Search

ImportanceEarly detection of atrial fibrillation (AF) may help prevent adverse cardiovascular events such as stroke. Deep learning applied to electrocardiograms (ECGs) has been successfully used for early identification of several cardiovascular diseases.ObjectiveTo determine whether deep learning models applied to outpatient ECGs in sinus rhythm can predict AF in a large and diverse patient population.Design, setting, and participantsThis prognostic study was performed on ECGs acquired from January 1, 1987, to December 31, 2022, at 6 US Veterans Affairs (VA) hospital networks and 1 large non-VA academic medical center. Participants included all outpatients with 12-lead ECGs in sinus rhythm.Main outcomes and measuresA convolutional neural network using 12-lead ECGs from 2 US VA hospital networks was trained to predict the presence of AF within 31 days of sinus rhythm ECGs. The model was tested on ECGs held out from training at the 2 VA networks as well as 4 additional VA networks and 1 large non-VA academic medical center.ResultsA total of 907 858 ECGs from patients across 6 VA sites were included in the analysis. These patients had a mean (SD) age of 62.4 (13.5) years, 6.4% were female, and 93.6% were male, with a mean (SD) CHA2DS2-VASc (congestive heart failure, hypertension, age, diabetes mellitus, prior stroke or transient ischemic attack or thromboembolism, vascular disease, age, sex category) score of 1.9 (1.6). A total of 0.2% were American Indian or Alaska Native, 2.7% were Asian, 10.7% were Black, 4.6% were Latinx, 0.7% were Native Hawaiian or Other Pacific Islander, 62.4% were White, 0.4% were of other race or ethnicity (which is not broken down into subcategories in the VA data set), and 18.4% were of unknown race or ethnicity. At the non-VA academic medical center (72 483 ECGs), the mean (SD) age was 59.5 (15.4) years and 52.5% were female, with a mean (SD) CHA2DS2-VASc score of 1.6 (1.4). A total of 0.1% were American Indian or Alaska Native, 7.9% were Asian, 9.4% were Black, 2.9% were Latinx, 0.03% were Native Hawaiian or Other Pacific Islander, 74.8% were White, 0.1% were of other race or ethnicity, and 4.7% were of unknown race or ethnicity. A deep learning model predicted the presence of AF within 31 days of a sinus rhythm ECG on held-out test ECGs at VA sites with an area under the receiver operating characteristic curve (AUROC) of 0.86 (95% CI, 0.85-0.86), accuracy of 0.78 (95% CI, 0.77-0.78), and F1 score of 0.30 (95% CI, 0.30-0.31). At the non-VA site, AUROC was 0.93 (95% CI, 0.93-0.94); accuracy, 0.87 (95% CI, 0.86-0.88); and F1 score, 0.46 (95% CI, 0.44-0.48). The model was well calibrated, with a Brier score of 0.02 across all sites. Among individuals deemed high risk by deep learning, the number needed to screen to detect a positive case of AF was 2.47 individuals for a testing sensitivity of 25% and 11.48 for 75%. Model performance was similar in patients who were Black, female, or younger than 65 years or who had CHA2DS2-VASc scores of 2 or greater.Conclusions and relevanceDeep learning of outpatient sinus rhythm ECGs predicted AF within 31 days in populations with diverse demographics and comorbidities. Similar models could be used in future AF screening efforts to reduce adverse complications associated with this disease.

ImportanceThe Centers for Medicare & Medicaid Services uses a new peer group-based payment system to compare hospital performance as part of its Hospital Readmissions Reduction Program, which classifies hospitals into quintiles based on their share of dual-eligible beneficiaries for Medicare and Medicaid. However, little is known about the association of a hospital's share of dual-eligible beneficiaries with the quality of care and outcomes for patients with heart failure (HF).ObjectiveTo evaluate the association between a hospital's proportion of patients with dual eligibility for Medicare and Medicaid and HF quality of care and outcomes.Design, setting, and participantsThis retrospective cohort study evaluated 436 196 patients hospitalized for HF using the Get With The Guidelines-Heart Failure registry from January 1, 2010, to December 31, 2017. The analysis included patients 65 years or older with available data on dual-eligibility status. Hospitals were divided into quintiles based on their share of dual-eligible patients. Quality and outcomes were analyzed using unadjusted and adjusted multivariable logistic regression models. Data analysis was performed from April 1, 2020, to January 1, 2021.Main outcomes and measuresThe primary outcome was 30-day all-cause readmission. The secondary outcomes included in-hospital mortality, 30-day HF readmissions, 30-day all-cause mortality, and HF process of care measures.ResultsA total of 436 196 hospitalized HF patients 65 years or older from 535 hospital sites were identified, with 258 995 hospitalized patients (median age, 81 years; interquartile range, 74-87 years) at 455 sites meeting the study criteria and included in the primary analysis. A total of 258 995 HF hospitalizations from 455 sites were included in the primary analysis of the study. Hospitals in the highest dual-eligibility quintile (quintile 5) tended to care for patients who were younger, were more likely to be female, belonged to racial minority groups, or were located in rural areas compared with quintile 1 sites. After multivariable adjustment, hospitals with the highest quintile of dual eligibility were associated with lower rates of key process measures, including evidence-based β-blocker prescription, measure of left ventricular function, and anticoagulation for atrial fibrillation or atrial flutter. Differences in clinical outcomes were seen with higher 30-day all-cause (adjusted odds ratio, 1.24; 95% CI, 1.14-1.35) and HF (adjusted odds ratio, 1.14; 95% CI, 1.03-1.27) readmissions in higher dual-eligible quintile 5 sites compared with quintile 1 sites. Risk-adjusted in-hospital and 30-day mortality did not significantly differ in quintile 1 vs quintile 5 hospitals.Conclusions and relevanceIn this cohort study, hospitals with a higher share of dual-eligible patients provided care with lower rates of some of the key HF quality of care process measures and with higher 30-day all-cause or HF readmissions compared with lower dual-eligibility quintile hospitals.

ImportancePublic reporting of hospitals' 30-day risk-standardized readmission rates following heart failure hospitalization and the financial penalization of hospitals with higher rates have been associated with a reduction in 30-day readmissions but have raised concerns regarding the potential for unintended consequences.ObjectiveTo examine the association of the Hospital Readmissions Reduction Program (HRRP) with readmission and mortality outcomes among patients hospitalized with heart failure within a prospective clinical registry that allows for detailed risk adjustment.Design, setting, and participantsInterrupted time-series and survival analyses of index heart failure hospitalizations were conducted from January 1, 2006, to December 31, 2014. This study included 115 245 fee-for-service Medicare beneficiaries across 416 US hospital sites participating in the American Heart Association Get With The Guidelines-Heart Failure registry. Data analysis took place from January 1, 2017, to June 8, 2017.ExposuresTime intervals related to the HRRP were before the HRRP implementation (January 1, 2006, to March 31, 2010), during the HRRP implementation (April 1, 2010, to September 30, 2012), and after the HRRP penalties went into effect (October 1, 2012, to December 31, 2014).Main outcomes and measuresRisk-adjusted 30-day and 1-year all-cause readmission and mortality rates.ResultsThe mean (SD) age of the study population (n = 115 245) was 80.5 (8.4) years, 62 927 (54.6%) were women, and 91 996 (81.3%) were white and 11 037 (9.7%) were black. The 30-day risk-adjusted readmission rate declined from 20.0% before the HRRP implementation to 18.4% in the HRRP penalties phase (hazard ratio (HR) after vs before the HRRP implementation, 0.91; 95% CI, 0.87-0.95; P