key: cord-0259771-pr2vzygr authors: Zaman, A.; Calcagno, S.; Zoccai, G. B.; Campbell, N.; Koulaouzidis, G.; Tsipas, D.; Kecskes, I. title: Unraveling diagnostic co-morbidity makeup of each HF category as characteristically derived by ECG- and ECHO-findings, a prevalence analysis date: 2021-10-01 journal: nan DOI: 10.1101/2021.09.30.21264236 sha: 5fb7cf10d57ccf3e9018a4e8339a83268fd27b7b doc_id: 259771 cord_uid: pr2vzygr Heart Failure (HF) relies mainly on measurements from Echocardiography, in particular Echo-Findings, to estimate Left Ventricle Ejection Fraction (LVEF) and evaluating structural heart disease criteria. As Echocardiography is not available in primary care, the key structural (heart chamber enlargements) and functional abnormality related measurements are not available precluding the ability to diagnose HF other than through mainly symptomatic means. The opportunity for earlier detection of HF is lost. In this work, we first explore each of the three HF categories, preserved EF, mild-reduced EF, and reduced EF, using various morphological and functional etiology-specific characteristics supported by a literature review and an extensive analysis of a large, dedicated database accumulated over 8 years. We then explore the typical signs and co-morbidities of HF using prevalence analysis to unravel the diagnostic makeup of each HF category as characteristically derived by ECG- and ECHO-findings. From this, we then conduct a principal component analysis (PCA) of the data to interpret patterns of comorbidities, showing groups of comorbidities frequently associated with each other. Lastly, we delve into the role of breakthrough methods for the analysis of bio-signals to replicate common ECHO-findings, as alternatives for detecting and diagnosing HF similarly to Echocardiography, thereby providing a simple device for the effective detection of HF for use in Primary Care. Heart Failure (HF) relies mainly on measurements from Echocardiography, in particular Echo-27 Findings, to estimate Left Ventricle Ejection Fraction (LVEF) and evaluating structural heart 28 disease criteria. As Echocardiography is not available in primary care, the key structural (heart 29 chamber enlargements) and functional abnormality related measurements are not available 30 precluding the ability to diagnose HF other than through mainly symptomatic means. The 31 opportunity for earlier detection of HF is lost. 32 In this work, we first explore each of the three HF categories, preserved EF, mild-reduced EF, 33 and reduced EF, using various morphological and functional etiology-specific characteristics 34 supported by a literature review and an extensive analysis of a large, dedicated database 35 accumulated over 8 years. 36 We then explore the typical signs and co-morbidities of HF using prevalence analysis to unravel 37 the diagnostic makeup of each HF category as characteristically derived by ECG-and ECHO-38 findings. From this, we then conduct a principal component analysis (PCA) of the data to 39 interpret patterns of comorbidities, showing groups of comorbidities frequently associated with 40 each other. 41 Lastly, we delve into the role of breakthrough methods for the analysis of bio-signals to replicate 42 common ECHO-findings, as alternatives for detecting and diagnosing HF similarly to 43 Echocardiography, thereby providing a simple device for the effective detection of HF for use in 44 Primary [44]; primarily mild LVSD, but with features of DD. The diagnostic criteria for HFmrEF 103 includes any relevant structural heart disease, such as LVH or LAE or DD beside symptoms and 104 the mildly reduced EF (LVEF of 40-49%) [1] . Priority patients who transitioned to HFrEF were 105 more likely to have LAE and had a tendency to have AFib and more comorbidities [32] . 106 However, HFmrEF is associated with different characteristics and a more favorable prognosis 107 than HFrEF [25] . 108 Mild asymptomatic LVSD (ALVSD) might be a predictor of adverse events mainly in subjects 109 with combined DD [ . The ground truth of ECHO-findings was established by cardiologist consensus and confirmed 157 with ECHO measurements (listed in Table 4 ). Table 1 lists the ECHO-findings having 158 cardiologist consensus ground truth, selected based on relevance for referral decision and high 159 prevalence of diseases in the intended population (prevalence >3%). 160 includes normal patients and patients with non-HF CVD -typically with a "mild" CVD 199 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The limitation of the database is the lack of natriuretic peptides level confirmation. However, this 203 was partly substituted with the use of a broader range of ECG and ECHO findings. 204 Both HF Scoring systems were used to verify the HF categorization in Cardio Phoenix database. 205 Fig. 1 shows the reference three HF categories with the absent HF patients in function of LVEF 206 (vertical axis) and HF Scores (horizontal axis): A) graph -H2FPEF Score, B) graph -HFA-PEFF 207 Score. The two HF Scores have similar results, and the C) graph compares them to each other. 208 The HFmrEF and HFrEF categories are primarily confirmed by LVEF, but the two HF Scores 209 also show non-normal ranges, similarly for HFpEF categories. Both the HF Scores are calculated 210 using sigmoid probability function instead of strict binary threshold in order to ensure higher 211 resolution and to better resolve borderline cases. 212 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted October 1, 2021. . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted October 1, 2021. deviation is also highlighted, which denotes the highest variability. 227 Table 3 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted October 1, 2021. . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted October 1, 2021. . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted October 1, 2021. 3.3.2 Heart failure and co-morbidities 294 Table 6 shows the HF categories and the most important associated co-morbidities as the 295 summary of prevalence analysis. The list of typical abnormalities or co-morbidities are listed 296 using both the literature and statistical analysis based on the presented data. 297 298 In the presented analysis LVEF was used as an independent variable to estimates the prevalence 308 of ECG and ECHO abnormalities. LVEF as a continual measurement is split into 9 categories 309 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted October 1, 2021. The results are plotted in Fig. 2 together with the average value of some principal ECHO 315 measurements. 316 317 The 1 st graph of Fig. 2 shows a deteriorating trend of all the key ECHO-measurements as a 318 function of LVEF. Obesity is measured by average BMI shows a slight increasing trend, and the 319 percentage of Female patients drastically decrease with the decreasing LVEF. This confirms that 320 the HFpEF patient is biased to females, and the HFrEF patient biased to males. 321 The 2 nd graph of Fig. 2 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted October 1, 2021. ; https://doi.org/10.1101/2021.09.30.21264236 doi: medRxiv preprint The 3 nd graph of Fig. 2 shows constant increasing prevalence of almost all ECG-findings, except: 326 S. Bradycardia having maximum around LVEF=51%; Ischemic ST-T having maximum around 327 LVEF=41%; Rightward Axis having maximum around LVEF=71% and LVEF=36%. Typically, 328 the HF patient has LVH with leftward axis, the Rightward Axis decreasing, but the typical right-329 side disease patients belong to HFrEF, that is why it is increasing again at the LVEF<40% 330 groups. 331 The 4 th graph of Fig. 2 shows the prevalence of the discussed HF categories, primarily separated 332 based on LVEF. The maximum prevalence of HFpEF is around LVEF=51%, hence below 50% 333 the HFmrEF replacing the HFpEF classification and below 40% HFrEF replacing HFmrEF 334 classification. 335 5 PCA comorbidity statistics CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted October 1, 2021. ; https://doi.org/10.1101/2021.09.30.21264236 doi: medRxiv preprint L V E F % L V E j e c t i o n F r a c t i o n ( Q u i n o n e s E q u a t i o n ) L V M I g / m 2 L e f t V e n t r i c u l a r M a s s I n d e x L V I regurgitation, and PH. 439 2 Cardio-HART™ from Cardio-Phoenix 3 FDA require that Cardio-Phoenix not use the term Echo-findings as they are derived from images, whereas HART-findings™ are bio-signal derived. Echo-findings are disease equivalent to HART-findings™ . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 1, 2021. ; https://doi.org/10.1101/2021.09.30.21264236 doi: medRxiv preprint As such, bio-signal derived ECHO-findings processed through the CHART AI system, can 440 predict the structural and functional abnormalities and measurements essential to the detection 441 accuracy for HF. This means that on initial patient presentation to Primary Care, practitioners 442 will be able to recognize HF, including classifying its severity and category, precluding purely 443 symptomatic based detection of HF through repeated and costly clinical visits and testing, that 444 delays access to treatment especially when symptoms remain inconclusive. 445 . The right side of the heart is less represented by the included ECHO-parameters and ECG even 447 less, as such their place in the statistical results is limited. 448 Nonetheless, right-side heart failure and its structural, functional abnormalities and related 449 pulmonary disease co-morbidities (COPD, Covid-19) have attracted growing attention in the last 450 few years. Right-heart side study through a bio-signal approach shows strong potential. 451 Further research is warranted to investigate in more detail the relationship between HF categories 452 and right-side HF through the window of bio-signals, risk factors, detectable abnormalities, 453 COPD and other co-morbidities and symptoms. 454 The prevalence-based comorbidity analysis shows that the above-mentioned ECHO findings are 456 strong indicators for HF and its category. More precisely, Echo shows the enlarged heart with 457 decreased myocardial contractility and general abnormal heart functioning. 458 As Echocardiography is not available in primary care, the key structural (heart chamber 459 enlargements) and functional abnormalities related measurements are not available precluding 460 the ability to diagnose HF other than through mainly symptomatic means. 461 The current results suggest that a key set of ECHO-findings are sufficiently representative of HF 462 when taken together as opposed to a single measurement, like LVEF. The two HF score 463 techniques discussed in this study, the HFA-PEFF score [4] and the H2FPEF score [36], validate 464 this approach, as they rely on Echo-finding indicators for the critical components of their score, 465 precisely to avoid reliance on LVEF. The use of LVEF only would lead to a limited HF 466 prediction capability focusing only on the HFrEF category, ignoring HFpEF altogether and much 467 of HFmrEF, resulting in a high incidence of FN. 468 As such, the bio-signal based approach classifying a set of ECHO-findings is recommended, in 469 contrast, single LVEF would inherit its limitation. 470 The use of novel bio-signals of a physiological nature such as those found in Cardio-HART™ 471 can provide to primary care, and users of the various Scoring techniques, the missing 472 echocardiography elements to understanding the structural and functional abnormalities and 473 thereby increase HF detection accuracy for all 3 types of HF, even in asymptomatic patients. 474 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 1, 2021. ; https://doi.org/10.1101/2021.09.30.21264236 doi: medRxiv preprint ESC Guidelines for the Diagnosis and Treatment of Acute 476 and Chronic Heart Failure Heart failure with mid-range ejection 478 fraction-a new category of heart failure or still a gray zone Universal Definition and Classification of Heart Failure: A Report of the Heart 480 Heart Failure Association of the European Society of Cardiology Heart Failure Society and Writing Committee of the Universal Definition of Heart Failure How to diagnose heart failure with preserved ejection fraction: the HFA-PEFF 484 diagnostic algorithm: a consensus recommendation from the Heart Failure Association (HFA) of the European Society of Cardiology (ESC) Diagnosis and management of heart failure with preserved ejection 487 fraction: 10 key lessons Epidemiology and clinical course of heart failure with 489 preserved ejection fraction Heart failure with preserved ejection fraction: diagnosis and 491 management Cellular and molecular differences between HFpEF and HFrEF: a step ahead 493 in an improved pathological understanding Principal Component Analysis Patterns of co-occurring comorbidities in people living with HIV Section 5: Management of Asymptomatic Patients With Reduced Left 498 Ventricular Ejection Fraction Natural history of asymptomatic left ventricular systolic dysfunction in the 500 community Role of global 502 longitudinal strain in assessment of left ventricular systolic function in patients with heart failure with 503 preserved ejection fraction How should we measure global and regional left ventricular systolic 505 function? Impaired systolic function by strain imaging in heart failure with 507 preserved ejection fraction Right Ventricular Damage in COVID-19: Association Between 509 Myocardial Injury and COVID-19 COVID-19 illness and heart failure: a missing link Heart failure subgroups: HFrEF, HFmrEF, and HFpEF with or without mitral 513 regurgitation Outcomes after aortic valve replacement in patients with severe aortic regurgitation 515 and markedly reduced left ventricular function A new, simplified and accurate method for determining ejection fraction with 517 two-dimensional echocardiography Characterization of static and dynamic left ventricular diastolic 519 function in patients with heart failure with a preserved ejection fraction Pulmonary hypertension in patients with heart failure and preserved ejection 521 fraction: differential diagnosis and management Association of systolic dysfunction with left ventricular hypertrophy and diastolic 523 dysfunction in hypertensive patients Obesity is associated with left atrial enlargement, E/A 526 reversal and left ventricular hypertrophy Mild systolic dysfunction in heart failure (left ventricular ejection fraction> 35%): 528 baseline characteristics, prognosis and response to therapy in the Vasodilator in Is mild asymptomatic left ventricular systolic dysfunction always predictive of adverse 531 events in high-risk populations? Insights from the DAVID-Berg study Tricuspid regurgitation in acute heart failure: is there any incremental risk? Differentiation of heart failure related to dilated cardiomyopathy and coronary 536 artery disease using gadolinium-enhanced cardiovascular magnetic resonance Prognostic implications of moderate aortic stenosis in patients with left ventricular 539 systolic dysfunction Atrial fibrillation and heart failure: treatment 541 considerations for a dual epidemic Electrocardiographic abnormalities in patients with heart 543 failure Is heart failure with mid range ejection fraction (HFmrEF) a distinct clinical entity or 545 an overlap group? Relation of 547 disease pathogenesis and risk factors to heart failure with preserved or reduced ejection fraction: Insights 548 from the framingham heart study of the national heart, lung, and blood institute Recommendations for cardiac chamber quantification by echocardiography in 551 adults: an update from the American Society of Echocardiography and the European Association of 552 Cardiovascular Imaging Diagnostic And Prognostic Implications of Heart Failure With Preserved Ejection 554 Fraction Scoring Systems A simple, evidence-based approach to help guide diagnosis of heart failure with 556 preserved ejection fraction Prognostic importance of pulmonary hypertension in patients with heart failure From left ventricular hypertrophy to congestive heart failure: 560 management of hypertensive heart disease Left ventricular hypertrophy patterns and incidence of heart failure with 562 preserved versus reduced ejection fraction Relationship between left atrial volume and diastolic dysfunction in 564 500 Brazilian patients Screening for cardiac contractile dysfunction using an artificial intelligence-enabled 566 electrocardiogram Identifying the most important ECG predictors of reduced ejection 568 fraction in patients with suspected acute coronary syndrome Epidemiology of heart failure Who are patients classified within the new terminology of heart failure from the 573 2016 ESC guidelines?