Background: Atrial fibrillation (AF) is associated with substantial morbidity, especially when it goes undetected. If new-onset AF could be predicted, targeted screening could be used to find it early. We hypothesized that a deep neural network could predict new-onset AF from the resting 12-lead ECG and that this prediction may help identify those at risk of AF-related stroke. Methods: We used 1.6 M resting 12-lead digital ECG traces from 430 000 patients collected from 1984 to 2019. Deep neural networks were trained to predict new-onset AF (within 1 year) in patients without a history of AF. Performance was evaluated using areas under the receiver operating characteristic curve and precision-recall curve. We performed an incidence-free survival analysis for a period of 30 years following the ECG stratified by model predictions. To simulate real-world deployment, we trained a separate model using all ECGs before 2010 and evaluated model performance on a test set of ECGs from 2010 through 2014 that were linked to our stroke registry. We identified the patients at risk for AF-related stroke among those predicted to be high risk for AF by the model at different prediction thresholds. Results: The area under the receiver operating characteristic curve and area under the precision-recall curve were 0.85 and 0.22, respectively, for predicting new-onset AF within 1 year of an ECG. The hazard ratio for the predicted high- versus low-risk groups over a 30-year span was 7.2 (95% CI, 6.9–7.6). In a simulated deployment scenario, the model predicted new-onset AF at 1 year with a sensitivity of 69% and specificity of 81%. The number needed to screen to find 1 new case of AF was 9. This model predicted patients at high risk for new-onset AF in 62% of all patients who experienced an AF-related stroke within 3 years of the index ECG. Conclusions: Deep learning can predict new-onset AF from the 12-lead ECG in patients with no previous history of AF. This prediction may help identify patients at risk for AF-related strokes.

Several studies have associated hydroxychloroquine use with decreased risk of diabetes mellitus (diabetes) or improved glycemic control in rheumatoid arthritis patients, but the studies were small or used data from self-report. The present study sought to replicate this protective relationship in a health system using electronic health records with laboratory data and physician diagnoses. This study is a retrospective cohort of 1127 adults with newly diagnosed rheumatoid arthritis and no diabetes within the Geisinger Health System between January 1, 2003, and March 31, 2008. Patients were classified as ever users (n = 333) or never users (n = 794) of hydroxychloroquine. Incident diabetes cases were defined using 2010 American Diabetes Association criteria. The median follow-up times for the ever and never hydroxychloroquine users were 26.0 and 23.0 months, respectively (P = 0.28). The median duration of hydroxychloroquine exposure was 14.0 months. Of the 48 cases developing diabetes during observation, 3 were exposed to hydroxychloroquine at time of development and 45 were nonexposed, yielding incidence rates of 6.2 and 22.0 per 1000 per year (P = 0.03), respectively. In time-varying Cox proportional hazards regression models adjusting for sex, age, body mass index, positive rheumatoid factor and anti-cyclic citrullinated peptide antibodies, erythrocyte sedimentation rate, and nonsteroidal anti-inflammatory drug, glucocorticoid, methotrexate, and tumor necrosis factor α inhibitor use, the hazard ratio for incident diabetes among hydroxychloroquine users was 0.29 (95% confidence interval, 0.09-0.95; P = 0.04) compared with nonusers. Our findings support the potential benefit of hydroxychloroquine in attenuating the risk of diabetes in rheumatoid arthritis patients. Further work is needed to determine its potential preventive role in other groups at high risk for diabetes.

Background: Familial hypercholesterolemia (FH) is the most common cardiovascular genetic disorder and, if left untreated, is associated with increased risk of premature atherosclerotic cardiovascular disease, the leading cause of preventable death in the United States. Although FH is common, fatal, and treatable, it is underdiagnosed and undertreated due to a lack of systematic methods to identify individuals with FH and limited uptake of cascade testing. Methods and Results: This mixed-method, multi-stage study will optimize, test, and implement innovative approaches for both FH identification and cascade testing in 3 aims. To improve identification of individuals with FH, in Aim 1, we will compare and refine automated phenotype-based and genomic approaches to identify individuals likely to have FH. To improve cascade testing uptake for at-risk individuals, in Aim 2, we will use a patient-centered design thinking process to optimize and develop novel, active family communication methods. Using a prospective, observational pragmatic trial, we will assess uptake and effectiveness of each family communication method on cascade testing. Guided by an implementation science framework, in Aim 3, we will develop a comprehensive guide to identify individuals with FH. Using the Conceptual Model for Implementation Research, we will evaluate implementation outcomes including feasibility, acceptability, and perceived sustainability as well as health outcomes related to the optimized methods and tools developed in Aims 1 and 2. Conclusions: Data generated from this study will address barriers and gaps in care related to underdiagnosis of FH by developing and optimizing tools to improve FH identification and cascade testing.

Children with obesity have hypertrophic cardiac remodeling. Hypertension is common in pediatric obesity, and may independently contribute to hypertrophy. We hypothesized that both the degree of obesity and ambulatory blood pressure (ABP) would independently associate with measures of hypertrophic cardiac remodeling in children. Children, aged 8–17 years, prospectively underwent cardiovascular magnetic resonance (CMR) and ABP monitoring. Left ventricular (LV) mass indexed to height2.7 (LVMI), myocardial thickness and end-diastolic volume were quantified from a 3D LV model reconstructed from cine balanced steady state free precession images. Categories of remodeling were determined based on cutoff values for LVMI and mass/volume. Principal component analysis was used to define a “hypertrophy score” to study the continuous relationship between concentric hypertrophy and ABP. Seventy-two children were recruited, and 68 of those (37 healthy weight and 31 obese/overweight) completed both CMR and ABP monitoring. Obese/overweight children had increased LVMI (27 ± 4 vs 22 ± 3 g/m2.7, p < 0.001), myocardial thickness (5.6 ± 0.9 vs 4.9 ± 0.7 mm, p < 0.001), mass/volume (0.69 ± 0.1 vs 0.61 ± 0.06, p < 0.001), and hypertrophy score (1.1 ± 2.2 vs −0.96 ± 1.1, p < 0.001). Thirty-five percent of obese/overweight children had concentric hypertrophy. Ambulatory hypertension was observed in 26% of the obese/overweight children and none of the controls while masked hypertension was observed in 32% of the obese/overweight children and 16% of the controls. Univariate linear regression showed that BMI z-score, systolic BP (24 h, day and night), and systolic load correlated with LVMI, thickness, mass/volume and hypertrophy score, while 24 h and nighttime diastolic BP and load also correlated with thickness and mass/volume. Multivariate analysis showed body mass index z-score and systolic blood pressure were both independently associated with left ventricular mass index (β=0.54 [p < 0.001] and 0.22 [p = 0.03]), thickness (β=0.34 [p < 0.001] and 0.26 [p = 0.001]) and hypertrophy score (β=0.47 and 0.36, both p < 0.001). In children, both the degree of obesity and ambulatory blood pressures are independently associated with measures of cardiac hypertrophic remodeling, however the correlations were generally stronger for the degree of obesity. This suggests that interventions targeted at weight loss or obesity-associated co-morbidities including hypertension may be effective in reversing or preventing cardiac remodeling in obese children.

Tuberculosis (TB) preventive therapy reduces TB risk in children. However, the effectiveness of TB preventive therapy in children living in high TB burden settings is unclear.In a prospective observational community-based cohort study in Cape Town, South Africa, we assessed the effectiveness of routine TB preventive therapy in children ≤15 years of age in a high TB and HIV prevalence setting.Among 966 children (median (interquartile range) age 5.07 (2.52–8.72) years), 676 (70%) reported exposure to an adult with TB in the past 3 months and 240 out of 326 (74%) eligible children initiated isoniazid preventive therapy under programmatic guidelines. Prevalent (n=73) and incident (n=27) TB were diagnosed among 100 out of 966 (10%) children. Children who initiated isoniazid preventive therapy were 82% less likely to develop incident TB than children who did not (adjusted OR 0.18, 95% CI 0.06–0.52; p=0.0014). Risk of incident TB increased if children were <5 years of age, living with HIV, had a positiveMycobacterium tuberculosis-specific immune response or recent TB exposure. The risk of incident TB was not associated with sex orMycobacterium bovisbacille Calmette–Guérin vaccination status. Number needed to treat (NNT) was lowest in children living with HIV (NNT=15) and children <5 years of age (NNT=19) compared with children of all ages (NNT=82).In communities with high TB prevalence, TB preventive therapy substantially reduces the risk of TB among children who are <5 years of age or living with HIV, especially those with recent TB exposure or a positiveM. tuberculosis-specific immune response in the absence of disease.

Internationally adopted children often arrive from institutional settings where they have experienced medical, nutritional and psychosocial deprivation. This study uses a validated research assessment tool to prospectively assess the impact of baseline (immediately post adoption) nutritional status on fifty-eight children as measured by weight-for-age, height-for-age, weight-for-height and head circumference-for-age z scores, as a determinant of cognitive (MDI) and psychomotor development (PDI) scores longitudinally. A statistical model was developed to allow for different ages at time of initial assessment as well as variable intervals between follow up visits. The study results show that both acute and chronic measures of malnutrition significantly affect baseline developmental status as well as the rate of improvement in both MDI and PDI scores. This study contributes to the body of literature with its prospective nature, unique statistical model for longitudinal evaluation, and use of a validated assessment tool to assess outcomes.