A Machine Learning Methodology for Identification and Triage of Heart Failure Exacerbations

James Morrill,Nicholas Wysham,Ted Smith,Andrew Ambrosy,Klajdi Qirko,Botros Toro,Jacob Kelly,Marat Fudim,Sumanth Swaminathan

doi:10.1007/s12265-021-10151-7

Abstract

Inadequate at-home management and self-awareness of heart failure (HF) exacerbations are known to be leading causes of the greater than 1 million estimated HF-related hospitalizations in the USA alone. Most current at-home HF management protocols include paper guidelines or exploratory health applications that lack rigor and validation at the level of the individual patient. We report on a novel triage methodology that uses machine learning predictions for real-time detection and assessment of exacerbations. Medical specialist opinions on statistically and clinically comprehensive, simulated patient cases were used to train and validate prediction algorithms. Model performance was assessed by comparison to physician panel consensus in a representative, out-of-sample validation set of 100 vignettes. Algorithm prediction accuracy and safety indicators surpassed all individual specialists in identifying consensus opinion on existence/severity of exacerbations and appropriate treatment response. The algorithms also scored the highest sensitivity, specificity, and PPV when assessing the need for emergency care.Lay summaryHere we develop a machine-learning approach for providing real-time decision support to adults diagnosed with congestive heart failure. The algorithm achieves higher exacerbation and triage classification performance than any individual physician when compared to physician consensus opinion.Graphical abstract

Highlights

Heart failure (HF) is a chronic, progressive condition in which the heart muscle is unable to pump sufficient blood to meet the body’s circulatory and oxygenation needs
We found that for each of triage, exacerbation, and treatment identification, the topperforming algorithm consisted of a combination of linear discriminant analysis and naive Bayes classifiers combined through a soft voting strategy
This study has shown that a machine-learning approach to triaging patients with HF is a viable and accurate method of facilitating at-home triage and exacerbation self-identification when compared to individual heart specialists

Summary

Introduction

Heart failure (HF) is a chronic, progressive condition in which the heart muscle is unable to pump sufficient blood to meet the body’s circulatory and oxygenation needs. Hospitalizations from HF exacerbations account for 6.5 million hospital days, the leading cause of hospitalization in the USA and Europe [3]. Acute exacerbations of HF are due to a maladaptive accumulation of intravascular volume eventually leading to dyspnea and respiratory failure but can be identified by changes in symptoms and physiologic parameters. If recognized in a timely manner, exacerbations can be safely managed at home through changes to medication and diet. Despite the recognized impact of these exacerbations, there is no universally accepted clinical approach for self-identification of HF exacerbations by patients at home. The result is delayed treatment (which leads to more severe exacerbation episodes), long-term debilitation in quality-of-life, and severe economic burden. Recognition and treatment of HF exacerbations is an unmet need that leads to morbidity, mortality, and unnecessary healthcare utilization

Methods

Results

Discussion

Conclusion