Fair prediction of 2-year stroke risk in patients with atrial fibrillation.

OBJECTIVE: This study aims to develop machine learning models that provide both accurate and equitable predictions of 2-year stroke risk for patients with atrial fibrillation across diverse racial groups. MATERIALS AND METHODS: Our study utilized structured electronic health records (EHR) data from the All of Us Research Program. Machine learning models (LightGBM) were utilized to capture the relations between stroke risks and the predictors used by the widely recognized CHADS2 and CHA2DS2-VASc scores. We mitigated the racial disparity by creating a representative tuning set, customizing tuning criteria, and setting binary thresholds separately for subgroups. We constructed a hold-out test set that not only supports temporal validation but also includes a larger proportion of Black/African Americans for fairness validation. RESULTS: Compared to the original CHADS2 and CHA2DS2-VASc scores, significant improvements were achieved by modeling their predictors using machine learning models (Area Under the Receiver Operating Characteristic curve from near 0.70 to above 0.80). Furthermore, applying our disparity mitigation strategies can effectively enhance model fairness compared to the conventional cross-validation approach. DISCUSSION: Modeling CHADS2 and CHA2DS2-VASc risk factors with LightGBM and our disparity mitigation strategies achieved decent discriminative performance and excellent fairness performance. In addition, this approach can provide a complete interpretation of each predictor. These highlight its potential utility in clinical practice. CONCLUSIONS: Our research presents a practical example of addressing clinical challenges through the All of Us Research Program data. The disparity mitigation framework we proposed is adaptable across various models and data modalities, demonstrating broad potential in clinical informatics.

Poorer outcomes associated with more invasive lead management strategies for Abbott Riata® leads: a large, multicenter experience.

BACKGROUND: Over 100,000 Abbott Riata® were implanted in the United States before they were recalled in 2010. There are still a significant number of Abbott Riata® leads in use, and it is unclear how these leads should be managed at the time of generator change or lead malfunction. Although data comparing both Sprint Fidelis® and Abbott Riata® leads in this setting is available, there are no multicenter comparative studies of outcomes for various lead management strategies, including lead extraction (LE), lead abandonment/revision (LA), and generator change (GC) only at the time of device at elective replacement interval (ERI) for Abbott Riata® leads. METHODS: A retrospective, multicenter study was undertaken to compare short-term outcomes (major complications-MC, death, extended or re-hospitalizations within 60 days-RH, lead malfunction-LM) and total outcomes (short-term outcomes & lead malfunction during follow-up) of patients with Riata® leads undergoing LE, LA, or GC. RESULTS: 152 patients (65 ± 13 years, 68% male) were followed for a mean 33 ± 30 months following intervention. Out of 166 procedures, 13 patients underwent LE, 16 patients underwent LA, and 137 patients underwent GC. There was 1 major complication in each group, yielding an event rate of 7.7% for LE, 6.3% for LA, and 0.7% for GC cohorts. There were significantly more short-term and total adverse outcomes in the group of patients getting LE and LA versus GC only (38.5% & 31.3% vs 7.3%, P < 0.001). Total Riata® lead dwell time follow-up was 17,067 months. A total of 3 Riata® lead malfunctions were noted during long-term follow-up. Inappropriate shocks were similar between LE 7.7% (1/13), LA 6.3% (1/16). and GC 11.0% (4/136); P = 0.57. CONCLUSIONS: There were more short-term and total adverse outcomes in more invasive management strategies (LE and LA) versus GC alone. The failure rate of Riata® leads was substantially lower compared to previous reports. Therefore, we recommend only performing battery exchange when a device with an active Riata® lead is at ERI, unless there is malfunction of the Riata® lead noted on testing. There were significantly more short-term adverse outcomes in the lead extraction (5/13) and lead abandonment/revision (5/16) groups than the generator only (8/137) group (P < 0.001). GIB - Gastrointestinal bleed, CHF - congestive heart failure, NSTEMI - non-ST elevation MI.

Radiation-associated Arrhythmias: Putative Pathophysiological Mechanisms, Prevalence, Screening and Management Strategies.

Radiation-associated cardiovascular disease, an increasingly recognised disease process, is a significant adverse effect of radiation therapy for common malignancies that involve the chest, and include lymphomas, lung, mediastinal and breast cancers. Two factors contribute to the increasing incidence of radiation-associated cardiovascular disease: advances in malignancy detection and the improved survival of cancer patients, by which many symptoms of radiation-associated cardiovascular disease, specifically radiation-associated arrhythmias, present years and/or decades following initial radiotherapy. We present a focused overview of the currently understood pathophysiology, prevalence and management strategies of radiation-associated arrhythmias, which include bradyarrhythmias, tachyarrhythmias and autonomic dysfunction.

Comparison of quantifiable electrocardiographic changes associated with severe hyperkalemia.

BACKGROUND: Hyperkalemia (HK) is a life-threatening condition that is frequently evaluated by electrocardiogram (ECG). ECG changes in severe HK (≥ 6.3 mEq/L) are not well-characterized. This study sought to compare and correlate ECG metrics in severe HK to baseline normokalemic ECGs and serum potassium. METHODS: A retrospective analysis of 340 severe HK encounters with corresponding normokalemic ECGs was performed. RESULTS: Various ECG metrics were analyzed. P wave amplitude in lead II, QRS duration, T wave slope, ratio of T wave amplitude: duration, and ratios of T wave: QRS amplitudes were significantly different between normokalemic and HK ECGs. P wave amplitude attenuation in lead II correlated better with serum potassium than in V1. T wave metrics that incorporated both T wave and QRS amplitudes correlated better than metrics utilizing T wave metrics alone. CONCLUSION: Multiple statistically significant and quantifiable differences among ECG metrics were observed between normokalemic and HK ECGs and correlated with increasing degrees of serum potassium and along the continuum of serum potassium. When incorporated into a logistic regression model, the ability to distinguish HK versus normokalemia on ECG improved significantly. These findings could be integrated into an ECG acquisition system that can more accurately identify severe HK.

Electrophysiological evaluation following development of new and persistent left bundle branch block after transcatheter aortic valve replacement: A single center pilot study.

INTRODUCTION: New and persistent left bundle branch block (NP-LBBB) following Transcatheter Aortic Valve Replacement (TAVR) is an ongoing concern with incidence ranging from as low as 4% to up to 65% (varying for different types of valves). Such patients are at risk of developing high-grade atrioventricular block (HAVB) warranting permanent pacemaker (PPM) implantation. However, currently, there are no consensus guidelines or large prospective studies to risk stratify these patients for safer discharge after TAVR. OBJECTIVES: To provide insight from a single center study on using modified electrophysiology (EP) study to risk stratify post-TAVR patients to outpatient monitoring for low-risk versus pacemaker implantation for high-risk patients. METHODS AND RESULTS: Between June 2020 and March 2023, all patients who underwent a TAVR procedure (324 patients) at our institution were screened for development of NP-LBBB post-operatively. Out of 26 patients who developed NP-LBBB, after a pre-specified period of observation, 18 patients were deemed eligible for a modified EP study to assess His-Ventricular (HV) interval. 11 out of 18 patients (61.1%) had normal HV interval (HV < 55 ms). Three out of 18 patients (16.7%) had HV prolongation (55 ms < HV < 70 ms) without significant HV prolongation (defined as an increase in HV interval > 30%) with intra-procedural procainamide challenge. Four out of 18 patients (22.2%) had significant HV prolongation (HV > 70 ms) warranting PPM implantation based on a multidisciplinary approach and shared decision-making with the patients. Total of 50% of patients discharged with PPM (two out of four patients) were noted to be pacemaker dependent based on serial device interrogations. All patients who did not receive PPM were discharged with ambulatory monitoring with 30-day event monitor and did not develop HAVB on serial follow-up. CONCLUSION: Normal HV interval up to 55 ms on modified EP study after TAVR and development of NP-LBBB can be utilized as a threshold for risk stratification to facilitate safe discharge. The optimal upper limit of HV interval threshold remains unclear in determining appropriate candidacy for PPM.

Echocardiographic changes and heart failure hospitalizations following rhythm control for arrhythmia-induced cardiomyopathy: results from a multicenter, retrospective study.

BACKGROUND: The incidence and prevalence of arrhythmia-induced cardiomyopathy (AIC) are unclear but likely underrecognized. LV dysfunction is common among patients with atrial fibrillation (AF), atrial flutter (AFL), and frequent premature ventricular contractions (PVC). The hallmark of AIC is the improvement of left ventricular ejection fraction (LVEF) following arrhythmia treatment. Changes in echocardiographic parameters and their effect on outcomes after rhythm control for AIC are not well understood. We aimed to study echocardiographic parameters and outcomes following rhythm control for AIC. METHODS: A multicenter, retrospective study was conducted at 4 different medical centers involving patients with AIC. Clinical, echocardiographic, and outcome (mortality and heart failure hospitalizations [HFH]) parameters were extracted from the medical record. RESULTS: Two hundred fifty-five patients (age 66 ± 11 years, 73% male) with AIC caused by AF (51%), atrial tachycardia/AFL (20%), and PVCs (29%) were included and followed for a median period of 6 months after successful rhythm control. Significant improvements in left ventricular (LV) ejection fraction (P < 0.0001), LV end-systolic volume (ml) (90 ± 48 to 58 ± 30; P < 0.0001), LV internal diameter end diastole (cm) (5.5 ± 0.78 to 5.3 ± 0.64; P = 0.0001) and end systole (4.7 ± 0.95 to 4.3 ± 1.02; P < 0.0001), right atrial pressure (mmHg) (11.3 ± 5.0 to 7.4 ± 3.2; P = 0.0001), and right ventricular function (n (%)) (42 (44) to 9 (11); P < 0.0001) were noted following arrhythmia treatment. No deaths occurred during follow-up. HFH occurred in 7 patients. Arrhythmia recurrence rate was 50.5%. Neither echocardiographic parameters nor recurrence of arrhythmia correlated with HFH. CONCLUSION: Arrhythmia treatment significantly improved echocardiographic LV dimensions, LVEF, and RAP in this multicenter AIC cohort, underscoring the need for early recognition and aggressive rhythm control in suspected AIC patients. The event rate was too low to assess for outcome predictors.

Drug Interactions Affecting Antiarrhythmic Drug Use.

Antiarrhythmic drugs (AAD) play an important role in the management of arrhythmias. Drug interactions involving AAD are common in clinical practice. As AADs have a narrow therapeutic window, both pharmacokinetic as well as pharmacodynamic interactions involving AAD can result in serious adverse drug reactions ranging from arrhythmia recurrence, failure of device-based therapy, and heart failure, to death. Pharmacokinetic drug interactions frequently involve the inhibition of key metabolic pathways, resulting in accumulation of a substrate drug. Additionally, over the past 2 decades, the P-gp (permeability glycoprotein) has been increasingly cited as a significant source of drug interactions. Pharmacodynamic drug interactions involving AADs commonly involve additive QT prolongation. Amiodarone, quinidine, and dofetilide are AADs with numerous and clinically significant drug interactions. Recent studies have also demonstrated increased morbidity and mortality with the use of digoxin and other AAD which interact with P-gp. QT prolongation is an important pharmacodynamic interaction involving mainly Vaughan-Williams class III AAD as many commonly used drug classes, such as macrolide antibiotics, fluoroquinolone antibiotics, antipsychotics, and antiemetics prolong the QT interval. Whenever possible, serious drug-drug interactions involving AAD should be avoided. If unavoidable, patients will require closer monitoring and the concomitant use of interacting agents should be minimized. Increasing awareness of drug interactions among clinicians will significantly improve patient safety for patients with arrhythmias.

Drug Interactions Affecting Oral Anticoagulant Use.

Oral anticoagulants (OACs) are medications commonly used in patients with atrial fibrillation and other cardiovascular conditions. Both warfarin and direct oral anticoagulants are susceptible to drug-drug interactions (DDIs). DDIs are an important cause of adverse drug reactions and exact a large toll on the health care system. DDI for warfarin mainly involve moderate to strong inhibitors/inducers of cytochrome P450 (CYP) 2C9, which is responsible for the elimination of the more potent S-isomer of warfarin. However, inhibitor/inducers of CYP3A4 and CYP1A2 may also cause DDI with warfarin. Recognition of these precipitating agents along with increased frequency of monitoring when these agents are initiated or discontinued will minimize the impact of warfarin DDI. Direct oral anticoagulants are mainly affected by medications strongly affecting the permeability glycoprotein (P-gp), and to a lesser extent, strong CYP3A4 inhibitors/inducers. Dabigatran and edoxaban are affected by P-gp modulation. Strong inducers of CYP3A4 or P-gp should be avoided in all patients taking direct oral anticoagulant unless previously proven to be otherwise safe. Simultaneous strong CYP3A4 and P-gp inhibitors should be avoided in patients taking apixaban and rivaroxaban. Concomitant antiplatelet/anticoagulant use confers additive risk for bleeding, but their combination is unavoidable in many cases. Minimizing duration of concomitant anticoagulant/antiplatelet therapy as indicated by evidence-based clinical guidelines is the best way to reduce the risk of bleeding.

More Generalizable Models For Sepsis Detection Under Covariate Shift.

Sepsis is a major cause of mortality in the intensive care units (ICUs). Early intervention of sepsis can improve clinical outcomes for sepsis patients1,2,3. Machine learning models have been developed for clinical recognition of sepsis4,5,6. A common assumption of supervised machine learning models is that the covariates in the testing data follow the same distributions as those in the training data. When this assumption is violated (e.g., there is covariate shift), models that performed well for training data could perform badly for testing data. Covariate shift happens when the relationships between covariates and the outcome stay the same, but the marginal distributions of the covariates differ among training and testing data. Covariate shift could make clinical risk prediction model nongeneralizable. In this study, we applied covariate shift corrections onto common machine learning models and have observed that these corrections can help the models be more generalizable under the occurrence of covariate shift when detecting the onset of sepsis.

Atrial flutter mimicking ST-elevation myocardial infarction.

A 12 lead electrocardiogram provides an important diagnostic tool for atrial flutter recognition. However, rarely, atrial flutter waves can cause diagnostic challenges by producing ST segment abnormalities mimicking ST segment elevation and result in unnecessary workup and treatment. ​.

Supraventricular tachycardia with the use of phentermine: case report and review of literature.

Case: A 34-year-old woman with no significant past medical history presented to the hospital with sudden onset of palpitations with associated dyspnea and chest discomfort. She denied any similar previous episodes. Initial electrocardiogram (EKG) was consistent with a short R-P interval supraventricular tachycardia (SVT). Her transthoracic echocardiogram (TTE) revealed no structural abnormalities, TSH levels were normal, and urine drug screen was negative for any recreational drugs. However, the patient had been taking phentermine for weight loss.Discussion: The exact mechanism is not clear; however, we postulate that the sympathomimetic effects of phentermine likely contribute to SVT induction through enhanced AV nodal conduction or increased atrial ectopy. Conclusions: The only medication she was taking at home was phentermine, and the palpitations did not recur after discontinuation of the drug during follow-up. It is important to collect a thorough medication history when patients present with AV nodal reentrant tachycardia (AVNRT) or other SVT.

Basic surface electrocardiogram interpretation for the pharmacist.

PURPOSE: The electrocardiogram (ECG) is an invaluable tool for clinicians that provides important information about a patient's heart. As clinical pharmacists play an ever-increasing role in cardiovascular care, ECG interpretation is an important skill with which to become familiar. SUMMARY: The ECG provides information on both electrical and biomechanical aspects of the heart. Electrical information such as the rhythm, rate, and axis of the electrical activity can all be provided by the ECG. Biomechanical information about the heart, such as the presence of ventricular hypertrophy and repolarization changes that may be associated with ischemia or myocardial injury, can also easily be gleaned from the ECG. Furthermore, the ECG plays a central role in both the diagnosis and treatment of common clinical conditions such as atrial fibrillation, ischemic heart disease, and QT interval prolongation. CONCLUSION: The ECG is one of the most commonly performed diagnostic tests, and clinicians should become familiar with its basic interpretation.

Predictors of myocardial recovery in arrhythmia-induced cardiomyopathy: A multicenter study.

BACKGROUND: Arrhythmia-induced cardiomyopathy (AIC) is characterized by improvement in left ventricular ejection fraction (LVEF) following arrhythmia treatment. Predictors of recovery in LVEF are not well understood. OBJECTIVE: We evaluated predictors of AIC recovery in a large multicenter cohort. METHODS: In total, 243 patients (age 65 ± 11, 73% male) with AIC caused by atrial fibrillation (49%), atrial tachycardia (20%), and premature ventricular contractions (PVCs; 31%) were treated and included. LVEF was assessed before and after treatment. Patients were stratified by arrhythmia duration (known [KN, n = 132] vs. unknown [UKN, n = 111]), arrhythmia type, LVEF, and presence of structural heart disease (SHD). RESULTS: Arrhythmia treatment was rhythm control in 95%. Median arrhythmia duration in the KN group was 47 months (25-75th percentile, 24-80 months). Post treatment LVEF was higher in KN group (55.9 ± 7 vs. 46.2 ± 12%; p < .0001) but the degree of LVEF improvement was similar (21.2 ± 9 vs. 19.4 ± 11; p = .16). Comparing highest quartile (longest arrhythmia duration) versus the rest of the KN group, the extent of LVEF improvement was similar (21.5 ± 8 vs. 21 ± 9%; p = .1). Patients in lowest index LVEF quartile (n = 74) had more PVC-induced AIC, greater EF improvement after treatment (24 ± 17 vs. 19 ± 7%; p < .0001) but lower post treatment EF (45 ± 14 vs. 54 ± 8%; p < .0001) versus other patients. Patients with SHD had lower index EF (28 ± 8 vs. 34 ± 8%; p < .0001) and lower final EF (47 ± 12 vs. 56 ± 7; p ≪ .0001). In multivariate regression, low index LVEF predicted myocardial recovery (odds ratio, 11.4; p < .005). CONCLUSIONS: In this AIC cohort, LVEF improved regardless of arrhythmia duration or type but those with PVCs had lower index LVEF and had less recovery. Low index LVEF predicted LVEF recovery following arrhythmia treatment.

Entrapment of diagnostic catheter within Advisor HD grid mapping catheter.

The potential for catheter entanglement with the HD Grid mapping catheter is explicitly stated in the manufacturer's product manual. A case of an entrapped 6 French quadripolar diagnostic catheter within an HD Grid mapping catheter is presented. We discuss the diagnosis, management, and resolution of this complication in our patient. The patient's arrhythmia was successfully eliminated, and no vascular complication in the postprocedural setting nor arrhythmia recurrence at follow-up were observed. Strategies to prevent and safely manage this complication, while salvaging access, are also discussed.