Effect of Torsemide vs Furosemide After Discharge on All-Cause Mortality in Patients Hospitalized With Heart Failure: The TRANSFORM-HF Randomized Clinical Trial

Natural Language Processing for Adjudication of Heart Failure Hospitalizations in a Multi-Center Clinical Trial

Background

The gold standard for outcome adjudication in clinical trials is chart review by a physician clinical events committee (CEC), which requires substantial time and expertise. Automated adjudication by natural language processing (NLP) may offer a more resource-efficient alternative. We previously showed that the Community Care Cohort Project (C3PO) NLP model adjudicates heart failure (HF) hospitalizations accurately within one healthcare system.

Methods

This study externally validated the C3PO NLP model against CEC adjudication in the INVESTED trial. INVESTED compared influenza vaccination formulations in 5260 patients with cardiovascular disease at 157 North American sites. A central CEC adjudicated the cause of hospitalizations from medical records. We applied the C3PO NLP model to medical records from 4060 INVESTED hospitalizations and evaluated agreement between the NLP and final consensus CEC HF adjudications. We then fine-tuned the C3PO NLP model (C3PO+INVESTED) and trained a de novo model using half the INVESTED hospitalizations, and evaluated these models in the other half. NLP performance was benchmarked to CEC reviewer inter-rater reproducibility.

Results

1074 hospitalizations (26%) were adjudicated as HF by the CEC. There was high agreement between the C3PO NLP and CEC HF adjudications (agreement 87%, kappa statistic 0.69). C3PO NLP model sensitivity was 94% and specificity was 84%. The fine-tuned C3PO and de novo NLP models demonstrated agreement of 93% and kappa of 0.82 and 0.83, respectively. CEC reviewer inter-rater reproducibility was 94% (kappa 0.85).

Conclusion

Our NLP model developed within a single healthcare system accurately identified HF events relative to the gold-standard CEC in an external multi-center clinical trial. Fine-tuning the model improved agreement and approximated human reproducibility. NLP may improve the efficiency of future multi-center clinical trials by accurately identifying clinical events at scale.

Management of Worsening Heart Failure With Reduced Ejection Fraction: JACC Focus Seminar 3/3

Despite worsening heart failure (HF) being extremely common, expensive, and associated with substantial risk of death, there remain no dedicated clinical practice guidelines for the specific management of these patients. The lack of a management guideline is despite a rapidly evolving evidence-base, as a number of recent clinical trials have demonstrated multiple therapies to be safe and efficacious in this high-risk population. Herein, we propose a framework for treating worsening HF with reduced ejection fraction with the sense of urgency it deserves. This includes treating congestion; managing precipitants; and establishing a foundation of rapid-sequence, simultaneous, and/or in-hospital initiation of quadruple medical therapy for HF with reduced ejection fraction, with the top priority being at least low doses of all 4 medications. Moreover, to maximally reduce residual clinical risk, we further propose consideration of upfront simultaneous use of vericiguat (ie, quintuple medical therapy) and administration of intravenous iron for those who are iron deficient.

Tricuspid Regurgitation Management for Heart Failure

There is growing evidence that severe tricuspid regurgitation (TR) impacts clinical outcomes in a variety of cardiovascular disease states. The late presentation of patients with advanced TR highlights the underappreciation of the disease, as well as the pitfalls of current guideline-directed medical management. Given the high in-hospital mortality associated with isolated tricuspid valve surgery, transcatheter options continue to be explored with the hope of improved survival and reduced heart failure hospitalizations. In this review, we explore the physiology of TR, discuss the etiologic classes of TR, and explore the transcatheter options for treatment and who might benefit from device therapy.

Dendrogram of transparent feature importance machine learning statistics to classify associations for heart failure: A reanalysis of a retrospective cohort study of the Medical Information Mart for Intensive Care III (MIMIC-III) database

BACKGROUND

There is a continual push for developing accurate predictors for Intensive Care Unit (ICU) admitted heart failure (HF) patients and in-hospital mortality.

OBJECTIVE

The study aimed to utilize transparent machine learning and create hierarchical clustering of key predictors based off of model importance statistics gain, cover, and frequency.

METHODS

Inclusion criteria of complete patient information for in-hospital mortality in the ICU with HF from the MIMIC-III database were randomly divided into a training (n = 941, 80%) and test (n = 235, 20%). A grid search was set to find hyperparameters. Machine Learning with XGBoost were used to predict mortality followed by feature importance with Shapely Additive Explanations (SHAP) and hierarchical clustering of model metrics with a dendrogram and heat map.

RESULTS

Of the 1,176 heart failure ICU patients that met inclusion criteria for the study, 558 (47.5%) were males. The mean age was 74.05 (SD = 12.85). XGBoost model had an area under the receiver operator curve of 0.662. The highest overall SHAP explanations were urine output, leukocytes, bicarbonate, and platelets. Average urine output was 1899.28 (SD = 1272.36) mL/day with the hospital mortality group having 1345.97 (SD = 1136.58) mL/day and the group without hospital mortality having 1986.91 (SD = 1271.16) mL/day. The average leukocyte count in the cohort was 10.72 (SD = 5.23) cells per microliter. For the hospital mortality group the leukocyte count was 13.47 (SD = 7.42) cells per microliter and for the group without hospital mortality the leukocyte count was 10.28 (SD = 4.66) cells per microliter. The average bicarbonate value was 26.91 (SD = 5.17) mEq/L. Amongst the group with hospital mortality the average bicarbonate value was 24.00 (SD = 5.42) mEq/L. Amongst the group without hospital mortality the average bicarbonate value was 27.37 (SD = 4.98) mEq/L. The average platelet value was 241.52 platelets per microliter. For the group with hospital mortality the average platelet value was 216.21 platelets per microliter. For the group without hospital mortality the average platelet value was 245.47 platelets per microliter. Cluster 1 of the dendrogram grouped the temperature, platelets, urine output, Saturation of partial pressure of Oxygen (SPO2), Leukocyte count, lymphocyte count, bicarbonate, anion gap, respiratory rate, PCO2, BMI, and age as most similar in having the highest aggregate gain, cover, and frequency metrics.

CONCLUSION

Machine Learning models that incorporate dendrograms and heat maps can offer additional summaries of model statistics in differentiating factors between in patient ICU mortality in heart failure patients.

Effect of Torsemide Versus Furosemide on Symptoms and Quality of Life Among Patients Hospitalized for Heart Failure: The TRANSFORM-HF Randomized Clinical Trial

BACKGROUND

Loop diuretics are a primary therapy for the symptomatic treatment of heart failure (HF), but whether torsemide improves patient symptoms and quality of life better than furosemide remains unknown. As prespecified secondary end points, the TRANSFORM-HF trial (Torsemide Comparison With Furosemide for Management of Heart Failure) compared the effect of torsemide versus furosemide on patient-reported outcomes among patients with HF.

METHODS

TRANSFORM-HF was an open-label, pragmatic, randomized trial of 2859 patients hospitalized for HF (regardless of ejection fraction) across 60 hospitals in the United States. Patients were randomly assigned in a 1:1 ratio to a loop diuretic strategy of torsemide or furosemide with investigator-selected dosage. This report examined effects on prespecified secondary end points, which included Kansas City Cardiomyopathy Questionnaire Clinical Summary Score (KCCQ-CSS; assessed as adjusted mean difference in change from baseline; range, 0-100 with 100 indicating best health status; clinically important difference, ≥5 points) and Patient Health Questionnaire-2 (range, 0-6; score ≥3 supporting evaluation for depression) over 12 months.

RESULTS

Baseline data were available for 2787 (97.5%) patients for KCCQ-CSS and 2624 (91.8%) patients for Patient Health Questionnaire-2. Median (interquartile range) baseline KCCQ-CSS was 42 (27-60) in the torsemide group and 40 (24-59) in the furosemide group. At 12 months, there was no significant difference between torsemide and furosemide in change from baseline in KCCQ-CSS (adjusted mean difference, 0.06 [95% CI, -2.26 to 2.37]; P=0.96) or the proportion of patients with Patient Health Questionnaire-2 score ≥3 (15.1% versus 13.2%: P=0.34). Results for KCCQ-CSS were similar at 1 month (adjusted mean difference, 1.36 [95% CI, -0.64 to 3.36]; P=0.18) and 6-month follow-up (adjusted mean difference, -0.37 [95% CI, -2.52 to 1.78]; P=0.73), and across subgroups by ejection fraction phenotype, New York Heart Association class at randomization, and loop diuretic agent before hospitalization. Irrespective of baseline KCCQ-CSS tertile, there was no significant difference between torsemide and furosemide on change in KCCQ-CSS, all-cause mortality, or all-cause hospitalization.

CONCLUSIONS

Among patients discharged after hospitalization for HF, a strategy of torsemide compared with furosemide did not improve symptoms or quality of life over 12 months. The effects of torsemide and furosemide on patient-reported outcomes were similar regardless of ejection fraction, previous loop diuretic use, and baseline health status.

REGISTRATION

URL: https://www.

CLINICALTRIALS

gov; Unique identifier: NCT03296813.

The Potential Role of Torsemide in Optimizing Loop Diuretic Therapy for Heart Failure Patients

Heart failure is associated with an increased frequency of hospitalization, reduced life span, and greater risk to public health, thus posing a challenge. In India, torsemide is one of the commonly used loop diuretics for decongestion in heart failure. However, this use of torsemide, including its dosing, and up/down titration, is based on practical experience. Loop diuretic therapy for heart failure patients poses several dilemmas due to the lack of robust evidence based on which treatment decisions can be made. To guide physicians on the optimal use of torsemide in heart failure patients with or without renal impairment, a panel of expert cardiologists and nephrologists from India convened to develop this expert opinion document for the use of torsemide. This expert opinion on torsemide will pave the way for optimal management with loop diuretic therapy in real-world heart failure patients.

Leveraging electronic health records to streamline the conduct of cardiovascular clinical trials

Conventional randomized controlled trials (RCTs) can be expensive, time intensive, and complex to conduct. Trial recruitment, participation, and data collection can burden participants and research personnel. In the past two decades, there have been rapid technological advances and an exponential growth in digitized healthcare data. Embedding RCTs, including cardiovascular outcome trials, into electronic health record systems or registries may streamline screening, consent, randomization, follow-up visits, and outcome adjudication. Moreover, wearable sensors (i.e. health and fitness trackers) provide an opportunity to collect data on cardiovascular health and risk factors in unprecedented detail and scale, while growing internet connectivity supports the collection of patient-reported outcomes. There is a pressing need to develop robust mechanisms that facilitate data capture from diverse databases and guidance to standardize data definitions. Importantly, the data collection infrastructure should be reusable to support multiple cardiovascular RCTs over time. Systems, processes, and policies will need to have sufficient flexibility to allow interoperability between different sources of data acquisition. Clinical research guidelines, ethics oversight, and regulatory requirements also need to evolve. This review highlights recent progress towards the use of routinely generated data to conduct RCTs and discusses potential solutions for ongoing barriers. There is a particular focus on methods to utilize routinely generated data for trials while complying with regional data protection laws. The discussion is supported with examples of cardiovascular outcome trials that have successfully leveraged the electronic health record, web-enabled devices or administrative databases to conduct randomized trials.

Medical Treatment in Heart Failure with Reduced Ejection Fraction: A Proposed Algorithm Based on the Patient's Electrolytes and Congestion Status

In heart failure (HF) with reduced ejection fraction (HFrEF), four classes of drugs (β-blockers, angiotensin-converting enzyme inhibitors/angiotensin receptor neprilysin inhibitors, mineralocorticoid receptor antagonists, and the most recent Sodium-Glucose Co-Transporters 2 Inhibitors) have demonstrated positive results in randomized controlled trials (RCTs). Nevertheless, the latest RCTs are not proper for comparison since they were carried out at various times with dissimilar background therapies and the patients enrolled did not have the same characteristics. The difficulty of extrapolating from these trials and proposing a common framework appropriate for all cases is thus obvious. Despite the fact that these four agents are now the fundamental pillars of HFrEF treatment, the built-up algorithm of initiation and titration is a matter of debate. Electrolyte disturbances are common in HFrEF patients and can be attributed to several factors, such as the use of diuretics, renal impairment, and neurohormonal activation. We have identified several HFrEF phenotypes according to their sodium (Na+) and potassium (K+) status in a "real world" setting and suggest an algorithm on how to introduce the most appropriate drug and set up therapy based on the patients' electrolytes and the existence of congestion.

Congestion and Use of Diuretics in Heart Failure and Cardiomyopathies: a Practical Guide

PURPOSE OF REVIEW

Heart failure is a highly prevalent condition caused by many different aetiologies and characterised by cardiac dysfunction and congestion. Once developed, congestion leads to signs (peripheral oedema) and symptoms (breathlessness on exertion), adverse cardiac remodelling, and an increased risk of hospitalisation and premature death. This review summarises strategies that could enable early identification and a more objective management of congestion in patients with heart failure.

RECENT FINDINGS

For patients with suspected or diagnosed heart failure, combining an echocardiogram with assessment of great veins, lungs, and kidneys by ultrasound might facilitate recognition and quantification of congestion, the management of which is still difficult and highly subjective. Congestion is a one of the key drivers of morbidity and mortality in patients with heart failure and is often under-recognised. The use of ultrasound allows for a timely, simultaneous identification of cardiac dysfunction and multiorgan congestion; ongoing and future studies will clarify how to tailor diuretic treatments in those with or at risk of heart failure.

How to Manage Heart Failure With Preserved Ejection Fraction: Practical Guidance for Clinicians

Although patients with heart failure with preserved ejection fraction (HFpEF) (left ventricular ejection fraction ≥50%) comprise nearly half of those with chronic heart failure, evidence-based treatment options for this population have historically been limited. Recently, however, emerging data from prospective, randomized trials enrolling patients with HFpEF have greatly altered the range of pharmacologic options to modify disease progression in selected patients with HFpEF. In the context of this evolving landscape, clinicians are increasingly in need of practical guidance regarding the best approach to management of this growing population. In this review, we build on the recently published heart failure guidelines by integrating contemporary data from recent randomized trials to provide a contemporary framework for diagnosis and evidence-based treatment of patients with HFpEF. Where gaps in knowledge persist, we provide "best available" data from post hoc analyses of clinical trials or data from observational studies to guide management until more definitive studies are available.

Heart Failure With Preserved Ejection Fraction: A Review

IMPORTANCE

Heart failure with preserved ejection fraction (HFpEF), defined as HF with an EF of 50% or higher at diagnosis, affects approximately 3 million people in the US and up to 32 million people worldwide. Patients with HFpEF are hospitalized approximately 1.4 times per year and have an annual mortality rate of approximately 15%.

OBSERVATIONS

Risk factors for HFpEF include older age, hypertension, diabetes, dyslipidemia, and obesity. Approximately 65% of patients with HFpEF present with dyspnea and physical examination, chest radiographic, echocardiographic, or invasive hemodynamic evidence of HF with overt congestion (volume overload) at rest. Approximately 35% of patients with HFpEF present with "unexplained" dyspnea on exertion, meaning they do not have clear physical, radiographic, or echocardiographic signs of HF. These patients have elevated atrial pressures with exercise as measured with invasive hemodynamic stress testing or estimated with Doppler echocardiography stress testing. In unselected patients presenting with unexplained dyspnea, the H2FPEF score incorporating clinical (age, hypertension, obesity, atrial fibrillation status) and resting Doppler echocardiographic (estimated pulmonary artery systolic pressure or left atrial pressure) variables can assist with diagnosis (H2FPEF score range, 0-9; score >5 indicates more than 95% probability of HFpEF). Specific causes of the clinical syndrome of HF with normal EF other than HFpEF should be identified and treated, such as valvular, infiltrative, or pericardial disease. First-line pharmacologic therapy consists of sodium-glucose cotransporter type 2 inhibitors, such as dapagliflozin or empagliflozin, which reduced HF hospitalization or cardiovascular death by approximately 20% compared with placebo in randomized clinical trials. Compared with usual care, exercise training and diet-induced weight loss produced clinically meaningful increases in functional capacity and quality of life in randomized clinical trials. Diuretics (typically loop diuretics, such as furosemide or torsemide) should be prescribed to patients with overt congestion to improve symptoms. Education in HF self-care (eg, adherence to medications and dietary restrictions, monitoring of symptoms and vital signs) can help avoid HF decompensation.

CONCLUSIONS AND RELEVANCE

Approximately 3 million people in the US have HFpEF. First-line therapy consists of sodium-glucose cotransporter type 2 inhibitors, exercise, HF self-care, loop diuretics as needed to maintain euvolemia, and weight loss for patients with obesity and HFpEF.