Scalable risk stratification of undiagnosed heart failure using routine health data and its association with imaging phenotypes and outcomes.

Risk-Guided Screening for Atrial Fibrillation Using Electronic Health Records

The EHR-based machine learning model, FIND-AF 2.0, identifies a high-risk subpopulation for AF diagnosis among patients at elevated risk of stroke and could enable scalable, EHR-driven, risk-guided AF screening.

Association of Charlson Comorbidity Index with Clinical Characteristics and Outcomes in Heart Failure.

BACKGROUND The Charlson comorbidity index (CCI) is a validated weighted measure of comorbidity burden, but has an unclear role in heart failure (HF). OBJECTIVES To characterize age-adjusted CCI (ACCI) in HF, its association with the use of guideline-directed medical therapy (GDMT) and with outcomes. METHODS In the Swedish HF registry, patients were divided into 3 groups based on ACCI scores (low 1-3, intermediate 4-6, and high ≥7). We studied its association with clinical characteristics and GDMT use. Multivariable multinominal and Cox regressions were used to analyze ACCI and its association with ejection fraction (EF) category (reduced [HFrEF], mildly reduced [HFmrEF], preserved [HFpEF]), and with outcomes up to 3 years. RESULTS Among 117,419 patients (age 75 [66-82], 36% women, 53% HFrEF, 24% HFmrEF, 23% HFpEF, median ACCI was 5, 6 and 6 respectively. ACCI ≥7 was present in 31% with HFrEF, 35% with HFmrEF and 42% with HFpEF. A higher ACCI score was associated with higher EF category, NYHA class, serum NT-proBNP, diuretic use, and lower quality of life and GDMT use (p<0.05 for all). The risk of composite CV mortality/first HF hospitalization was higher with intermediate (adjusted HR 1.34, 95% CI 1.29-1.39) and high vs low ACCI (1.76, 1.69-1.83). High ACCI category (vs low) was significantly associated with non-CV (adjusted HR 6.36, 95% CI 5.75-7.04), but also, CV mortality (4.35, 3.99-4.74), and total hospitalizations (adjusted IRR 2.16, 2.11-2.22) but also HF hospitalizations 1.23 (1.17-1.28). CONCLUSION The ACCI discriminates well in HF, with higher ACCI being associated with greater severity of HF, lesser GDMT use and worse quality of life. Higher ACCI was more strongly associated with death than hospitalization. Higher ACCI was more strongly associated with non-CV outcomes but also with CV and HF outcomes.

iScience Development and external validation of a risk stratification model for stroke-associated pneumonia

Frailty Attenuates the Prognostic Value of Multimorbidity in Heart Failure - Insights From the Largest Tertiary Center in the Middle East.

OBJECTIVE To evaluate whether frailty modifies the association of multimorbidity in HF. METHODS A retrospective cohort study of HF patients. Multimorbidity was quantified using 26 conditions and categorized as low (≤4), intermediate (5-7), or high (>8). Frailty was defined using a multimodal construct integrating functional, nutritional, and biochemical domains. All-cause mortality was assessed using cox models with frailty-multimorbidity interaction testing. Charlson Comorbidity Index (CCI) was used for validation. RESULTS Among 18,058 patients (median age 74 years; 62% male; 64% frail), 57% died over a median follow-up of 4.0 years (IQR 1.3-7.7). Multimorbidity burden comprised 30.5% low, 48% intermediate, and 21.5% high. Frailty was associated with a 91% higher mortality risk (adjusted HR 1.91; 95% CI 1.85-2.00). Compared with low multimorbidity, intermediate and high burden were associated with an independent 8% and 20% higher mortality risks, respectively (95% CI 1.02-1.14 and 1.13-1.27; both p<.001). However, the association of multimorbidity differed by frailty such that among non-frail patients, intermediate and high multimorbidity increased mortality by 18% and 42% whereas in frail patients, the association was attenuated, with no excess risk for intermediate burden and only 13% increase for high burden (p-for-interaction <.001). CCI analyses was consistent, though differed among females and HFpEF. CONCLUSION Frailty fundamentally attenuates risk in HF, while multimorbidity stratifies prognosis only when physiological reserve is preserved.

External Validation of the PREVENT Equations in a National Sample of US Adults.

BACKGROUND The American Heart Association Predicting Risk of Cardiovascular Disease EVENT (PREVENT) equations were developed from observational research cohorts and electronic health record data and provide sex-specific risk estimates for cardiovascular disease (CVD), atherosclerotic CVD (ASCVD), and heart failure (HF). External validation in large contemporary samples across multiple health systems in the United States is needed. METHODS We assembled a national electronic health record-based cohort of US adults with individual-level patient data pooled from a collective of 30 health systems (Truveta) to externally validate the outcome-specific 10-year PREVENT equations (PREVENT-CVD, PREVENT-ASCVD, and PREVENT-HF). We included patients aged 30 to 79 years without a history of prior CVD and with an ambulatory encounter in the electronic health record between 2013 and 2018. The outcomes were defined as total CVD (composite of ASCVD and HF), ASCVD, and HF through December 2024 using diagnosis codes. Model performance of the outcome-specific PREVENT base equations was assessed with the Harrell C statistic and calibration slope, stratified by sex. RESULTS Of the 680 864 adults included, the mean (SD) age was 55 (13) years, and 56% were female. Over a mean (SD) follow-up of 6.8 (2.3) years, there were 29 535 incident CVD events, 19 280 incident ASCVD events, and 16 824 incident HF events. The median (interquartile range) 10-year predicted risk of PREVENT-CVD among women was 3.6% (1.3%-8.8%), and among men was 5.8% (2.5%-11.7%). The C statistic (95% CI) was 0.788 (0.786-0.790), and the calibration slope (95% CI) was 0.98 (0.95-1.01) for PREVENT-CVD. PREVENT-ASCVD and PREVENT-HF demonstrated similar C statistics (0.774 [0.771-0.777] and 0.824 [0.820-0.828]) and calibration slopes (1.07 [1.04-1.10] and 1.01 [0.97-1.04]) for prediction of the 10-year risk of ASCVD and HF, respectively. CONCLUSIONS The PREVENT equations accurately and precisely estimate the 10-year risk of CVD, ASCVD, and HF in a large sample of US adults. These findings support the generalizability of the PREVENT equations to inform guideline-recommended risk assessment and preventive efforts.

Development and validation of a population-based prediction model for prevalent hypertension: evidence from Qatar biobank

Hypertension is a major cause of cardiovascular morbidity and mortality globally. Despite its high burden, validated population-based models and scores for prevalent hypertension remain limited. This study aimed to develop and internally validate a diagnostic model and scores to predict prevalent hypertension among adults in Qatar. This population-based study included 6,927 adults aged ≥18 years with complete baseline data for hypertension status from the Qatar Biobank (QBB). A multivariable logistic regression model was developed using a randomly selected training cohort (70%; n  = 4,835) and internally validated in an independent testing cohort (30%; n  = 2,092). Model performance was evaluated for discrimination using the area under the receiver operating characteristic curve (AUC), calibration using calibration plots and the Hosmer–Lemeshow goodness-of-fit test, and stability via bootstrap resampling (1,000 replications). Regression coefficients were converted into simplified point-based scores to facilitate practical application. Overall, 1,097 (15.8%) participants had prevalent hypertension. The final model included predictors such as age, gender, nationality, body mass index, smoking, physical activity, sleep duration, diabetes, hyperlipidemia, and cardiovascular diseases. The model demonstrated good discrimination, with AUCs of 0.803 (95% CI: 0.78–0.83) and 0.797 (95% CI: 0.76–0.84) in the training and testing cohorts, respectively. Calibration was satisfactory, with close agreement between predicted and observed outcomes and non-significant Hosmer–Lemeshow tests in both the training ( p  = 0.102) and testing ( p  = 0.393) cohorts. Bootstrap bias across all predictors was negligible, further supporting the model's stability. A simplified 100-point predictive score was derived from the final model to support practical classification of hypertension. We developed and internally validated a population-based prediction model and simplified predictive score for prevalent hypertension in Qatar. The model demonstrated good discrimination, calibration, and stability. The model may support population-level hypertension stratification, targeted screening, and complication-preventive interventions.