Utilizing median and maximum QTc values improves prediction of breakthrough cardiac events in pediatric long QT syndrome.

INTRODUCTION: Although prior studies indicate that a QTc > 500 ms on a single baseline 12-lead electrocardiogram (ECG) is associated with significantly increased risk of arrhythmic events in long QT syndrome (LQTS), less is known about the risk of persistent QT prolongation. We sought to determine QTc persistence and its prognostic effect on breakthrough cardiac events (BCEs) among pediatric patients treated for LQTS. METHODS: We performed a retrospective analysis of 433 patients with LQTS evaluated, risk-stratified, and undergoing active guideline-based LQTS treatment between 1999 and 2019. BCEs were defined as arrhythmogenic syncope/seizure, sudden cardiac arrest (SCA), appropriate VF-terminating ICD shock, and sudden cardiac death (SCD). RESULTS: During the median follow-up of 5.5 years (interquartile range [IQR] = 3-9), 32 (7%) patients experienced a total of 129 BCEs. A maximum QTc threshold of 520 ms and median QTc threshold of 490 ms were determined to be strong predictors for BCEs. A landmark analysis controlling for age, sex, genotype, and symptomatic status demonstrated models utilizing both the median QTc and maximum QTc demonstrated the highest discriminatory value (c-statistic = 0.93-0.95). Patients in the high-risk group (median QTc > 490 ms and maximum QTc > 520 ms) had a significantly lower BCE free survival (70%-81%) when compared to patients in both medium-risk (93%-97%) and low-risk (98%-99%) groups. CONCLUSIONS: The risk of BCE among patients treated for LQTS increases not only based upon their maximum QTc, but also their median QTc (persistence of QTc prolongation). Patients with a maximum QTc > 520 ms and median QTc > 490 ms over serial 12-lead ECGs are at the highest risk of BCE while on guideline-directed medical therapy.

Management of Premature Ventricular Complexes in the Outpatient Setting.

With an aging population and an explosion in ambulatory electrocardiographic monitoring, primary care clinicians will undoubtedly see an increase in patients presenting with premature ventricular complexes (PVCs), especially given how common they are in the general population. A sizable number of patients with PVCs are asymptomatic, and these PVCs have no significant clinical implications. In contrast, PVCs can predispose patients to or are a clinical marker of an underlying cardiomyopathy, heart failure, or sudden cardiac death. Such a dichotomy can create fear in dealing with PVCs in the outpatient setting, both in immediate situations and for surveillance and follow-up. In this review, we provide a comprehensive overview of the pathophysiologic basis of PVCs, appropriate diagnostic testing, management, and prognostic considerations necessary when faced with PVCs in the outpatient clinic. We also provide a simplistic approach to help guide the initial work-up of PVCs, basic treatment strategies, and indications of when to refer for cardiovascular specialty consultation to improve physician comfort and to enhance patient care.

Cardiac fludeoxyglucose-18 positron emission tomography in genotype-positive arrhythmogenic cardiomyopathy.

BACKGROUND: Myocardial inflammation contributes to the pathogenesis of arrhythmogenic cardiomyopathy (ACM), a clinically and genetically heterogenous disorder. Due to phenotypic overlap, some patients with genetic ACM may be evaluated for an underlying inflammatory cardiomyopathy. However, the cardiac fludeoxyglucose (FDG) positron emission tomography (PET) findings in ACM patients have not been elucidated. METHODS: All genotype-positive patients in the Mayo Clinic ACM registry (n = 323) who received a cardiac FDG PET were included in this study. Pertinent data were extracted from the medical record. RESULTS: Collectively, 12/323 (4%; 67% female) genotype-positive ACM patients received a cardiac PET FDG scan as part of their clinical evaluation (median age at scan 49 ± 13 years). Amongst these patients, pathogenic/likely pathogenic variants were detected in LMNA (n = 7), DSP (n = 3), FLNC (n = 1) and PLN (n = 1). Of note, 6/12 (50%) had abnormal myocardial FDG uptake, including diffuse (entire myocardium) uptake in 2/6 (33%), focal (1-2 segments) uptake in 2/6 (33%) and patchy (3+ segments) in 2/6 (33%). Median myocardial standardized uptake value ratio was 2.1. Interestingly, LMNA-positive patients accounted for 3 out of 6 (50%) positive studies (diffuse uptake in 2 and focal uptake in 1). CONCLUSION: Abnormal myocardial FDG uptake is common in genetic ACM patients undergoing cardiac FDG PET. This study further supports the role of myocardial inflammation in ACM. Further investigation is needed to determine role of FDG PET in diagnosis and management of ACM and investigate the role of inflammation in ACM.

Limb and cardiovascular event risk in type 1 and 2 diabetic patients with peripheral artery disease.

Background: Peripheral artery disease (PAD) is a risk factor for adverse limb events (LE) and cardiovascular events (CVE) that coexists with type 1 (T1) and 2 (T2) diabetes mellitus (DM). Little is known about comparative risk of LE and CVE in T1/T2 DM patients with PAD. Patients and methods: We queried our database of 40,144 patients ≥18 years old who underwent ankle brachial index (ABI) measurement from 01/1996-02/2020. We isolated T1/T2 DM patients with PAD diagnosed by ankle brachial index (ABI; low [<1.0] or elevated [>1.4]) and retrieved demographics including glycated hemoglobin (HbA1c). Primary outcomes were LE (critical limb ischemia/vascular amputation) and CVE (myocardial infarction/ischemic stroke). All-cause mortality was a secondary outcome. Multivariable Cox proportional regression yielded hazard ratios (HR) with 95% confidence intervals (CI) after adjusting for pertinent risk factors including age, hypertension, hyperlipidemia, smoking, and HbA1c. Results: Our study found 10,156 patients with PAD and DM (34% T1DM, 66% T2DM) with median follow-up time 34 mo (IQR 85 mo). T1DM patients were younger than T2DM (mean age 67 vs. 70 years), with higher median HbA1c (7.7 [IQR 1.9] vs. 6.7% [IQR 1.6]), and more prevalent hypertension, hyperlipidemia, CAD, and CKD. Antiplatelet and statin use was equivocal. Elevated ABI was more common in T1DM (47 vs. 28%). LE occurred in 23% and CVE in 12% patients. LE risk was higher in T1 than T2 DM patients (HR 1.58 [95% CI 1.44, 1.73], p<0.0001), but CVE and all-cause mortality were equivocal. These observations were preserved across ABI and HbA1c subgroup analyses. Conclusions: PAD patients with T1DM had a higher LE risk than those with T2DM, even after adjustment for glycemic control and pertinent risk factors, but CVE risk and all-cause mortality were equivocal. These data suggest a potential role for more intensive LE risk modification in PAD patients with T1DM, but further investigation is needed.

Antiarrhythmic drugs for atrial fibrillation in the outpatient setting.

With burgeoning evidence for early rhythm control in patients with atrial fibrillation (AF), the management of AF in the outpatient setting has become more challenging. The primary care clinician often serves as the frontline in the pharmacologic management of AF. Because of drug interactions and the risk of proarrhythmia, many clinicians remain hesitant about the initiation and chronic management of antiarrhythmic drug prescriptions. However, with the likely increased use of antiarrhythmics for early rhythm control, understanding and familiarity with these medications have likewise become more important, especially since patients with AF probably have other non-cardiac medical conditions that can impact their antiarrhythmic therapy. In this comprehensive review, we provide informative, high-yield cases and edifying references that will help primary care providers become comfortable handling various clinical scenarios.

Prediction of permanent pacemaker implantation after transcatheter aortic valve replacement: The role of machine learning.

BACKGROUND: Atrioventricular block requiring permanent pacemaker (PPM) implantation is an important complication of transcatheter aortic valve replacement (TAVR). Application of machine learning could potentially be used to predict pre-procedural risk for PPM. AIM: To apply machine learning to be used to predict pre-procedural risk for PPM. METHODS: A retrospective study of 1200 patients who underwent TAVR (January 2014-December 2017) was performed. 964 patients without prior PPM were included for a 30-d analysis and 657 patients without PPM requirement through 30 d were included for a 1-year analysis. After the exclusion of variables with near-zero variance or ≥ 50% missing data, 167 variables were included in the random forest gradient boosting algorithm (GBM) optimized using 5-fold cross-validations repeated 10 times. The receiver operator curve (ROC) for the GBM model and PPM risk score models were calculated to predict the risk of PPM at 30 d and 1 year. RESULTS: Of 964 patients included in the 30-d analysis without prior PPM, 19.6% required PPM post-TAVR. The mean age of patients was 80.9 ± 8.7 years. 42.1 % were female. Of 657 patients included in the 1-year analysis, the mean age of the patients was 80.7 ± 8.2. Of those, 42.6% of patients were female and 26.7% required PPM at 1-year post-TAVR. The area under ROC to predict 30-d and 1-year risk of PPM for the GBM model (0.66 and 0.72) was superior to that of the PPM risk score (0.55 and 0.54) with a P value < 0.001. CONCLUSION: The GBM model has good discrimination and calibration in identifying patients at high risk of PPM post-TAVR.

One-year outcomes of catheter ablation for atrial fibrillation in young patients.

BACKGROUND: Atrial fibrillation (AF) is relatively less frequent in younger patients (age < 50). Recently, studies have suggested that early restoration of sinus rhythm may lead to improved outcomes compared with rate control, however the efficacy of catheter ablation for AF in young is scarce. METHODS: We included all hospitalized patients between 18 and 50 years with a diagnosis of AF from the Nationwide Readmission Database 2016-2017 from the Healthcare Cost and Utilization Project. Demographic and comorbidity data were collected and analyzed. Outcomes assessed included one-year AF readmission rates, all-cause readmission, ischemic stroke, and all-cause mortality. Subgroup analyses were performed for all demographic and comorbidity variables. RESULTS: Overall, 52,598 patients (medium age 44, interquartile range 38-48, female 25.7%) were included in the study, including 2,146 (4.0%) who underwent catheter ablation for AF. Patients who underwent catheter ablation had a significantly lower rate of readmission for AF or any cause at one year (adjusted hazard ratios (HR) of 0.52 [95% confidence interval (CI): 0.43-0.63] and HR of 0.81 [95% CI: 0.72-0.89], respectively). There was no difference in 1-year readmission for stroke or all-cause mortality between the two groups. Subgroup analyses showed a consistent reduction in the risk of AF readmission among major demographic and comorbidity subgroups. CONCLUSION: Catheter ablation in young patients with AF was associated with a reduction in 1-year AF related and all-cause readmissions. These data merit further prospective investigation for validation, through dedicated registries and multicenter collaborations to include young AF from diverse population.

An intervention strategy to improve genetic testing for dilated cardiomyopathy in a heart failure clinic.

PURPOSE: Despite its clinical implications in screening and therapy, genetic testing in dilated cardiomyopathy (DCM) is underused. This study evaluated implementing a practice intervention in a heart failure clinic to automate and streamline the process of genetic testing. METHODS: Eligible patients with DCM were compared for frequency of pretest genetic education and testing during pre- and postintervention periods. The intervention comprised automated prescheduling of a cardiovascular genomics e-consult that served as a placeholder for downstream, pretest education, testing, and post-test review of genetic results. RESULTS: Patients with DCM were more likely to undergo pretest genetic education after intervention than before intervention (33.5% vs 14.8%, P < .0001). Similarly, patients with DCM were more likely to undergo genetic testing after intervention than before intervention (27.3% vs 13.0%, P = .0006). The number of patients who were diagnosed to have likely pathogenic or pathogenic genetic variants were 2 of 21 (9.5%) and 6 of 53 (11.1%) before and after intervention, respectively, and variants were present in the following genes: FLNC, TTN, DES, LMNA, PLN, and TNNT2. CONCLUSION: An intervention strategy in a heart failure clinic to increase the rates of pretest genetic education and testing in eligible patients with DCM was feasible and efficacious and may have important implications for the management of DCM.

Peripheral artery disease and the risk of venous thromboembolism.

Background: Peripheral artery disease (PAD) impacts 3-12% of patients worldwide and is characterized by endothelial dysfunction and inflammatory pathways which are also common to venous thromboembolism (VTE), but there is a paucity of evidence regarding VTE risk in PAD patients. We investigated whether PAD is an independent risk factor for VTE. Patients and methods: We reviewed medical records of patients undergoing ABI studies at Mayo Clinic from 01/1996-02/2020. We classified patients by ABI (low [<1.0], normal [1.0-1.4], or elevated [>1.4]), as well as by specific low ABI subgroup: severely reduced (ABI: 0.00-0.39), moderately reduced (0.40-0.69), mildly reduced (0.70-0.90), and borderline reduced (0.91-0.99). The primary outcome was incident VTE event (acute lower extremity deep vein thrombosis or pulmonary embolism) after ABI measurement. Multivariable Cox proportional regression was used to calculate hazard ratios (HR) with 95% confidence intervals (CI) after adjusting for age, sex, active smoking, cancer, previous VTE, thrombophilia, anticoagulation, and revascularization. Results: 39,834 unique patients (mean age 66.3±14.3 years, median follow-up 34 months) were identified. 2,305 VTE events occurred in patients without PAD (13.0%), 2,218 in low ABI patients (13.0%), and 751 in elevated ABI patients (14.8%). After risk factor adjustment, VTE risk was modestly increased for PAD overall (HR: 1.12, 95% CI [1.06, 1.18]), including low ABI (HR: 1.11, 95% CI [1.04, 1.18]) and elevated ABI groups (HR: 1.15, 95% CI [1.04, 1.26]), compared to patients without PAD. The greatest VTE risk was in severely low ABI patients (HR: 1.46, 95% CI [1.31, 1.64]). Conclusions: In a large longitudinal cohort, we present strong clinical evidence of PAD, with low and elevated ABI, as an independent VTE risk factor, with the highest risk seen in patients with severely low ABI. Continued research is required to further investigate this relationship and its intersection with functional performance status to optimize VTE risk reduction or anticoagulation strategies in the PAD population.

Substrate Characterization and Outcomes of Catheter Ablation of Ventricular Arrhythmias in Patients With Mitral Annular Disjunction.

BACKGROUND: Mitral annular disjunction (MAD) has recently been recognized as an arrhythmogenic entity. Data on the electrophysiological substrate as well as the outcomes of catheter ablation of ventricular arrhythmias in patients with MAD is limited. METHODS: Forty patients with MAD (mean age 47±15 years; 70% female) underwent catheter ablation for ventricular arrhythmias. Detailed clinical, electrocardiographic, cardiac imaging, and procedural data were collected. Clinical outcomes were compared between patients who had substrate modification in the MAD area and those who did not. RESULTS: Twenty-three (57.5%) patients had ablation for premature ventricular contractions, 10 (25%) patients for sustained ventricular tachycardia, and 7 (17.5%) patients for premature ventricular contraction-triggered ventricular fibrillation. Mean end-systolic MAD length was 10.58±3.49 mm on transthoracic echocardiography. Seventeen (42.5%) patients had preprocedural cardiac magnetic resonance imaging, and 5 (29%) patients had late gadolinium enhancement. Among the 18 (45%) patients who had abnormal local electrograms (low voltage, long-duration, fractionated, isolated mid-diastolic potentials) during electroanatomical mapping, 10 (25%) patients had abnormal electrograms in the anterolateral mitral annulus and/or MAD area. Substrate modification was performed in 10 (25%) patients. Catheter ablation was acutely successful in 36 (90%) patients (elimination of premature ventricular contraction or noninducibility of ventricular tachycardia). After a median follow-up duration of 54.08 (interquartile range, 10.67-89.79) months, premature ventricular contraction burden decreased from a median of 9.75% (interquartile range, 3.25-14) before the ablation to a median of 4% (interquartile range, 1-7.75) after the ablation (P=0.03 [95% CI, 0.055-6.5]). Eight (20.5%) patients had repeat ablation for ventricular arrhythmias. Substrate modification of the MAD was associated with a trend toward lower rates of repeat ablation (0% versus 26.7%; P=0.16). CONCLUSIONS: Patients with MAD have a complex arrhythmogenic substrate, and catheter ablation is effective in reducing recurrence of ventricular arrhythmias. Substrate mapping and ablation may be considered in these patients.

Patient Age at Diagnosis of Peripheral Artery Disease and Its Impact on Cardiovascular and Limb Outcomes.

Peripheral artery disease (PAD) prevalence increases with age, but the relation between age at PAD diagnosis and outcomes is unclear. We investigated the cardiovascular and limb outcomes of patients diagnosed with PAD at different ages. We studied patients with PAD aged ≥18 years who were diagnosed between 1996 and 2020 at Mayo Clinic. Patients were grouped by diagnosis age (<50, 50 to 59, 60 to 69, ≥70 years) and ankle brachial index (ABI): low ABI (<1.0) or elevated ABI (>1.4). Primary outcomes were cardiovascular events (CVEs; myocardial infarction or ischemic stroke) and limb events (LEs; critical limb ischemia or amputation). Competing risk analysis was performed to calculate adjusted hazard ratios. The cohort included 22,073 patients with PAD (low ABI: 77.1%; elevated ABI: 22.9%). CVEs were observed in 8.2% of patients and LEs in 15.6%. The highest CVE risk was observed in patients diagnosed with PAD before age 50 (compared with patients diagnosed after age 70; hazard ratio 2.33 [95% confidence interval 1.95 to 2.78]). CVE risk decreased with older age at diagnosis. Although younger groups demonstrated higher LE risk, there was no clear association with diagnosis age. These patterns of risk were seen both in low and elevated ABI subgroups but in greater magnitude with elevated ABI. Younger patients diagnosed with PAD face increased risk of myocardial infarction and ischemic stroke compared with patients diagnosed at an older age. CVE risk notably exceeds LE risk. In conclusion, younger age at PAD diagnosis may be an important risk factor, which warrants more aggressive interventions focused on CVE prevention.

Silent cerebral infarcts in patients with atrial fibrillation: Clinical implications of an imaging-adjusted CHA2DS2-VASc score.

BACKGROUND: The CHA2DS2-VASc score does not include silent infarcts on neuroimaging in stroke risk estimation for patients with atrial fibrillation (AF). The inclusion of silent infarcts into CHA2DS2-VASc scoring and its impact on stroke prophylaxis recommendations in patients with AF has not been previously studied. The present study sought to quantify the prevalence of silent infarcts in patients with AF and describe potential changes in management based on magnetic resonance imaging (MRI) findings. METHODS: Participants from the Mayo Clinic Study of Aging with AF and brain MRI were included. Silent infarcts were identified. "Standard" CHA2DS2-VASc scores were calculated for each subject based on clinical history alone and "imaging-adjusted" CHA2DS2-VASc scores based on evidence of cerebral infarction on MRI. Standard and imaging-adjusted scores were compared. RESULTS: One hundred and forty-seven participants (average age 77, 28% female) were identified with AF, MRI, and no clinical history of stroke. Overall, 41 (28%) patients had silent infarcts on MRI, corresponding with a 2-point increase in CHA2DS2-VASc score. Of these participants, only 39% (16/41) with silent infarct were on anticoagulation despite having standard CHA2DS2-VASc scores supportive of anticoagulation. After incorporating silent infarcts, 13% (19/147) would have an indication for periprocedural bridging compared to 0.6% (1/147) at baseline. CONCLUSIONS: Incorporation of silent infarcts into the CHA2DS2-VASc score may change the risk- -benefit ratio of anticoagulation. It may also increase the number of patients who would benefit from periprocedural bridging. Future research should examine whether an anticoagulation strategy based on imaging-adjusted CHA2DS2-VASc scores could result in a greater reduction of stroke and cognitive decline.

A real-world experience of atrioventricular synchronous pacing with leadless ventricular pacemakers.

AIMS: The MicraTM transcatheter pacing system (TPS) (Medtronic) is the only leadless pacemaker that promotes atrioventricular (AV) synchrony via accelerometer-based atrial sensing. Data regarding the real-world experience with this novel system are scarce. We sought to characterize patients undergoing MicraTM -AV implants, describe percentage AV synchrony achieved, and analyze the causes for suboptimal AV synchrony. METHODS: In this retrospective cohort study, electronic medical records from 56 consecutive patients undergoing MicraTM -AV implants at the Mayo Clinic sites in Minnesota, Florida, and Arizona with a minimum follow-up of 3 months were reviewed. Demographic data, comorbidities, echocardiographic data, and clinical outcomes were compared among patients with and without atrial synchronous ventricular pacing (AsVP) ≥ 70%. RESULTS: Sixty-five percent of patients achieved AsVP ≥ 70%. Patients with adequate AsVP had smaller body mass indices, a lower proportion of congestive heart failure, and prior cardiac surgery. Echocardiographic parameters and procedural characteristics were similar across the two groups. Active device troubleshooting was associated with higher AsVP. The likely reasons for low AsVP were small A4-wave amplitude, high ventricular pacing burden, and inadequate device reprogramming. Importantly, in patients with low AsVP, subjective clinical worsening was not noted during follow-up. CONCLUSION: With the increasing popularity of leadless pacemakers, it is paramount for device implanting teams to be familiar with common predictors of AV synchrony and troubleshooting with MicraTM -AV devices.

Transvenous laser lead extraction in patients with congenital complete heart block.

BACKGROUND: Data on lead management in patients with congenital complete heart block (CCHB) with cardiac implantable electronic devices are lacking. OBJECTIVE: The purpose of this study was to describe the natural history and outcomes in patients with CCHB with cardiac implantable electronic devices undergoing transvenous lead extraction (TLE). METHODS: Data on all attempted TLE procedures in patients with CCHB at 2 institutions between 2011 and 2021 were collected from a retrospective registry. RESULTS: Overall, 16 patients (mean age at transvenous device implant 13.8 ± 4.7 years) were included. Before TLE, patients underwent an average of 2.25 ± 1.3 generator changes, 3 (19%) underwent cardiac resynchronization therapy upgrade, and 7 (44%) underwent a lead revision with subsequently abandoned leads. Mean patient age at TLE was 34.4 ± 9.4 years with a mean duration of lead implant of 19.2 ± 6.9 years. Lead malfunction (n = 11 [69%]) and infection (n = 5 [31%]) were the most common indications for TLE. A total of 38 leads were removed, with complete procedural success achieved in 14 of 16 (87.5%). Two (12.5%) major complications occurred, including right ventricular laceration and superior vena cava tear requiring sternotomies. All patients survived at 1-year follow-up. CONCLUSION: Patients with CCHB represent a unique cohort highlighted by several generator changes, lead revisions, and abandoned leads at a young age, along with a long duration of lead dwelling time and a high prevalence of lead malfunction requiring TLE. There may be a high risk of major complications during TLE, suggesting TLE should be performed only in experienced centers. Larger studies are needed to confirm these findings.

Classifying social determinants of health from unstructured electronic health records using deep learning-based natural language processing.

OBJECTIVE: Social determinants of health (SDOH) are non-medical factors that can profoundly impact patient health outcomes. However, SDOH are rarely available in structured electronic health record (EHR) data such as diagnosis codes, and more commonly found in unstructured narrative clinical notes. Hence, identifying social context from unstructured EHR data has become increasingly important. Yet, previous work on using natural language processing to automate extraction of SDOH from text (a) usually focuses on an ad hoc selection of SDOH, and (b) does not use the latest advances in deep learning. Our objective was to advance automatic extraction of SDOH from clinical text by (a) systematically creating a set of SDOH based on standard biomedical and psychiatric ontologies, and (b) training state-of-the-art deep neural networks to extract mentions of these SDOH from clinical notes. DESIGN: A retrospective cohort study. SETTING AND PARTICIPANTS: Data were extracted from the Medical Information Mart for Intensive Care (MIMIC-III) database. The corpus comprised 3,504 social related sentences from 2,670 clinical notes. METHODS: We developed a framework for automated classification of multiple SDOH categories. Our dataset comprised narrative clinical notes under the "Social Work" category in the MIMIC-III Clinical Database. Using standard terminologies, SNOMED-CT and DSM-IV, we systematically curated a set of 13 SDOH categories and created annotation guidelines for these. After manually annotating the 3,504 sentences, we developed and tested three deep neural network (DNN) architectures - convolutional neural network (CNN), long short-term memory (LSTM) network, and the Bidirectional Encoder Representations from Transformers (BERT) - for automated detection of eight SDOH categories. We also compared these DNNs to three baselines models: (1) cTAKES, as well as (2) L2-regularized logistic regression and (3) random forests on bags-of-words. Model evaluation metrics included micro- and macro- F1, and area under the receiver operating characteristic curve (AUC). RESULTS: All three DNN models accurately classified all SDOH categories (minimum micro-F1 = 0.632, minimum macro-AUC = 0.854). Compared to the CNN and LSTM, BERT performed best in most key metrics (micro-F1 = 0.690, macro-AUC = 0.907). The BERT model most effectively identified the "occupational" category (F1 = 0.774, AUC = 0.965) and least effectively identified the "non-SDOH" category (F = 0.491, AUC = 0.788). BERT outperformed cTAKES in distinguishing social vs non-social sentences (BERT F1 = 0.87 vs. cTAKES F1 = 0.06), and outperformed logistic regression (micro-F1 = 0.649, macro-AUC = 0.696) and random forest (micro-F1 = 0.502, macro-AUC = 0.523) trained on bag-of-words. CONCLUSIONS: Our study framework with DNN models demonstrated improved performance for efficiently identifying a systematic range of SDOH categories from clinical notes in the EHR. Improved identification of patient SDOH may further improve healthcare outcomes.

Prediction of Atrial Fibrillation Using Machine Learning: A Review.

There has been recent immense interest in the use of machine learning techniques in the prediction and screening of atrial fibrillation, a common rhythm disorder present with significant clinical implications primarily related to the risk of ischemic cerebrovascular events and heart failure. Prior to the advent of the application of artificial intelligence in clinical medicine, previous studies have enumerated multiple clinical risk factors that can predict the development of atrial fibrillation. These clinical parameters include previous diagnoses, laboratory data (e.g., cardiac and inflammatory biomarkers, etc.), imaging data (e.g., cardiac computed tomography, cardiac magnetic resonance imaging, echocardiography, etc.), and electrophysiological data. These data are readily available in the electronic health record and can be automatically queried by artificial intelligence algorithms. With the modern computational capabilities afforded by technological advancements in computing and artificial intelligence, we present the current state of machine learning methodologies in the prediction and screening of atrial fibrillation as well as the implications and future direction of this rapidly evolving field.