a random-effects meta-analysis to pool discrimination in models with multiple validation studies. We also performed a descriptive synthesis of calibration and we assessed the risk of bias and certainty of evidence (high, moderate, low).

RESULTS

Fifty-five studies reporting on 58 models were identified: (1) models developed in patients with T2D for HF prediction (n=43), (2) models predicting HF developed in non-diabetic cohorts and externally validated in patients with T2D (n=3), and (3) models originally predicting a different outcome and externally validated for HF (n=12). RECODe (C-statistic=0.75 95% CI (0.72, 0.78), 95% prediction interval (PI) (0.68, 0.81); high certainty), TRS-HFDM (C-statistic=0.75 95% CI (0.69, 0.81), 95% PI (0.58, 0.87); low certainty) and WATCH-DM (C-statistic=0.70 95% CI (0.67, 0.73), 95% PI (0.63, 0.76); moderate certainty) showed the best performance. QDiabetes-HF demonstrated also good discrimination but was externally validated only once and not meta-analysed.

CONCLUSIONS

Among the prognostic models identified, four models showed promising performance and, thus, could be implemented in current clinical practice.

A systematic comparison of machine learning algorithms to develop and validate prediction model to predict heart failure risk in middle-aged and elderly patients with periodontitis (NHANES 2009 to 2014)

Periodontitis is increasingly associated with heart failure, and the goal of this study was to develop and validate a prediction model based on machine learning algorithms for the risk of heart failure in middle-aged and elderly participants with periodontitis. We analyzed data from a total of 2876 participants with a history of periodontitis from the National Health and Nutrition Examination Survey (NHANES) 2009 to 2014, with a training set of 1980 subjects with periodontitis from the NHANES 2009 to 2012 and an external validation set of 896 subjects from the NHANES 2013 to 2014. The independent risk factors for heart failure were identified using univariate and multivariate logistic regression analysis. Machine learning algorithms such as logistic regression, k-nearest neighbor, support vector machine, random forest, gradient boosting machine, and multilayer perceptron were used on the training set to construct the models. The performance of the machine learning models was evaluated using 10-fold cross-validation on the training set and receiver operating characteristic curve (ROC) analysis in the validation set. Based on the results of univariate logistic regression and multivariate logistic regression, it was found that age, race, myocardial infarction, and diabetes mellitus status were independent predictors of the risk of heart failure in participants with periodontitis. Six machine learning models, including logistic regression, K-nearest neighbor, support vector machine, random forest, gradient boosting machine, and multilayer perceptron, were built on the training set, respectively. The area under the ROC for the 6 models was obtained using 10-fold cross-validation with values of 0 848, 0.936, 0.859, 0.889, 0.927, and 0.666, respectively. The areas under the ROC on the external validation set were 0.854, 0.949, 0.647, 0.933, 0.855, and 0.74, respectively. K-nearest neighbor model got the best prediction performance across all models. Out of 6 machine learning models, the K-nearest neighbor algorithm model performed the best. The prediction model offers early, individualized diagnosis and treatment plans and assists in identifying the risk of heart failure occurrence in middle-aged and elderly patients with periodontitis.

Improvement of a prediction model for heart failure survival through explainable artificial intelligence

Cardiovascular diseases and their associated disorder of heart failure (HF) are major causes of death globally, making it a priority for doctors to detect and predict their onset and medical consequences. Artificial Intelligence (AI) allows doctors to discover clinical indicators and enhance their diagnoses and treatments. Specifically, "eXplainable AI" (XAI) offers tools to improve the clinical prediction models that experience poor interpretability of their results. This work presents an explainability analysis and evaluation of two HF survival prediction models using a dataset that includes 299 patients who have experienced HF. The first model utilizes survival analysis, considering death events and time as target features, while the second model approaches the problem as a classification task to predict death. The model employs an optimization data workflow pipeline capable of selecting the best machine learning algorithm as well as the optimal collection of features. Moreover, different post hoc techniques have been used for the explainability analysis of the model. The main contribution of this paper is an explainability-driven approach to select the best HF survival prediction model balancing prediction performance and explainability. Therefore, the most balanced explainable prediction models are Survival Gradient Boosting model for the survival analysis and Random Forest for the classification approach with a c-index of 0.714 and balanced accuracy of 0.74 (std 0.03) respectively. The selection of features by the SCI-XAI in the two models is similar where "serum_creatinine", "ejection_fraction", and "sex" are selected in both approaches, with the addition of "diabetes" for the survival analysis model. Moreover, the application of post hoc XAI techniques also confirm common findings from both approaches by placing the "serum_creatinine" as the most relevant feature for the predicted outcome, followed by "ejection_fraction". The explainable prediction models for HF survival presented in this paper would improve the further adoption of clinical prediction models by providing doctors with insights to better understand the reasoning behind usually "black-box" AI clinical solutions and make more reasonable and data-driven decisions.

External Validation of the BRAVO Diabetes Model Using the EXSCEL Clinical Trial Data

INTRODUCTION

We have developed the Building, Relating, Assessing, and Validating Outcomes (BRAVO) diabetes model, an individual-level, discrete-time microsimulation model specifically designed for type 2 diabetes (T2D) management. This study aims to validate the model's performance when populated exclusively with a fully de-identified dataset to ensure its applicability in secure settings.

METHODS

Patient-level data from the Exenatide Study of Cardiovascular Event Lowering (EXSCEL) trial were fully de-identified by removing all identifiable information and masking numerical values (e.g., age, body mass index) within ranges to minimize the risk of re-identification. To populate the simulation, we imputed the masked numerical values using data from the National Health and Nutrition Examination Survey (NHANES). We applied the BRAVO model to the baseline data to predict 7-year study outcomes for the EXSCEL trial and assessed its discrimination power and calibration using C-statistics and Brier scores.

RESULTS

The model demonstrated acceptable discrimination and calibration in predicting the first occurrence of non-fatal myocardial infarction, non-fatal stroke, heart failure, revascularization, and all-cause mortality. Even with the fully deidentified data from the EXSCEL trial primarily presented in ranges rather than specific values, the BRAVO model exhibited robust prediction performance for diabetes complications and mortality.

CONCLUSIONS

This study demonstrates the feasibility of using the BRAVO model in settings where only fully de-identified patient-level data are available.

Rapid kidney function decline and increased risk of heart failure in patients with type 2 diabetes: findings from the ACCORD cohort : Rapid kidney function decline and heart failure in T2D

BACKGROUND

Impaired kidney function and albuminuria are associated with increased risk of heart failure (HF) in patients with type 2 diabetes (T2D). We investigated whether rapid kidney function decline over time is an additional determinant of increased HF risk in patients with T2D, independent of baseline kidney function, albuminuria, and other HF predictors.

METHODS

Included in the study were 7,539 participants in the Action to Control Cardiovascular Risk in Diabetes (ACCORD) study with baseline urinary albumin-to-creatinine ratio (UACR) data, who had completed 4 years of follow-up and had ≥ 3 eGFR measurements during that period (median eGFR/year = 1.9, IQR 1.7-3.2). The association between rapid kidney function decline (eGFR loss ≥ 5 ml/min/1.73 m2/year) and odds of HF hospitalization or HF death during the first 4 years of follow-up was estimated by logistic regression. The improvement in risk discrimination provided by adding rapid kidney function decline to other HF risk factors was evaluated as the increment in the area under the Receiving Operating Characteristics curve (ROC AUC) and integrated discrimination improvement (IDI).

RESULTS

Over 4 years of follow-up, 1,573 participants (20.9%) experienced rapid kidney function decline and 255 (3.4%) experienced a HF event. Rapid kidney function decline was associated with a ~ 3.2-fold increase in HF odds (3.23, 95% CI, 2.51-4.16, p < 0.0001), independent of baseline CVD history. This estimate was not attenuated by adjustment for potential confounders, including eGFR and UACR at baseline as well as at censoring (3.74; 95% CI 2.63-5.31). Adding rapid kidney function decline during follow-up to other clinical predictors (WATCH-DM score, eGFR, and UACR at study entry and end of follow-up) improved HF risk classification (ROC AUC = + 0.02, p = 0.027; relative IDI = + 38%, p < 0.0001).

CONCLUSIONS

In patients with T2D, rapid kidney function decline is associated with a marked increase in HF risk, independent of starting kidney function and/or albuminuria. These findings highlight the importance of serial eGFR measurements over time to improve HF risk estimation in T2D.

Predicting and preventing heart failure in type 2 diabetes

The burden of heart failure among people with type 2 diabetes is increasing globally. People with comorbid type 2 diabetes and heart failure often have worse outcomes than those with only one of these conditions-eg, higher hospitalisation and mortality rates. Therefore, it is essential to implement optimal heart failure prevention strategies for people with type 2 diabetes. A detailed understanding of the pathophysiology underlying the occurrence of heart failure in type 2 diabetes can aid clinicians in identifying relevant risk factors and lead to early interventions that can help prevent heart failure. In this Review, we discuss the pathophysiology and risk factors of heart failure in type 2 diabetes. We also review the risk assessment tools for predicting heart failure incidence in people with type 2 diabetes as well as the data from clinical trials that have assessed the efficacy of lifestyle and pharmacological interventions. Finally, we discuss the potential challenges in implementing new management approaches and offer pragmatic recommendations to help overcome these challenges.

Study on the prediction model of atherosclerotic cardiovascular disease in the rural Xinjiang population based on survival analysis

PURPOSE

With the increase in aging and cardiovascular risk factors, the morbidity and mortality of atherosclerotic cardiovascular disease (ASCVD), represented by ischemic heart disease and stroke, continue to rise in China. For better prevention and intervention, relevant guidelines recommend using predictive models for early detection of ASCVD high-risk groups. Therefore, this study aims to establish a population ASCVD prediction model in rural areas of Xinjiang using survival analysis.

METHODS

Baseline cohort data were collected from September to December 2016 and followed up till June 2022. A total of 7975 residents (4054 males and 3920 females) aged 30-74 years were included in the analysis. The data set was divided according to different genders, and the training and test sets ratio was 7:3 for different genders. A Cox regression, Lasso-Cox regression, and random survival forest (RSF) model were established in the training set. The model parameters were determined by cross-validation and parameter tuning and then verified in the training set. Traditional ASCVD prediction models (Framingham and China-PAR models) were constructed in the test set. Different models' discrimination and calibration degrees were compared to find the optimal prediction model for this population according to different genders and further analyze the risk factors of ASCVD.

RESULTS

After 5.79 years of follow-up, 873 ASCVD events with a cumulative incidence of 10.19% were found (7.57% in men and 14.44% in women). By comparing the discrimination and calibration degrees of each model, the RSF showed the best prediction performance in males and females (male: Area Under Curve (AUC) 0.791 (95%CI 0.767,0.813), C statistic 0.780 (95%CI 0.730,0.829), Brier Score (BS):0.060, female: AUC 0.759 (95%CI 0.734,0.783) C statistic was 0.737 (95%CI 0.702,0.771), BS:0.110). Age, systolic blood pressure (SBP), apolipoprotein B (APOB), Visceral Adiposity Index (VAI), hip circumference (HC), and plasma arteriosclerosis index (AIP) are important predictors of ASCVD in the rural population of Xinjiang.

CONCLUSION

The performance of the ASCVD prediction model based on the RSF algorithm is better than that based on Cox regression, Lasso-Cox, and the traditional ASCVD prediction model in the rural population of Xinjiang.

WATCH-DM risk score predicts the prognosis of diabetic phenotype patients with heart failure and preserved ejection fraction

BACKGROUND

Heart failure with preserved ejection fraction (HFpEF) is a heterogeneous syndrome. Diabetes may identify an essential phenotype that significantly affects the prognosis of these patients. The WATCH-DM risk score has been validated for predicting the risk of heart failure in outpatients with type 2 diabetes mellitus (T2DM), but its ability to predict clinical outcomes in HFpEF patients with T2DM is unknown. We aimed to assess whether this risk score could predict the prognosis of diabetic phenotype patients with heart failure and preserved ejection fraction.

METHODS

We enrolled retrospectively 414 patients with HFpEF (70.03 ± 8.654 years, 58.70% female), including 203 (49.03%) type 2 diabetics. Diabetic HFpEF patients were stratified by baseline WATCH-DM risk score.

RESULTS

Diabetic HFpEF patients exhibited a trend toward more concentric remodeling/hypertrophy than nondiabetic HFpEF patients. When analyzed as a continuous variable, per 1-point increase in the WATCH-DM risk score was associated with increased risks of all-cause death (HR 1.181), cardiovascular death (HR 1.239), any hospitalization (HR 1.082), and HF hospitalization (HR 1.097). The AUC for the WATCH-DM risk score in predicting incident cardiovascular death (0.7061, 95% CI 0.6329-0.7792) was higher than that of all-cause death, any hospitalization, or HF hospitalization.

CONCLUSIONS

As a high-risk phenotype for heart failure, diabetic HFpEF necessitates early risk stratification and specific treatment. To the best of our knowledge, the current study is the first to demonstrate that the WATCH-DM score predicts poor outcomes in diabetic HFpEF patients. Its convenience may allow for quick risk assessments in busy clinical settings.

A Metabolism-Based Interpretable Machine Learning Prediction Model for Diabetic Retinopathy Risk: A Cross-Sectional Study in Chinese Patients with Type 2 Diabetes

The burden of diabetic retinopathy (DR) is increasing, and the sensitive biomarkers of the disease were not enough. Studies have found that the metabolic profile, such as amino acid (AA) and acylcarnitine (AcylCN), in the early stages of DR patients might have changed, indicating the potential of metabolites to become new biomarkers. We are amid to construct a metabolite-based prediction model for DR risk. This study was conducted on type 2 diabetes (T2D) patients with or without DR. Logistic regression and extreme gradient boosting (XGBoost) prediction models were constructed using the traditional clinical features and the screening features, respectively. Assessing the predictive power of the models in terms of both discrimination and calibration, the optimal model was interpreted using the Shapley Additive exPlanations (SHAP) to quantify the effect of features on prediction. Finally, the XGBoost model incorporating AA and AcylCN variables had the best comprehensive evaluation (ROCAUC = 0.82, PRAUC = 0.44, Brier score = 0.09). C18 : 1OH lower than 0.04 μmol/L, C18 : 1 lower than 0.70 μmol/L, threonine higher than 27.0 μmol/L, and tyrosine lower than 36.0 μmol/L were associated with an increased risk of developing DR. Phenylalanine higher than 52.0 μmol/L was associated with a decreased risk of developing DR. In conclusion, our study mainly used AAs and AcylCNs to construct an interpretable XGBoost model to predict the risk of developing DR in T2D patients which is beneficial in identifying high-risk groups and preventing or delaying the onset of DR. In addition, our study proposed possible risk cut-off values for DR of C18 : 1OH, C18 : 1, threonine, tyrosine, and phenylalanine.

Effect of Canagliflozin on Heart Failure Hospitalization in Diabetes According to Baseline Heart Failure Risk

BACKGROUND

In the CANVAS (Canagliflozin Cardiovascular Assessment Study) program, canagliflozin reduced the risk of heart failure (HF) hospitalization among individuals with type 2 diabetes mellitus (T2DM).

OBJECTIVES

The purpose of this study was to evaluate heterogeneity in absolute and relative treatment effects of canagliflozin on HF hospitalization according to baseline HF risk as assessed by diabetes-specific HF risk scores (WATCH-DM [Weight (body mass index), Age, hyperTension, Creatinine, HDL-C, Diabetes control (fasting plasma glucose) and QRS Duration, MI and CABG] and TRS-HF_DM [TIMI Risk Score for HF in Diabetes]).

METHODS

Participants in the CANVAS trial were categorized into low, medium, and high risk for HF using the WATCH-DM score (for participants without prevalent HF) and the TRS-HF_DM score (for all participants). The outcome of interest was time to first HF hospitalization. The treatment effect of canagliflozin vs placebo for HF hospitalization was compared across risk strata.

RESULTS

Among 10,137 participants with available HF data, 1,446 (14.3%) had HF at baseline. Among participants without baseline HF, WATCH-DM risk category did not modify the treatment effect of canagliflozin (vs placebo) on HF hospitalization (P interaction = 0.56). However, the absolute and relative risk reduction with canagliflozin was numerically greater in the high-risk group (cumulative incidence, canagliflozin vs placebo: 8.1% vs 12.7%; HR: 0.62 [95% CI: 0.37-0.93]; P = 0.03; number needed to treat: 22) than in the low- and intermediate-risk groups. When overall study participants were categorized according to the TRS-HF_DM score, a statistically significant difference in the treatment effect of canagliflozin across risk strata was observed (P interaction = 0.04). Canagliflozin significantly reduced the risk of HF hospitalization by 39% in the high-risk group (HR: 0.61 [95% CI: 0.48-0.78]; P < 0.001; number needed to treat: 20) but not in the intermediate- or low-risk groups.

CONCLUSIONS

Among participants with T2DM, the WATCH-DM and TRS-HF_DM can reliably identify those at high risk for HF hospitalization and most likely to benefit from canagliflozin.

Electrocardiographic markers of atrial cardiomyopathy and risk of heart failure in the multi-ethnic study of atherosclerosis (MESA) cohort

Background

The association of electrocardiographic (ECG) markers of atrial cardiomyopathy with heart failure (HF) and its subtypes is unclear.

Methods

This analysis included 6,754 participants free of clinical cardiovascular disease (CVD), including atrial fibrillation (AF), from the Multi-Ethnic Study of Atherosclerosis. Five ECG markers of atrial cardiomyopathy (P-wave terminal force in V1 [PTFV1], deep-terminal negativity in V1 [DTNV1], P-wave duration [PWD], P-wave axis [PWA], advanced intra-atrial block [aIAB]) were derived from digitally recorded electrocardiograms. Incident HF events through 2018 were centrally adjudicated. An ejection fraction (EF) of 50% at the time of HF was used to classify HF as HF with reduced EF (HFrEF), HF with preserved EF (HFpEF), or unclassified HF. Cox proportional hazard models were used to examine the associations of markers of atrial cardiomyopathy with HF. The Lunn-McNeil method was used to compare the associations in HFrEF vs. HFpEF.

Results

413 HF events occurred over a median follow-up of 16 years. In adjusted models, abnormal PTFV1 (HR (95%CI): 1.56(1.15-2.13), abnormal PWA (HR (95%CI):1.60(1.16-2.22), aIAB (HR (95%CI):2.62(1.47-4.69), DTNPV1 (HR (95%CI): 2.99(1.63-7.33), and abnormal PWD (HR (95%CI): 1.33(1.02-1.73), were associated with increased HF risk. These associations persisted after further adjustments for intercurrent AF events. No significant differences in the strength of association of each ECG predictor with HFrEF and HFpEF were noted.

Conclusions

Atrial cardiomyopathy defined by ECG markers is associated with HF, with no differences in the strength of association between HFrEF and HFpEF. Markers of atrial Cardiomyopathy may help identify individuals at risk of developing HF.

Prediction models for major adverse cardiovascular events following ST-segment elevation myocardial infarction and subgroup-specific performance

Background

ST-segment elevation myocardial infarction (STEMI) patients are at a high residual risk of major adverse cardiovascular events (MACEs) after revascularization. Risk factors modify prognostic risk in distinct ways in different STEMI subpopulations. We developed a MACEs prediction model in patients with STEMI and examined its performance across subgroups.

Methods

Machine-learning models based on 63 clinical features were trained in patients with STEMI who underwent PCI. The best-performing model (the iPROMPT score) was further validated in an external cohort. Its predictive value and variable contribution were studied in the entire population and subgroups.

Results

Over 2.56 and 2.84 years, 5.0% and 8.33% of patients experienced MACEs in the derivation and external validation cohorts, respectively. The iPROMPT score predictors were ST-segment deviation, brain natriuretic peptide (BNP), low-density lipoprotein cholesterol (LDL-C), estimated glomerular filtration rate (eGFR), age, hemoglobin, and white blood cell (WBC) count. The iPROMPT score improved the predictive value of the existing risk score, with an increase in the area under the curve to 0.837 [95% confidence interval (CI): 0.784-0.889] in the derivation cohort and 0.730 (95% CI: 0.293-1.162) in the external validation cohort. Comparable performance was observed between subgroups. The ST-segment deviation was the most important predictor, followed by LDL-C in hypertensive patients, BNP in males, WBC count in females with diabetes mellitus, and eGFR in patients without diabetes mellitus. Hemoglobin was the top predictor in non-hypertensive patients.

Conclusion

The iPROMPT score predicts long-term MACEs following STEMI and provides insights into the pathophysiological mechanisms for subgroup differences.

Analysis of Publication Activity and Research Trends in the Field of AI Medical Applications: Network Approach

Artificial intelligence (AI) has revolutionized numerous industries, including medicine. In recent years, the integration of AI into medical practices has shown great promise in enhancing the accuracy and efficiency of diagnosing diseases, predicting patient outcomes, and personalizing treatment plans. This paper aims at the exploration of the AI-based medicine research using network approach and analysis of existing trends based on PubMed. Our findings are based on the results of PubMed search queries and analysis of the number of papers obtained by the different search queries. Our goal is to explore how are the AI-based methods used in healthcare research, which approaches and techniques are the most popular, and to discuss the potential reasoning behind the obtained results. Using analysis of the co-occurrence network constructed using VOSviewer software, we detected the main clusters of interest in AI-based healthcare research. Then, we proceeded with the thorough analysis of publication activity in various categories of medical AI research, including research on different AI-based methods applied to different types of medical data. We analyzed the results of query processing in the PubMed database over the past 5 years obtained via a specifically designed strategy for generating search queries based on the thorough selection of keywords from different categories of interest. We provide a comprehensive analysis of existing applications of AI-based methods to medical data of different modalities, including the context of various medical fields and specific diseases that carry the greatest danger to the human population.

Elevated urine albumin-to-creatinine ratio increases the risk of new-onset heart failure in patients with type 2 diabetes

BACKGROUND

Although albuminuria has been linked to heart failure in the general population, the relationship between urine albumin-to-creatinine ratio (uACR) and heart failure in type 2 diabetes patients is not well understood. We aimed to investigate the relationship between uACR and new-onset heart failure (HF) in type 2 diabetics.

METHODS

We included 9287 Chinese participants with type 2 diabetes (T2D) but no heart failure (HF) who were assessed with uACR between 2014 and 2016. The participants were divided into three groups based on their baseline uACR: normal (< 3 mg/mmol), microalbuminuria (3-30 mg/mmol), and macroalbuminuria (≥ 30 mg/mmol). The relationship between uACR and new-onset HF was studied using Cox proportional hazard models and restricted cubic spline. The area under the receiver operating characteristic curve (AUC), net reclassification improvement (NRI), and integrated discrimination improvement (IDI) were used to see if incorporating uACR into existing models could improve performance.

RESULTS

216 new-onset HF cases (2.33%) were recorded after a median follow-up of 4.05 years. When compared to normal uACR, elevated uACR was associated with a progressively increased risk of new-onset HF, ranging from microalbuminuria (adjusted HR, 2.21; 95% CI 1.59-3.06) to macroalbuminuria (adjusted HR, 6.02; 95% CI 4.11-8.80), and 1 standard deviation (SD) in ln (uACR) (adjusted HR, 1.89; 95% CI 1.68-2.13). The results were consistent across sex, estimated glomerular filtration rate, systolic blood pressure, and glycosylated hemoglobin subgroups. The addition of uACR to established HF risk models improved the HF risk prediction efficacy.

CONCLUSIONS

Increasing uACR, even below the normal range, is an independent risk factor for new-onset HF in a type 2 diabetic population. Furthermore, uACR may improve HF risk prediction in community-based T2D patients.

Evaluation of Machine Learning Methods Developed for Prediction of Diabetes Complications: A Systematic Review

BACKGROUND

With the rising prevalence of diabetes, machine learning (ML) models have been increasingly used for prediction of diabetes and its complications, due to their ability to handle large complex data sets. This study aims to evaluate the quality and performance of ML models developed to predict microvascular and macrovascular diabetes complications in an adult Type 2 diabetes population.

METHODS

A systematic review was conducted in MEDLINE®, Embase®, the Cochrane® Library, Web of Science®, and DBLP Computer Science Bibliography databases according to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) checklist. Studies that developed or validated ML prediction models for microvascular or macrovascular complications in people with Type 2 diabetes were included. Prediction performance was evaluated using area under the receiver operating characteristic curve (AUC). An AUC >0.75 indicates clearly useful discrimination performance, while a positive mean relative AUC difference indicates better comparative model performance.

RESULTS

Of 13 606 articles screened, 32 studies comprising 87 ML models were included. Neural networks (n = 15) were the most frequently utilized. Age, duration of diabetes, and body mass index were common predictors in ML models. Across predicted outcomes, 36% of the models demonstrated clearly useful discrimination. Most ML models reported positive mean relative AUC compared with non-ML methods, with random forest showing the best overall performance for microvascular and macrovascular outcomes. Majority (n = 31) of studies had high risk of bias.

CONCLUSIONS

Random forest was found to have the overall best prediction performance. Current ML prediction models remain largely exploratory, and external validation studies are required before their clinical implementation.

PROTOCOL REGISTRATION

Open Science Framework (registration number: 10.17605/OSF.IO/UP49X).

Machine Learning for Predicting Micro- and Macrovascular Complications in Individuals With Prediabetes or Diabetes: Retrospective Cohort Study

BACKGROUND

Micro- and macrovascular complications are a major burden for individuals with diabetes and can already arise in a prediabetic state. To allocate effective treatments and to possibly prevent these complications, identification of those at risk is essential.

OBJECTIVE

This study aimed to build machine learning (ML) models that predict the risk of developing a micro- or macrovascular complication in individuals with prediabetes or diabetes.

METHODS

In this study, we used electronic health records from Israel that contain information about demographics, biomarkers, medications, and disease codes; span from 2003 to 2013; and were queried to identify individuals with prediabetes or diabetes in 2008. Subsequently, we aimed to predict which of these individuals developed a micro- or macrovascular complication within the next 5 years. We included 3 microvascular complications: retinopathy, nephropathy, and neuropathy. In addition, we considered 3 macrovascular complications: peripheral vascular disease (PVD), cerebrovascular disease (CeVD), and cardiovascular disease (CVD). Complications were identified via disease codes, and, for nephropathy, the estimated glomerular filtration rate and albuminuria were considered additionally. Inclusion criteria were complete information on age and sex and on disease codes (or measurements of estimated glomerular filtration rate and albuminuria for nephropathy) until 2013 to account for patient dropout. Exclusion criteria for predicting a complication were diagnosis of this specific complication before or in 2008. In total, 105 predictors from demographics, biomarkers, medications, and disease codes were used to build the ML models. We compared 2 ML models: logistic regression and gradient-boosted decision trees (GBDTs). To explain the predictions of the GBDTs, we calculated Shapley additive explanations values.

RESULTS

Overall, 13,904 and 4259 individuals with prediabetes and diabetes, respectively, were identified in our underlying data set. For individuals with prediabetes, the areas under the receiver operating characteristic curve for logistic regression and GBDTs were, respectively, 0.657 and 0.681 (retinopathy), 0.807 and 0.815 (nephropathy), 0.727 and 0.706 (neuropathy), 0.730 and 0.727 (PVD), 0.687 and 0.693 (CeVD), and 0.707 and 0.705 (CVD); for individuals with diabetes, the areas under the receiver operating characteristic curve were, respectively, 0.673 and 0.726 (retinopathy), 0.763 and 0.775 (nephropathy), 0.745 and 0.771 (neuropathy), 0.698 and 0.715 (PVD), 0.651 and 0.646 (CeVD), and 0.686 and 0.680 (CVD). Overall, the prediction performance is comparable for logistic regression and GBDTs. The Shapley additive explanations values showed that increased levels of blood glucose, glycated hemoglobin, and serum creatinine are risk factors for microvascular complications. Age and hypertension were associated with an elevated risk for macrovascular complications.

CONCLUSIONS

Our ML models allow for an identification of individuals with prediabetes or diabetes who are at increased risk of developing micro- or macrovascular complications. The prediction performance varied across complications and target populations but was in an acceptable range for most prediction tasks.

Early intervention and intensive management of patients with diabetes, cardiorenal, and metabolic diseases

Increasing rates of obesity and diabetes have driven corresponding increases in related cardiorenal and metabolic diseases. In many patients, these conditions occur together, further increasing morbidity and mortality risks to the individual. Yet all too often, the risk factors for these disorders are not addressed promptly in clinical practice, leading to irreversible pathologic progression. To address this gap, we convened a Task Force of experts in cardiology, nephrology, endocrinology, and primary care to develop recommendations for early identification and intervention in obesity, diabetes, and other cardiorenal and metabolic diseases. The recommendations include screening and diagnosis, early interventions with lifestyle, and when and how to implement medical therapies. These recommendations are organized into primary and secondary prevention along the continuum from obesity through the metabolic syndrome, prediabetes, diabetes, hypertension, dyslipidemia, nonalcoholic fatty liver disease (NAFLD), atherosclerotic cardiovascular disease (ASCVD) and atrial fibrillation, chronic kidney disease (CKD), and heart failure (HF). The goal of early and intensive intervention is primary prevention of comorbidities or secondary prevention to decrease further worsening of disease and reduce morbidity and mortality. These efforts will reduce clinical inertia and may improve patients' well-being and adherence.

Predicting Stroke and Mortality in Mitral Regurgitation: A Machine Learning Approach

We hypothesized that an interpretable gradient boosting machine (GBM) model considering comorbidities, P-wave and echocardiographic measurements, can better predict mortality and cerebrovascular events in mitral regurgitation (MR). Patients from a tertiary center were analyzed. The GBM model was used as an interpretable statistical approach to identify the leading indicators of high-risk patients with either outcome of CVAs and all-cause mortality. A total of 706 patients were included. GBM analysis showed that age, systolic blood pressure, diastolic blood pressure, plasma albumin levels, mean P-wave duration (PWD), MR regurgitant volume, left ventricular ejection fraction (LVEF), left atrial dimension at end-systole (LADs), velocity-time integral (VTI) and effective regurgitant orifice were significant predictors of TIA/stroke. Age, sodium, urea and albumin levels, platelet count, mean PWD, LVEF, LADs, left ventricular dimension at end systole (LVDs) and VTI were significant predictors of all-cause mortality. The GBM demonstrates the best predictive performance in terms of precision, sensitivity c-statistic and F1-score compared to logistic regression, decision tree, random forest, support vector machine, and artificial neural networks. Gradient boosting model incorporating clinical data from different investigative modalities significantly improves risk prediction performance and identify key indicators for outcome prediction in MR.

Development and validation of a risk score for diabetic kidney disease prediction in type 2 diabetes patients: a machine learning approach

PURPOSE

This study aims to develop and validate a risk score to predict the occurrence of DKD in individuals with type 2 diabetes using a machine learning (ML) approach.

METHODS

By implementing Recursive Feature Elimination with Cross-Validation (RFECV) and RFE on the Diabetes Clinic of Imam Khomeini Hospital Complex (IKHC) dataset, the most critical features were identified. These features were used in the multivariate logistic regression (LR) analysis, and the discrimination and calibration of the model were evaluated. Finally, external validation of the model was assessed.

RESULTS

The development dataset included 1907 type 2 diabetic patients, 763 of whom developed DKD over 5 years. The predictive model performed well in the development dataset by implementing RFECV with the RF algorithm and considering six features (AUC: 79%). Using these features, the LR-based risk score indicated appropriate discrimination (AUC: 75.5%, 95% CI 73-78%) and acceptable calibration ([Formula: see text]= 7.44; p value = 0.49). This risk score was then used for 1543 diabetic patients in the validation dataset, including 633 patients with DKD over 5 years. The results showed sufficient discrimination (AUC: 75.8%, 95% CI 73-78%) of the risk score in the validation dataset.

CONCLUSIONS

We developed and validated a new risk score for predicting DKD via ML approach, which used common features in the periodic screening of type 2 diabetic patients that are readily available. In addition, a web-based online tool that is readily available to the public was developed to calculate the DKD risk score.

Stroke prevention in rural residents: development of a simplified risk assessment tool with artificial intelligence

BACKGROUND

Limited studies have focused on the risk assessment of stroke in rural regions. Moreover, the application of artificial intelligence in stroke risk scoring system is still insufficient. This study aims to develop a simplified and visualized risk score with good performance and convenience for rural stroke risk assessment, which is combined with a machine learning (ML) algorithm.

METHODS

Participants of the Henan Rural Cohort were enrolled in this study. The total participants (n = 38,322) were randomly split into a train set and a test set in the ratio of 7:3. An ML algorithm was used to select variables and the logistic regression was then applied to construct the scoring system. The C-statistic and the Brier score (BS) were used to evaluate the discrimination and calibration. The Framingham stroke risk profile (FSRP) and the self-reported stroke risk function (SRSRF) were chosen to be compared.

RESULTS

The Rural Stroke Risk Score (RSRS) was produced in this study, including age, drinking status, triglyceride, type 2 diabetes mellitus, hypertension, waist circumference, and family history of stroke. On validation, the C-statistic was 0.757 (95% CI 0.749-0.765) and the BS was 0.058 in the test set. In addition, the discrimination of RSRS was 6.02% and 7.34% higher than that of the FSRP and SRSRF, respectively.

CONCLUSIONS

A well-performed scoring system for assessing stroke risk in rural residents was developed in this study. This risk score would facilitate stroke screening and the prevention of cardiovascular disease in economically underdeveloped areas.

Deep Learning Using Electrocardiograms in Patients on Maintenance Dialysis

Cardiovascular morbidity and mortality occur with an extraordinarily high incidence in the hemodialysis-dependent end-stage kidney disease population. There is a clear need to improve identification of those individuals at the highest risk of cardiovascular complications in order to better target them for preventative therapies. Twelve-lead electrocardiograms are ubiquitous and use inexpensive technology that can be administered with minimal inconvenience to patients and at a minimal burden to care providers. The embedded waveforms encode significant information on the cardiovascular structure and function that might be unlocked and used to identify at-risk individuals with the use of artificial intelligence techniques like deep learning. In this review, we discuss the experience with deep learning-based analysis of electrocardiograms to identify cardiovascular abnormalities or risk and the potential to extend this to the setting of dialysis-dependent end-stage kidney disease.