Identification and risk stratification of coronary disease by artificial intelligence-enabled ECG.

Background: Atherosclerotic cardiovascular disease (ASCVD) is the leading cause of death worldwide, driven primarily by coronary artery disease (CAD). ASCVD risk estimators such as the pooled cohort equations (PCE) facilitate risk stratification and primary prevention of ASCVD but their accuracy is still suboptimal. Methods: Using deep electronic health record data from 7,116,209 patients seen at 70+ hospitals and clinics across 5 states in the USA, we developed an artificial intelligence-based electrocardiogram analysis tool (ECG-AI) to detect CAD and assessed the additive value of ECG-AI-based ASCVD risk stratification to the PCE. We created independent ECG-AI models using separate neural networks including subjects without known history of ASCVD, to identify coronary artery calcium (CAC) score ≥300 Agatston units by computed tomography, obstructive CAD by angiography or procedural intervention, and regional left ventricular akinesis in ≥1 segment by echocardiogram, as a reflection of possible prior myocardial infarction (MI). These were used to assess the utility of ECG-AI-based ASCVD risk stratification in a retrospective observational study consisting of patients with PCE scores and no prior ASCVD. The study period covered all available digitized EHR data, with the first available ECG in 1987 and the last in February 2023. Findings: ECG-AI for identifying CAC ≥300, obstructive CAD, and regional akinesis achieved area under the receiver operating characteristic (AUROC) values of 0.88, 0.85, and 0.94, respectively. An ensembled ECG-AI identified 3, 5, and 10-year risk for acute coronary events and mortality independently and additively to PCE. Hazard ratios for acute coronary events over 3-years in patients without ASCVD that tested positive on 1, 2, or 3 versus 0 disease-specific ECG-AI models at cohort entry were 2.41 (2.14-2.71), 4.23 (3.74-4.78), and 11.75 (10.2-13.52), respectively. Similar stratification was observed in cohorts stratified by PCE or age. Interpretation: ECG-AI has potential to address unmet need for accessible risk stratification in patients in whom PCE under, over, or insufficiently estimates ASCVD risk, and in whom risk assessment over time periods shorter than 10 years is desired. Funding: Anumana.

Risk assessment of post-discharge mortality among recently hospitalized Medicare heart failure patients with reduced or preserved ejection fraction.

Objective: Targeted care management for hospitalized patients with acute decompensated heart failure (ADHF) with reduced or preserved ejection fraction (HFrEF/HFpEF) who are at higher risk for post-discharge mortality may mitigate this outcome. However, identification of the most appropriate population for intervention has been challenging. This study developed predictive models to assess risk of 30 day and 1 year post-discharge all-cause mortality among Medicare patients with HFrEF or HFpEF recently hospitalized with ADHF.Methods: A retrospective study was conducted using the 100% Centers for Medicare Services fee-for-service sample with complementary Part D files. Eligible patients had an ADHF-related hospitalization and ICD-9-CM diagnosis code for systolic or diastolic heart failure between 1 January 2010 and 31 December 2014. Data partitioned into training (60%), validation (20%) and test sets (20%) were used to evaluate the three model approaches: classification and regression tree, full logistic regression, and stepwise logistic regression. Performance across models was assessed by comparing the receiver operating characteristic (ROC), cumulative lift, misclassification rate, the number of input variables and the order of selection/variable importance.Results: In the HFrEF (N = 83,000) and HFpEF (N = 123,644) cohorts, 30 day all-cause mortality rates were 6.6% and 5.5%, respectively, and 1 year all-cause mortality rates were 33.6% and 29.5%. The stepwise logistic regression models performed best across both cohorts, having good discrimination (test set ROC of 0.75 for both 30 day mortality models and 0.74 for both 1 year mortality models) and the lowest number of input variables (18-34 variables).Conclusions: Post-discharge mortality risk models for recently hospitalized Medicare patients with HFrEF or HFpEF were developed and found to have good predictive ability with ROCs of greater than or equal to 0.74 and a reasonable number of input variables. Applying this risk model may help providers and health systems identify hospitalized Medicare patients with HFrEF or HFpEF who may benefit from more targeted care management.

Speckle strain echocardiography predicts outcome in patients with heart failure with both depressed and preserved left ventricular ejection fraction.

BACKGROUND: While speckle imaging has been shown to predict outcome in patients with heart failure (HF), it remains unclear whether speckle strain predicts outcome in patients with HF with preserved ejection fraction (HFPEF). METHODS: Four hundred twenty patients with HF by Framingham criteria and either: left ventricular (LV) EF <50%, or elevated LV filling pressure by comprehensive echo Doppler study in the setting of left ventricular ejection fraction (LVEF) ≥50%, were enrolled. Speckle tracking was used to measure strain and strain rate in multiple vectors. The primary endpoint was HF hospitalization or cardiovascular death. RESULTS: Follow-up was completed in 380/420 patients (90%). The mean age was 55.7 ± 0.8 years, 191/380 (50%) were male, 319/380 (84%) were hypertensive, 183/380 (48%) were diabetic, and 152/380 (40%) had known coronary artery disease. At a mean follow-up of 369 ± 30 days, 107/380 patients (28%) reached the primary endpoint: 97 HF rehospitalizations and 10 cardiac deaths. The best univariate predictors of outcome were global longitudinal peak strain (GLPS) (χ(2) = 25.6, P < 0.001), mitral DT (χ(2) = 16.8, P < 0.001), LVEF (χ(2) = 16.7, P < 0.0001), longitudinal early diastolic strain (χ(2) = 8.7, P = 0.003), and circumferential peak strain (χ(2) = 7.9, P = 0.005). On multivariate analysis, GLPS (P < 0.0001), LVEF (P = 0.0002), and mitral DT (P = 0.005) were independent predictors of outcome. In the 100 HF patients with preserved LVEF, there were 17 events. Patients with GPLS ≤-15 had significantly better event-free survival than patients with GPLS >-15 (χ(2) = 4.1, P = 0.04), whereas LVEF did not predict event-free survival. CONCLUSION: Speckle strain echocardiography is an important predictor of outcome in HF patients with both depressed and preserved LVEF.

Usefulness of speckle tracking echocardiography in hypertensive crisis and the effect of medical treatment.

The acute impact of hypertensive crisis, and changes after treatment, on left ventricular (LV) systolic and diastolic function using comprehensive echocardiography, including speckle tracking, has not been well characterized. Thirty consecutive patients admitted to the hospital from the emergency room with hypertensive crisis underwent Doppler echocardiography at baseline and after blood pressure optimization. The mean age of the patients was 54 ± 13 years, with 19 men (63%). The most common presenting symptoms included dyspnea (70%), chest pain (43%), and altered mental status (13%). Mean systolic and diastolic blood pressures at presentation were 198 ± 12 and 122 ± 12 mm Hg, decreasing to 143 ± 15 and 77 ± 12 mm Hg (p <0.001 for both) after treatment. There was no significant change in LV ejection fraction between baseline and follow-up (48 ± 18% vs 46 ± 18%, p = 0.50); however, global longitudinal LV systolic strain (-10 ± 4% to -12 ± 4%, p = 0.01) and global systolic strain rate (-1.0 ± 0.4 vs -1.4 ± 0.6 s(-1), p = 0.01) significantly improved. Mean global early diastolic strain (-7.2 ± 4.0% to -9.4 ± 2.9%, p = 0.004) and early diastolic strain rate (0.3 ± 0.2 to 0.5 ± 0.4 s(-1), p = 0.05) also improved after treatment. On multivariate analysis, the independent predictors of LV longitudinal strain at follow-up were LV ejection fraction (p <0.001), heart rate (p = 0.005), systolic blood pressure (p = 0.04), and left atrial volume index (p = 0.05). In conclusion, as opposed to LV ejection fraction, LV systolic strain and strain rate were depressed during hypertensive crisis and significantly improved after medical treatment. LV diastolic function, assessed using conventional and speckle-tracking parameters, was also depressed and significantly improved after treatment.