Abstract
Background
The relationship between aortic stenosis (AS) severity, AS-related symptoms, and clinical outcomes is poorly understood.
Objectives
The purpose of this study was to evaluate whether potential AS-related symptoms at diagnosis are associated with outcomes, including death and AS-related hospitalization.
Methods
In this retrospective cohort study from a large, integrated health care system serving >4.5 M individuals, we applied validated natural language processing algorithms to echocardiogram reports to identify physician-assessed AS severity and potential AS-related symptoms (eg, chest pain, syncope, dyspnea, worsening heart failure) via diagnosis codes and natural language processing-applied physician notes. Of 602,821 adults with echocardiograms from 2010 to 2019, we identified 40,333 adults diagnosed with AS and applied Cox models to examine associations between AS-related symptoms, AS severity, and clinical outcomes over a median follow-up of 2.2 years.
Results
Most patients with AS had potential AS-related symptoms (mild: 80%, mild-moderate: 77%, moderate: 77%, moderate-severe: 85%, severe: 87%). Symptomatic patients were older (mean age 78 vs 75 years; P < 0.01), more often female (51% vs 47%; P < 0.01), and had greater comorbidity burden. After multivariable adjustment, symptom status strongly predicted risk. Patients with symptomatic moderate AS had a similar risk to those with severe AS without symptoms (adjusted HR: 1.77 [95% CI: 1.65-1.91] vs 1.81 [95% CI: 1.51-2.17] for death, P = 0.81; aHR: 2.27 [95% CI: 2.13-2.41] vs 2.40 [95% CI: 2.08-2.77] for AS-related hospitalization, P = 0.42).
Conclusions
These findings suggest that symptom status, independent of AS severity, is a key risk factor for adverse outcomes. Further research is needed to assess the benefits of early intervention in these high-risk groups.
Key words: aortic stenosis, natural history, outcomes, valvular heart disease
Central Illustration
Aortic stenosis (AS) is the most prevalent heart valve condition globally, and its prevalence is expected to rise significantly in the upcoming decades due to aging populations.1, 2, 3, 4, 5, 6, 7, 8 Currently, there are no pharmaceutical treatments available for AS, and the primary management is through surgical aortic valve replacement or transcatheter aortic valve replacement (TAVR).9, 10, 11 The widespread adoption of the minimally invasive TAVR procedure has broadened the pool of individuals with AS eligible for AVR by reducing the risk of treatment and has opened the possibility of extending treatment to patients with less severe disease based on grade and/or symptom status.11,12 Simultaneously, diagnosing AS has become more complex, with newly identified clinical phenotypes beyond high-gradient severe AS.9,13, 14, 15, 16
Taken together, these advancements and recent work17, 18, 19, 20, 21 have challenged the original AS paradigm born from the landmark studies of Braunwald and Ross,22 which suggested that mortality for asymptomatic AS patients was similar to the general population but declined precipitously once symptoms attributed to AS surfaced. Thus, treatment for AS has historically necessitated the presence of severe disease plus AS-related symptoms. Although recent work has demonstrated that mortality from untreated AS may be significant even at lower disease severity levels,17, 18, 19, 20, 21,23, 24, 25 the complex relationship between AS severity and symptoms attributed to AS has been less well studied.24,26 This includes a gap in understanding: 1) the prevalence of documented potential AS-related symptoms across the spectrum of AS severity; and 2) the association between symptom status with subsequent morbidity and mortality at different levels of AS severity. Understanding this relationship could help improve risk stratification of AS and support more personalized approaches to AS management.
Using a previously validated database of AS patients17 within a large, integrated health care delivery system, we examined the prevalence of documented potential AS-related symptoms surrounding the time of AS diagnosis and the association of symptoms with clinical outcomes including AS-related hospitalization and all-cause death.
Methods
Source population
Kaiser Permanente Northern California (KPNC) is a large integrated health care delivery system providing comprehensive, outpatient, emergency department, and inpatient medical care to >4.5 million persons in northern and central California. The KPNC membership is highly representative of the local and state-wide population with regard to age, sex, race and ethnicity, and socioeconomic status. Nearly all aspects of care are captured through an integrated electronic health record (EHR) system, with additional data standardization for research in the Kaiser Permanente Virtual Data Warehouse.
The study was approved by the KPNC Institutional Review Board, and waiver of informed consent was obtained due to the nature of the study.
Study eligibility
We identified all adult (age ≥18 years) KPNC members diagnosed with AS by physician interpretation of an echocardiogram report between January 1, 2010, and December 31, 2019, using previously validated natural language processing (NLP) algorithms. We excluded patients with <6 months of continuous prior health plan membership, sex other than male or female, a known history of bicuspid or prior prosthetic aortic valves, or a left ventricular ejection fraction <50% or missing left ventricular ejection fraction on the qualifying echocardiogram report. We also excluded any patients identified by inpatient echocardiograms if their hospitalization episode ended in death or included an AVR procedure.
Follow-up and outcome variables
We set each patient’s index date as the date of the first qualifying echocardiogram report indicating AS during the study period. We obtained follow-up data through December 31, 2019. Patients were censored at the time of death, disenrollment from the health plan, or receipt of a prosthetic aortic valve (surgical aortic valve replacement or TAVR).
Our primary outcome of interest was all-cause death, identified using a combination of EHR data including member proxy reporting, State of California death certificate data, and Social Security vital status information. We also identified the secondary outcome of AS-related hospitalization, defined as any hospitalization with primary discharge diagnosis of AS, dyspnea, chest pain, stable angina, acute coronary syndrome or presyncope/syncope using validated diagnosis codes (International Classification of Diseases [ICD]-9/-10); validated NLP algorithms applied to physician notes to identify dyspnea, orthopnea, paroxysmal nocturnal dyspnea; or a hospitalization for heart failure or worsening heart failure based on previously validated methods.27,28
Covariates
We assessed for recently documented symptoms potentially attributable to AS at the index date by searching for a combination of diagnosis code-based categories (ICD-9/-10) and NLP-derived categories for symptoms associated with AS. We searched in the 365 days before and up to 60 days after the index date to allow for a broad-based search for potential symptoms before the AS diagnosis date, as well as a limited time period after diagnosis, to ensure the full capture of documented symptoms that may have been associated with a clinical episode that prompted the initial echocardiogram identifying incident AS.
We used diagnosis codes to identify dyspnea, syncope or dizziness, chest pain, and acute coronary syndrome including unstable angina and acute myocardial infarction (CP/ACS), and we used validated NLP algorithms applied to physician notes (ie, provider notes, discharge summaries, imaging reports) to identify dyspnea, orthopnea, paroxysmal nocturnal dyspnea, and worsening heart failure. We also obtained assessments of AS severity from echocardiogram reports using validated NLP algorithms, categorizing AS severity as mild, mild-moderate, moderate, moderate-severe, or severe, based upon the interpreting physician’s assessment. We used the combination of AS symptom status and AS severity at index date as the primary exposure in our analysis, with a total of 10 possible categories (2 categories of symptom status × 5 categories of AS severity).
In addition to AS status and severity, we extracted additional echocardiographic parameters of interest using NLP algorithms, including stroke volume, left ventricular hypertrophy, and end-diastolic diameter.17 We also obtained data on demographic characteristics, comorbidities, medication use, laboratory results, and vital sign measurements using relevant EHR data. A full list of definitions and industry-standard ICD-9/-10 and Current Procedural Terminology-4 codes is available upon request.
Analytic approach
All analyses were conducted using SAS software, version 9.4. We calculated crude rates of time-to-event per 100 person-years for AS-related hospitalization and all-cause death events. We plotted cumulative incidence curves by the combination of index AS symptoms and severity for AS-related hospitalization and all-cause death, with the utilization-based outcome also accounting for the competing risk of mortality. To address missing laboratory data and maintain the linear relationship between lab values, we conducted multiple imputation of missing data for blood pressure, cholesterol, and hemoglobin with 22 imputations.
We conducted multivariable Cox proportional hazards models for AS symptom status and severity on the 2 outcomes of interest for each of the 22 imputations, and then calculated summary estimates. Each model followed patients until the time to first event or censoring and included adjustment for the selected demographics, comorbidities, laboratory results, and echocardiographic parameters (ie, all variables reported below in our baseline) (Table 1).
Table 1.
Baseline Characteristics by Baseline Aortic Stenosis Severity and Aortic Stenosis-Related Symptom Status
| Symptomatic AS (n = 32,155) | Asymptomatic AS (n = 8,178) | Mild AS (n = 24,666) | Mild-Moderate AS (n = 3,053) | Moderate AS (n = 7,333) | Moderate-Severe AS (n = 1,684) | Severe AS (n = 3,597) | |
|---|---|---|---|---|---|---|---|
| Mean (SD) age, y | 78.3 (10.4) | 75.2 (10.7) | 76.8 (10.6) | 78.0 (10.1) | 78.9 (10.0) | 79.3 (10.2) | 79.8 (10.3) |
| Women, n (%) | 16,493 (51.3) | 3,836 (46.9) | 12,807 (51.9) | 1,471 (48.2) | 3,470 (47.3) | 809 (48.0) | 1,772 (49.3) |
| Race, n (%) | |||||||
| White | 22,342 (69.5) | 5,703 (69.7) | 16,502 (66.9) | 2,183 (71.5) | 5,353 (73.0) | 1,289 (76.5) | 2,718 (75.6) |
| Black | 1,736 (5.4) | 424 (5.2) | 1,479 (6.0) | 150 (4.9) | 327 (4.5) | 65 (3.9) | 139 (3.9) |
| Asian or Pacific Islander | 2,769 (8.6) | 779 (9.5) | 2,501 (10.1) | 251 (8.2) | 509 (6.9) | 92 (5.5) | 195 (5.4) |
| Native American | 121 (0.4) | 26 (0.3) | 89 (0.4) | 17 (0.6) | 23 (0.3) | 8 (0.5) | 10 (0.3) |
| Multiracial | 2,604 (8.1) | 505 (6.2) | 1,909 (7.7) | 214 (7.0) | 576 (7.9) | 113 (6.7) | 297 (8.3) |
| Unknown | 2,583 (8.0) | 741 (9.1) | 2,186 (8.9) | 238 (7.8) | 545 (7.4) | 117 (6.9) | 238 (6.6) |
| Hispanic ethnicity, n (%) | 4,014 (12.5) | 930 (11.4) | 3,176 (12.9) | 356 (11.7) | 843 (11.5) | 181 (10.7) | 388 (10.8) |
| Tobacco use, n (%) | |||||||
| Current smoker | 1,612 (5.0) | 465 (5.7) | 1,279 (5.2) | 165 (5.4) | 345 (4.7) | 86 (5.1) | 202 (5.6) |
| Former smoker | 14,863 (46.2) | 3,266 (39.9) | 11,048 (44.8) | 1,382 (45.3) | 3,406 (46.4) | 748 (44.4) | 1,545 (43.0) |
| Passive smoker | 125 (0.4) | 37 (0.5) | 101 (0.4) | 6 (0.2) | 31 (0.4) | 7 (0.4) | 17 (0.5) |
| No tobacco use | 15,555 (48.4) | 4,410 (53.9) | 12,238 (49.6) | 1,500 (49.1) | 3,551 (48.4) | 843 (50.1) | 1,833 (51.0) |
| Alcohol use, n (%) | 10,760 (33.5) | 3,008 (36.8) | 8,532 (34.6) | 1,060 (34.7) | 2,453 (33.5) | 573 (34.0) | 1,150 (32.0) |
| Illicit drug use, n (%) | 473 (1.5) | 109 (1.3) | 385 (1.6) | 43 (1.4) | 91 (1.2) | 21 (1.2) | 42 (1.2) |
| Index year, n (%) | |||||||
| 2010 | 4,458 (13.9) | 1,146 (14.0) | 3,101 (12.6) | 265 (8.7) | 1,246 (17.0) | 259 (15.4) | 733 (20.4) |
| 2011 | 3,469 (10.8) | 946 (11.6) | 2,467 (10.0) | 288 (9.4) | 927 (12.6) | 223 (13.2) | 510 (14.2) |
| 2012 | 3,009 (9.4) | 798 (9.8) | 2,212 (9.0) | 296 (9.7) | 726 (9.9) | 192 (11.4) | 381 (10.6) |
| 2013 | 2,812 (8.7) | 713 (8.7) | 2,144 (8.7) | 301 (9.9) | 593 (8.1) | 167 (9.9) | 320 (8.9) |
| 2014 | 2,710 (8.4) | 633 (7.7) | 2,071 (8.4) | 288 (9.4) | 554 (7.6) | 146 (8.7) | 284 (7.9) |
| 2015 | 2,913 (9.1) | 712 (8.7) | 2,288 (9.3) | 307 (10.1) | 592 (8.1) | 147 (8.7) | 291 (8.1) |
| 2016 | 3,005 (9.3) | 753 (9.2) | 2,455 (10.0) | 268 (8.8) | 628 (8.6) | 143 (8.5) | 264 (7.3) |
| 2017 | 3,180 (9.9) | 822 (10.1) | 2,592 (10.5) | 330 (10.8) | 679 (9.3) | 118 (7.0) | 283 (7.9) |
| 2018 | 3,312 (10.3) | 830 (10.1) | 2,688 (10.9) | 349 (11.4) | 707 (9.6) | 139 (8.3) | 259 (7.2) |
| 2019 | 3,287 (10.2) | 825 (10.1) | 2,648 (10.7) | 361 (11.8) | 681 (9.3) | 150 (8.9) | 272 (7.6) |
| Medical history, n (%) | |||||||
| Heart failure | 7,686 (23.9) | 510 (6.2) | 5,128 (20.8) | 591 (19.4) | 1,377 (18.8) | 349 (20.7) | 751 (20.9) |
| Atrial fibrillation | 8,402 (26.1) | 1,056 (12.9) | 5,982 (24.3) | 653 (21.4) | 1,641 (22.4) | 397 (23.6) | 785 (21.8) |
| Atrial flutter | 1,165 (3.6) | 128 (1.6) | 865 (3.5) | 95 (3.1) | 197 (2.7) | 54 (3.2) | 82 (2.3) |
| Ventricular fibrillation | 39 (0.1) | 3 (0.0) | 26 (0.1) | 5 (0.2) | 9 (0.1) | 0 (0.0) | 2 (0.1) |
| Ventricular tachycardia | 195 (0.6) | 11 (0.1) | 149 (0.6) | 17 (0.6) | 31 (0.4) | 3 (0.2) | 6 (0.2) |
| Ischemic stroke | 1,173 (3.6) | 189 (2.3) | 870 (3.5) | 110 (3.6) | 220 (3.0) | 52 (3.1) | 110 (3.1) |
| Transient ischemic attack | 1,525 (4.7) | 208 (2.5) | 1,106 (4.5) | 121 (4.0) | 270 (3.7) | 71 (4.2) | 165 (4.6) |
| Peripheral artery disease | 2,613 (8.1) | 331 (4.0) | 1,906 (7.7) | 223 (7.3) | 532 (7.3) | 91 (5.4) | 192 (5.3) |
| Venous thromboembolism | 1,776 (5.5) | 152 (1.9) | 1,322 (5.4) | 142 (4.7) | 283 (3.9) | 69 (4.1) | 112 (3.1) |
| Nonvenous thromboembolism | 509 (1.6) | 70 (0.9) | 361 (1.5) | 51 (1.7) | 100 (1.4) | 24 (1.4) | 43 (1.2) |
| Diabetes mellitus | 13,424 (41.7) | 2,884 (35.3) | 10,247 (41.5) | 1,256 (41.1) | 2,906 (39.6) | 659 (39.1) | 1,240 (34.5) |
| Hypertension | 27,969 (87.0) | 6,563 (80.3) | 21,292 (86.3) | 2,624 (85.9) | 6,209 (84.7) | 1,445 (85.8) | 2,962 (82.3) |
| Dyslipidemia | 27,249 (84.7) | 6,490 (79.4) | 20,715 (84.0) | 2,598 (85.1) | 6,101 (83.2) | 1,405 (83.4) | 2,920 (81.2) |
| Hyperthyroidism | 1,518 (4.7) | 333 (4.1) | 1,167 (4.7) | 133 (4.4) | 335 (4.6) | 71 (4.2) | 145 (4.0) |
| Hypothyroidism | 6,983 (21.7) | 1,412 (17.3) | 5,104 (20.7) | 628 (20.6) | 1,569 (21.4) | 344 (20.4) | 750 (20.9) |
| Chronic liver disease | 1,732 (5.4) | 381 (4.7) | 1,457 (5.9) | 136 (4.5) | 331 (4.5) | 70 (4.2) | 119 (3.3) |
| Chronic lung disease | 12,806 (39.8) | 1,965 (24.0) | 9,442 (38.3) | 1,082 (35.4) | 2,544 (34.7) | 568 (33.7) | 1,135 (31.6) |
| Diagnosed depression | 6,213 (19.3) | 1,079 (13.2) | 4,654 (18.9) | 555 (18.2) | 1,230 (16.8) | 267 (15.9) | 586 (16.3) |
| Diagnosed dementia | 2,624 (8.2) | 428 (5.2) | 1,745 (7.1) | 249 (8.2) | 576 (7.9) | 144 (8.6) | 338 (9.4) |
| Hospitalized bleed | 2,040 (6.3) | 249 (3.0) | 1,347 (5.5) | 173 (5.7) | 451 (6.2) | 93 (5.5) | 225 (6.3) |
| Coronary artery bypass graft | 508 (1.6) | 28 (0.3) | 394 (1.6) | 49 (1.6) | 69 (0.9) | 9 (0.5) | 15 (0.4) |
| Percutaneous coronary intervention | 2,622 (8.2) | 229 (2.8) | 1,895 (7.7) | 228 (7.5) | 460 (6.3) | 111 (6.6) | 157 (4.4) |
| Frailty | 510 (1.6) | 59 (0.7) | 389 (1.6) | 44 (1.4) | 76 (1.0) | 24 (1.4) | 36 (1.0) |
| Hearing impairment | 11,496 (35.8) | 2,298 (28.1) | 8,432 (34.2) | 1,055 (34.6) | 2,526 (34.4) | 555 (33.0) | 1,226 (34.1) |
| Visual impairment | 28,129 (87.5) | 6,808 (83.2) | 21,510 (87.2) | 2,609 (85.5) | 6,366 (86.8) | 1,437 (85.3) | 3,015 (83.8) |
| Charlson Comorbidity Index, mean (SD) | 3.4 (2.6) | 2.0 (2.1) | 3.2 (2.6) | 3.1 (2.5) | 3.0 (2.5) | 2.9 (2.4) | 2.7 (2.4) |
| Body mass index, kg/m2, n (%) | |||||||
| <18.5 | 614 (1.9) | 105 (1.3) | 413 (1.7) | 50 (1.6) | 137 (1.9) | 29 (1.7) | 90 (2.5) |
| 18.5-24.9 | 8,384 (26.1) | 2,100 (25.7) | 6,167 (25.0) | 749 (24.5) | 2,001 (27.3) | 492 (29.2) | 1,075 (29.9) |
| 25.0-29.9 | 10,465 (32.5) | 2,903 (35.5) | 7,980 (32.4) | 1,094 (35.8) | 2,489 (33.9) | 572 (34.0) | 1,233 (34.3) |
| 30.0-39.9 | 10,141 (31.5) | 2,616 (32.0) | 8,072 (32.7) | 947 (31.0) | 2,266 (30.9) | 495 (29.4) | 977 (27.2) |
| ≥40.0 | 2,427 (7.5) | 421 (5.1) | 1,952 (7.9) | 195 (6.4) | 413 (5.6) | 85 (5.0) | 203 (5.6) |
| Unknown | 124 (0.4) | 33 (0.4) | 82 (0.3) | 18 (0.6) | 27 (0.4) | 11 (0.7) | 19 (0.5) |
| Systolic blood pressure, mm Hg, n (%) | |||||||
| <120 | 8,378 (26.1) | 1,689 (20.7) | 5,858 (23.7) | 721 (23.6) | 1,921 (26.2) | 454 (27.0) | 1,113 (30.9) |
| 120-129 | 5,480 (17.0) | 1,500 (18.3) | 4,205 (17.0) | 519 (17.0) | 1,268 (17.3) | 316 (18.8) | 672 (18.7) |
| 130-139 | 7,710 (24.0) | 2,238 (27.4) | 6,100 (24.7) | 761 (24.9) | 1,817 (24.8) | 425 (25.2) | 845 (23.5) |
| 140-159 | 7,046 (21.9) | 1,934 (23.6) | 5,635 (22.8) | 738 (24.2) | 1,600 (21.8) | 335 (19.9) | 672 (18.7) |
| 160-179 | 2,486 (7.7) | 603 (7.4) | 2,014 (8.2) | 211 (6.9) | 540 (7.4) | 110 (6.5) | 214 (5.9) |
| ≥180 | 951 (3.0) | 192 (2.3) | 777 (3.2) | 94 (3.1) | 171 (2.3) | 37 (2.2) | 64 (1.8) |
| Unknown | 104 (0.3) | 22 (0.3) | 77 (0.3) | 9 (0.3) | 16 (0.2) | 7 (0.4) | 17 (0.5) |
| Diastolic blood pressure, mm Hg, n (%) | |||||||
| <80 | 26,925 (83.7) | 6,654 (81.4) | 20,421 (82.8) | 2,530 (82.9) | 6,156 (83.9) | 1,408 (83.6) | 3,064 (85.2) |
| 81-84 | 1,841 (5.7) | 615 (7.5) | 1,510 (6.1) | 196 (6.4) | 453 (6.2) | 107 (6.4) | 190 (5.3) |
| 85-89 | 1,575 (4.9) | 458 (5.6) | 1,280 (5.2) | 155 (5.1) | 363 (5.0) | 78 (4.6) | 157 (4.4) |
| 90-99 | 1,241 (3.9) | 336 (4.1) | 1,025 (4.2) | 115 (3.8) | 247 (3.4) | 57 (3.4) | 133 (3.7) |
| 100-109 | 316 (1.0) | 66 (0.8) | 237 (1.0) | 34 (1.1) | 62 (0.8) | 18 (1.1) | 31 (0.9) |
| ≥110 | 153 (0.5) | 27 (0.3) | 116 (0.5) | 14 (0.5) | 36 (0.5) | 9 (0.5) | 5 (0.1) |
| Unknown | 104 (0.3) | 22 (0.3) | 77 (0.3) | 9 (0.3) | 16 (0.2) | 7 (0.4) | 17 (0.5) |
| Hemoglobin, g/dL, n (%) | |||||||
| <9.0 | 1,032 (3.2) | 60 (0.7) | 725 (2.9) | 90 (2.9) | 155 (2.1) | 40 (2.4) | 82 (2.3) |
| 9.0-9.9 | 1,461 (4.5) | 130 (1.6) | 1,035 (4.2) | 124 (4.1) | 260 (3.5) | 64 (3.8) | 108 (3.0) |
| 10.0-10.9 | 2,646 (8.2) | 358 (4.4) | 1,949 (7.9) | 227 (7.4) | 498 (6.8) | 112 (6.7) | 218 (6.1) |
| 11.0-11.9 | 4,413 (13.7) | 809 (9.9) | 3,198 (13.0) | 371 (12.2) | 962 (13.1) | 226 (13.4) | 465 (12.9) |
| 12.0-12.9 | 5,864 (18.2) | 1,349 (16.5) | 4,477 (18.2) | 521 (17.1) | 1,281 (17.5) | 307 (18.2) | 627 (17.4) |
| 13.0-13.9 | 6,225 (19.4) | 1,803 (22.0) | 4,867 (19.7) | 643 (21.1) | 1,484 (20.2) | 320 (19.0) | 714 (19.8) |
| ≥14.0 | 6,933 (21.6) | 2,464 (30.1) | 5,694 (23.1) | 689 (22.6) | 1,767 (24.1) | 395 (23.5) | 852 (23.7) |
| Unknown | 3,581 (11.1) | 1,205 (14.7) | 2,721 (11.0) | 388 (12.7) | 926 (12.6) | 220 (13.1) | 531 (14.8) |
| High-density lipoprotein, mg/dL, n (%) | |||||||
| <35 | 2,742 (8.5) | 524 (6.4) | 2,072 (8.4) | 254 (8.3) | 552 (7.5) | 140 (8.3) | 248 (6.9) |
| 35-39 | 3,022 (9.4) | 734 (9.0) | 2,349 (9.5) | 266 (8.7) | 669 (9.1) | 169 (10.0) | 303 (8.4) |
| 40-49 | 7,228 (22.5) | 1,939 (23.7) | 5,591 (22.7) | 703 (23.0) | 1,666 (22.7) | 385 (22.9) | 822 (22.9) |
| 50-59 | 5,674 (17.6) | 1,701 (20.8) | 4,453 (18.1) | 526 (17.2) | 1,400 (19.1) | 296 (17.6) | 700 (19.5) |
| ≥60 | 6,090 (18.9) | 1,832 (22.4) | 4,847 (19.7) | 618 (20.2) | 1,425 (19.4) | 330 (19.6) | 702 (19.5) |
| Unknown | 7,399 (23.0) | 1,448 (17.7) | 5,354 (21.7) | 686 (22.5) | 1,621 (22.1) | 364 (21.6) | 822 (22.9) |
| Low-density lipoprotein, mg/dL, n (%) | |||||||
| <70 | 7,062 (22.0) | 1,363 (16.7) | 5,248 (21.3) | 631 (20.7) | 1,478 (20.2) | 351 (20.8) | 717 (19.9) |
| 70-99 | 9,712 (30.2) | 2,553 (31.2) | 7,429 (30.1) | 936 (30.7) | 2,276 (31.0) | 528 (31.4) | 1,096 (30.5) |
| 100-129 | 5,408 (16.8) | 1,665 (20.4) | 4,337 (17.6) | 516 (16.9) | 1,301 (17.7) | 279 (16.6) | 640 (17.8) |
| 130-159 | 2,278 (7.1) | 801 (9.8) | 1,861 (7.5) | 208 (6.8) | 578 (7.9) | 136 (8.1) | 296 (8.2) |
| 160-199 | 921 (2.9) | 295 (3.6) | 734 (3.0) | 86 (2.8) | 230 (3.1) | 54 (3.2) | 112 (3.1) |
| ≥200 | 244 (0.8) | 57 (0.7) | 180 (0.7) | 31 (1.0) | 53 (0.7) | 11 (0.7) | 26 (0.7) |
| Unknown | 6,530 (20.3) | 1,444 (17.7) | 4,877 (19.8) | 645 (21.1) | 1,417 (19.3) | 325 (19.3) | 710 (19.7) |
| Estimated glomerular filtration rate, mL/min/1.73 m2, n (%) | |||||||
| >150 | 2 (0.0) | 0 (0.0) | 2 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) |
| 90-150 | 2,567 (8.0) | 1,028 (12.6) | 2,366 (9.6) | 274 (9.0) | 601 (8.2) | 119 (7.1) | 235 (6.5) |
| 60-89 | 13,960 (43.4) | 4,230 (51.7) | 10,974 (44.5) | 1,411 (46.2) | 3,334 (45.5) | 768 (45.6) | 1,703 (47.3) |
| 45-59 | 6,799 (21.1) | 1,498 (18.3) | 4,969 (20.1) | 606 (19.8) | 1,596 (21.8) | 349 (20.7) | 777 (21.6) |
| 30-44 | 4,445 (13.8) | 694 (8.5) | 3,103 (12.6) | 388 (12.7) | 955 (13.0) | 237 (14.1) | 456 (12.7) |
| 15-29 | 1,702 (5.3) | 218 (2.7) | 1,249 (5.1) | 136 (4.5) | 321 (4.4) | 82 (4.9) | 132 (3.7) |
| <15 | 574 (1.8) | 57 (0.7) | 429 (1.7) | 43 (1.4) | 103 (1.4) | 18 (1.1) | 38 (1.1) |
| Dialysis | 466 (1.4) | 47 (0.6) | 380 (1.5) | 38 (1.2) | 64 (0.9) | 11 (0.7) | 20 (0.6) |
| Transplant | 262 (0.8) | 42 (0.5) | 218 (0.9) | 26 (0.9) | 35 (0.5) | 10 (0.6) | 15 (0.4) |
| Unknown | 1,378 (4.3) | 364 (4.5) | 976 (4.0) | 131 (4.3) | 324 (4.4) | 90 (5.3) | 221 (6.1) |
| Left ventricular ejection fraction, %, mean (SD) | 61.5 (5.0) | 61.7 (4.6) | 61.5 (4.9) | 61.6 (4.8) | 61.7 (5.0) | 61.5 (5.1) | 61.4 (4.8) |
| Stroke volume, mL, mean (SD) | 79.7 (23.2) | 82.8 (22.5) | 81.3 (23.0) | 79.8 (23.5) | 80.7 (22.9) | 76.5 (23.2) | 75.0 (22.5) |
| Left ventricular hypertrophy, n (%) | |||||||
| Normal | 6,921 (21.5) | 2,094 (25.6) | 6,142 (24.9) | 733 (24.0) | 1,450 (19.8) | 263 (15.6) | 427 (11.9) |
| Mild | 11,373 (35.4) | 2,873 (35.1) | 8,263 (33.5) | 1,131 (37.0) | 2,802 (38.2) | 702 (41.7) | 1,348 (37.5) |
| Mild-moderate | 1,258 (3.9) | 275 (3.4) | 812 (3.3) | 145 (4.7) | 300 (4.1) | 85 (5.0) | 191 (5.3) |
| Moderate | 3,757 (11.7) | 787 (9.6) | 2,221 (9.0) | 292 (9.6) | 978 (13.3) | 262 (15.6) | 791 (22.0) |
| Moderate-severe | 271 (0.8) | 55 (0.7) | 158 (0.6) | 19 (0.6) | 54 (0.7) | 23 (1.4) | 72 (2.0) |
| Severe | 597 (1.9) | 91 (1.1) | 320 (1.3) | 40 (1.3) | 120 (1.6) | 44 (2.6) | 164 (4.6) |
| Unknown | 7,978 (24.8) | 2,003 (24.5) | 6,750 (27.4) | 693 (22.7) | 1,629 (22.2) | 305 (18.1) | 604 (16.8) |
AS = aortic stenosis.
In sensitivity analyses, we examined outcomes for matched patients with echocardiograms between January 1, 2010, and December 31, 2019, who did not have AS in a 1:1 ratio to our cohort of AS patients based on age (±3 years) at index date and sex.
Results
Study population and baseline characteristics
Among 602,821 adults with echocardiograms, we identified 40,333 adults with incident AS with a median (IQR) follow-up of 2.2 (0.8-4.3) years (Table 1). The breakdown of AS severity was as follows: 24,666 (61.2%) mild, 3,053 (7.6%) mild-moderate, 7,333 (18.2%) moderate, 1,684 (4.2%) moderate-severe, and 3,597 (8.9%) severe. Patients with recent symptoms were older (mean [SD] age 78 [10] vs 75 [11] years; P < 0.01), more likely female (51% vs 47%; P < 0.01), and had a higher comorbidity burden.
Presence and type of potential AS-related symptoms surrounding index AS diagnosis
Regardless of AS severity level, most patients had recently documented potential AS-related symptoms (mild: 80%, mild-moderate: 77%, moderate: 77%, moderate-severe: 85%, and severe AS: 87%) (Figure 1). Symptoms were broadly distributed across the categories defining AS-related symptoms, with a modest symptom burden among the codes-based diagnoses of CP/ACS, dizziness/syncope, and dyspnea. But the largest category of symptoms was identified through the NLP-based evaluation of heart failure and worsening signs and symptoms of heart failure from provider notes.
Figure 1.
Potential Aortic Stenosis-Related Symptoms Surrounding Aortic Stenosis Diagnosis
AS-related symptoms defined as any mention of dyspnea, orthopnea, paroxysmal nocturnal dyspnea, or worsening heart failure by the application of validated natural language processing techniques to physician notes, or dyspnea, syncope, dizziness or presyncope, chest pain, acute myocardial infarction, or unstable angina by validated algorithms using diagnosis and procedure codes. AS = aortic stenosis; CP/ACS = Chest pain or acute coronary syndrome; HF = heart failure; NLP = natural language processing.
AS severity, symptom status and outcomes
In unadjusted analyses, symptom status was a strong discriminator of risk for adverse outcomes across AS severity levels (Figure 2). After classifying patients into mutually exclusive categories of AS severity levels and symptom status, at each level of AS severity, outcomes were less frequent for patients without documented symptoms. Furthermore, over the first 5 years of follow-up, only severe AS patients without documented symptoms had death rates that approached levels of symptomatic patients of any AS severity level.
Figure 2.
Cumulative Incidence for Clinical Outcomes by Aortic Stenosis Severity and Symptoms
(A) Cumulative incidence of all-cause death, by aortic stenosis severity and symptoms. (B) Cumulative incidence of aortic stenosis-related hospitalization, by aortic stenosis severity and symptoms. Abbreviation as in Figure 1.
In multivariable models that examined the adjusted risk of outcomes by each severity-symptom status category (reference category mild, asymptomatic AS patients), AS-related symptoms remained strongly associated with AS-related hospitalization and all-cause death across the spectrum of AS severity (Figure 3; Central Illustration). Moderate symptomatic patients had risk comparable to severe patients without documented symptoms (adjusted HR: 1.77 [95% CI: 1.65-1.91] vs aHR: 1.81 [95% CI: 1.51-2.17] for all-cause death; P = 0.81; and aHR: 2.27 [95% CI: 2.13-2.41] vs aHR: 2.40 [95% CI: 2.08-2.77] for AS-related hospitalization; P = 0.42). Even mild-moderate symptomatic AS patients had risk similar to, though slightly lower in magnitude, than severe asymptomatic AS patients for the outcome of all-cause death (aHR: 1.57 [95% CI: 1.44-1.72] vs aHR: 1.81 [95% CI: 1.51-2.17]; P = 0.13), though a lower risk of AS-related hospitalization (aHR: 1.99 [95% CI: 1.84-2.15] vs aHR: 2.40 [95% CI: 2.08-2.77]; P = 0.013).
Figure 3.
Multivariable Results for the Association of Aortic Stenosis Severity and Potential Symptoms With Clinical Outcomes
(A) Association of aortic stenosis severity and potential symptoms with all-cause death. (B) Association of aortic stenosis severity and potential symptoms with aortic stenosis-related hospitalization. All models were also adjusted for age, sex, race and ethnicity, index year, left ventricular ejection fraction, stroke volume, end-diastolic diameter, left ventricular hypertrophy, tobacco use, alcohol use, illicit drug use, Charlson Comorbidity Index, atrial fibrillation, atrial flutter, ventricular fibrillation, ventricular tachycardia, ischemic stroke, transient ischemic attack, peripheral artery disease, venous thromboembolism, nonvenous thromboembolism, diabetes mellitus, hypertension, dyslipidemia, hyperthyroidism, hypothyroidism, chronic liver disease, chronic lung disease, diagnosed depression, diagnosed dementia, hospitalized bleed, coronary artery bypass graft, percutaneous coronary intervention, frailty, hearing impairment, visual impairment, body mass index, systolic blood pressure, diastolic blood pressure, hemoglobin, high-density lipoprotein, low-density lipoprotein, and estimated glomerular filtration rate.
Central Illustration.
Association of Aortic Stenosis Severity and Potential Aortic Stenosis-Related Symptoms With All-Cause Death
Results stratified by potential AS-related symptom status surrounding the time of diagnosis. Patients with symptomatic moderate AS had similarly poor outcomes compared to patients with severe AS without documented symptoms. Model adjusted for competing risks using Fine-Gray estimates. Abbreviation as in Figure 1.
Sensitivity analyses
In sensitivity analyses, age- and sex-matched patients without AS on echocardiogram had a lower comorbidity burden compared to patients with symptomatic AS (Supplemental Table 2). However, patients with no AS had a similar frequency of potential AS-related symptoms compared to patients with AS (Supplemental Figure 3), including both code-based symptom categories and NLP-based symptom category of heart failure and worsening signs and symptoms of heart failure from provider notes. In unadjusted analyses, patients with no AS had lower rates of all-cause death compared to symptomatic AS patients of any severity level (Supplemental Figure 4).
Discussion
In a large, real-world cohort of > 40,000 patients with AS, we assessed whether patients had recent symptoms potentially attributable to AS using a combination of methods including diagnosis and procedure codes and NLP-based approaches and assessed the associated risk of potential AS-related symptoms with adverse outcomes across the spectrum of AS severity levels. We found that the great majority of AS patients—approximately 80%—had recent or active symptoms potentially attributable to AS at the time of diagnosis, regardless of the level of AS severity. Furthermore, a recent history of potential AS-related symptoms was associated with a significantly increased risk of adverse outcomes, including AS-related hospitalization and all-cause death, at all levels of AS severity. The presence of recent symptoms was a considerable risk stratifier for AS patients, and after multivariable adjustment, those with moderate symptomatic AS had outcomes similar to those with severe AS without documented symptoms. Our findings suggest that: 1) a greater proportion of AS patients may have symptoms potentially attributable to AS than previous work suggests; 2) the risk of morbidity and mortality increases with worsening AS; and 3) symptom status is correlated with significant morbidity and mortality.
Poor clinical outcomes for patients with severe AS have been demonstrated since the seminal studies of Braunwald and Ross,22 but recent work has differed on the prognosis and outcomes for AS of lesser severity levels.17, 18, 19, 20, 21 Our prior work suggested that moderate AS represents a wide spectrum of patients, with significant and increasing rates of adverse outcomes as AS severity increased, but with varying outcomes across the range of moderate AS patients that were distinct from either mild or severe AS patients.17 The VALVENOR (Suivi d'une Cohorte de Patients Présentant une Sténose Valvulaire Aortique en Région Nord-Pas-de-Calais) study group, who prospectively followed AS patients in routine clinical practice in northern France, also demonstrated that symptoms status, measured through NYHA classifications at follow-up visits, was a strong risk discriminator and that moderate AS patients, particularly if asymptomatic, had demonstrably lower mortality rates compared to severe AS.21 In that study, the proportion of symptomatic AS patients (NYHA functional class ≥II) was lower than in ours, with 54% of mild AS, 64% of moderate AS, and 70% of their severe AS cohort having symptoms attributable to AS. And others have proposed staging systems for AS29 based upon the extent of cardiac damage, which suggests varying phenotypes may be present with different clinical trajectories across the traditional severity spectrum. Moderate AS patients with symptoms may represent a very different phenotype compared to those that develop high gradients without symptoms.
Although when assessing AS patients for symptoms, care must be taken to delineate whether symptoms are driven by AS or an alternate condition,30 some studies have suggested that only half of moderate or severe AS patients report attributable symptoms.31,32 Other work has found that 76% of severe AS patients present with symptoms attributable to AS,5 but that patients often delay reporting their symptoms.33 Our findings add to this evidence base, as many prior studies were limited by smaller sample sizes, variable assessments of AS severity, and a lack of an evaluation of potential AS symptoms across the entire spectrum of AS severity. Given the generally low adverse event rates for asymptomatic AS patients under medical management20,21,34,35 and the heightened risk of adverse outcomes associated with symptoms, rigorous follow-up and symptoms assessment is critical for AS patients.
Prior studies have also not used advanced EHR searches that combined validated NLP-based reviews of physician notes in combination with code-based algorithms to ascertain potential AS symptom status. Our work indicates that broad-based searches of the EHR may identify patients with symptoms attributable to AS that could be missed in regular clinical practice. It is well-known that patient histories can be incomplete or shift between providers36,37 and that few patients with AS are objectively tested for symptoms with procedures like exercise treadmill testing. Given that the current indications for AVR hinges upon the presence or absence of symptoms9 and that our study and related work21 demonstrates the prognostic importance of symptoms, a comprehensive evaluation for potential AS symptoms is critical to avoiding preventable morbidity and mortality.
We found that the vast majority of AS patients have clinical encounters in the EHR that suggest symptoms potentially attributable to AS. Interestingly, in sensitivity analyses that examined age- and sex-matched patients undergoing echocardiography, nearly as many patients among those without documented AS had a similar distribution of symptoms, suggesting that referral for echocardiography is likely often triggered by a concerning cardiac symptom. Though symptom assessment is subjective and varies by physician, our study underscores that there are few patients referred for echocardiography who may not manifest potential AS-related symptoms. Furthermore, we found the most common AS-related symptom across every category of AS severity was NLP-derived heart failure symptoms. One can imagine a future where implementing AI techniques that scour the EHR for potential AS-related symptoms could better identify those who may benefit from a more detailed evaluation to assess the need for AVR, and algorithms that search for symptoms using NLP or other artificial intelligence techniques may be superior to approaches based upon traditional diagnosis codes alone.
Our results shed light on the real-world outcomes for a broad population of patients with AS of varying severity, but, due to the variation in provider assessments of symptoms and many factors involved in clinical decision-making, do not necessarily indicate which populations or subpopulations would clearly benefit from AVR. Randomized trials have recently examined whether AVR improves outcomes for asymptomatic severe AS patients (EARLY TAVR [Evaluation of TAVR Compared to Surveillance for Patients With Asymptomatic Severe Aortic Stenosis;NCT03042104]),12,38 and are ongoing to assess whether AVR is beneficial for eligible moderate AS patients (PROGRESS [Prospective, Randomized, Controlled Trial to Assess the Management of Moderate Aortic Stenosis by Clinical Surveillance or Transcatheter Aortic Valve Replacement; NCT04889872]). Though our study suggests significant morbidity and mortality in these patient populations, the findings from these ongoing trials will provide insight whether we can improve outcomes for these vulnerable patients.
Study Limitations
Our study has several limitations. First, as a retrospective study, we cannot rule-out residual or unmeasured confounding. However, our richly detailed clinical database allows us to capture data on a patient’s entire health care journey, including comorbid conditions, medications received, laboratory results, and vital signs, allowing us to control for a large number of variables across diverse domains of care that may affect outcomes.
Second, it is possible that the symptoms we classified as AS-related may not be due the pathophysiologic sequelae of AS but rather to other comorbid conditions. However, the physician in the exam room can be similarly confounded, as determining AS symptoms is subjective in nature, and it is ultimately the physician’s best judgment to decide whether symptoms are attributable to AS. In today’s modern practice, physicians may not have the time to review all of the information fully and systematically in the rapidly expanding EHR, and computer-assisted symptoms assessment could provide useful clinical decision support. Conversely, it is also possible that our methods could have missed some AS-related symptoms in the asymptomatic group. Since current treatment thresholds depend upon the presence of symptoms, it makes it difficult to discern who should or should not be treated. Thus, more granular treatment indications warrant further development, and ongoing clinical trials will shed further light on the specific subgroups that may benefit from AVR.
Third, we did not examine concomitant valvular disease or do core lab review of echocardiograms, but our cohort and methodologies were developed and validated rigorously in prior studies.17,27,39 Thus, our data reflect a real-world assessment of AS in the community and its correlation with outcomes. Finally, although our data are drawn from a large and diverse population, reflecting the practice of a large, health care delivery system caring for nearly 1.5% of the United States, our findings may not be fully generalizable to other geographic areas or practice settings.
Conclusions
In summary, in a large, real-world cohort of patients with incident AS, we found that the majority of AS patients have symptoms potentially attributable to AS (as many as 80% of patients) regardless of the echocardiographic-based severity of the AS, outcomes worsened with worsening AS, and the presence of symptoms considerably increased the risk for the adverse outcomes of AS-related hospitalization and all-cause death. These results highlight the importance of rigorously screening AS patients for symptoms, which greatly affects their disease trajectory, and the potential for EHR-based clinical decision support tools to flag high-risk patients for evaluation. We also found that symptomatic patients with moderate AS have risk comparable to asymptomatic patients with severe AS. Forthcoming studies will clarify if early intervention with TAVR improves outcomes for either of these groups, or for subpopulations within them. In summary, incorporating potential AS symptom status across the individual grades of AS may further improve risk stratification beyond AS severity alone.
Perspectives.
COMPETENCY IN PRACTICE-BASED LEARNING: This study highlights the importance of recognizing symptoms in AS patients, showing that symptom status—regardless of echocardiographic severity—is strongly linked to worse outcomes. It underscores the need for routine symptom screening in clinical practice. The comparable risk between symptomatic moderate AS and asymptomatic severe AS challenges current severity-based approaches and supports ongoing learning to refine clinical decision-making.
TRANSLATIONAL OUTLOOK: This study supports the development of EHR-integrated clinical decision support tools that can identify potential AS symptoms to identify high-risk AS patients earlier and more comprehensively. It also lays the groundwork for prospective trials evaluating whether earlier intervention, such as TAVR in symptomatic moderate AS, improves outcomes, and encourages more nuanced risk stratification models that blend clinical symptoms, imaging, and real-world outcomes.
Funding support and author disclosures
This work was funded by an investigator-initiated grant from Edwards Lifesciences, Inc. Dr Solomon has received research funding through his institution from Edwards Lifesciences, LLC, Bristol Myers Squibb, The Permanente Medical Group, Northern California Community Benefits Programs, and the Garfield Memorial Fund. Dr Go has received research funding through his institution from National Heart, Lung, and Blood Institute, National Institute of Diabetes and Digestive and Kidney Diseases, Bristol Myers Squibb, and Novartis. Dr Ambrosy has received relevant research support through grants to his institution from the National Heart, Lung, and Blood Institute (K23HL150159), the American Heart Association, The Permanente Medical Group, Northern California Community Benefits Programs, Garfield Memorial Fund, Abbott Laboratories, Amarin Pharma, Inc, Bayer AG, Edwards Lifesciences LLC, Esperion Therapeutics, Inc, and Novartis. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
Footnotes
The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the Author Center.
Appendix
For supplemental tables and figures, please see the online version of this paper.
Supplementary data
References
- 1.Stewart B.F., Siscovick D., Lind B.K., et al. Clinical factors associated with calcific aortic valve disease. Cardiovascular Health Study. J Am Coll Cardiol. 1997;29(3):630–634. doi: 10.1016/s0735-1097(96)00563-3. [DOI] [PubMed] [Google Scholar]
- 2.Martinsson A., Li X., Andersson C., Nilsson J., Smith J.G., Sundquist K. Temporal trends in the incidence and prognosis of aortic stenosis: a nationwide study of the Swedish population. Circulation. 2015;131(11):988–994. doi: 10.1161/CIRCULATIONAHA.114.012906. [DOI] [PubMed] [Google Scholar]
- 3.d'Arcy J.L., Coffey S., Loudon M.A., et al. Large-scale community echocardiographic screening reveals a major burden of undiagnosed valvular heart disease in older people: the OxVALVE Population Cohort Study. Eur Heart J. 2016;37(47):3515–3522. doi: 10.1093/eurheartj/ehw229. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Nkomo V.T., Gardin J.M., Skelton T.N., Gottdiener J.S., Scott C.G., Enriquez-Sarano M. Burden of valvular heart diseases: a population-based study. Lancet. 2006;368(9540):1005–1011. doi: 10.1016/S0140-6736(06)69208-8. [DOI] [PubMed] [Google Scholar]
- 5.Osnabrugge R.L., Mylotte D., Head S.J., et al. Aortic stenosis in the elderly: disease prevalence and number of candidates for transcatheter aortic valve replacement: a meta-analysis and modeling study. J Am Coll Cardiol. 2013;62(11):1002–1012. doi: 10.1016/j.jacc.2013.05.015. [DOI] [PubMed] [Google Scholar]
- 6.Strange G.A., Stewart S., Curzen N., et al. Uncovering the treatable burden of severe aortic stenosis in the UK. Open Heart. 2022;9(1):e001783. doi: 10.1136/openhrt-2021-001783. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Eveborn G.W., Schirmer H., Heggelund G., Lunde P., Rasmussen K. The evolving epidemiology of valvular aortic stenosis. the Tromso study. Heart. 2013;99(6):396–400. doi: 10.1136/heartjnl-2012-302265. [DOI] [PubMed] [Google Scholar]
- 8.Ambrosy A.P., Go A.S., Leong T.K., et al. Temporal trends in the prevalence and severity of aortic stenosis within a contemporary and diverse community-based cohort. Int J Cardiol. 2023;384:107–111. doi: 10.1016/j.ijcard.2023.04.047. [DOI] [PubMed] [Google Scholar]
- 9.Otto C.M., Nishimura R.A., Bonow R.O., et al. 2020 ACC/AHA guideline for the management of patients with valvular heart disease: executive summary: a report of the American College of Cardiology/American Heart Association joint committee on clinical practice guidelines. J Am Coll Cardiol. 2021;77(4):e25–e197. doi: 10.1016/j.jacc.2020.11.018. [DOI] [PubMed] [Google Scholar]
- 10.Smith C.R., Leon M.B., Mack M.J., et al. Transcatheter versus surgical aortic-valve replacement in high-risk patients. N Engl J Med. 2011;364(23):2187–2198. doi: 10.1056/NEJMoa1103510. [DOI] [PubMed] [Google Scholar]
- 11.Mack M.J., Leon M.B., Thourani V.H., et al. Transcatheter aortic-valve replacement in low-risk patients at five years. N Engl J Med. 2023;389(21):1949–1960. doi: 10.1056/NEJMoa2307447. [DOI] [PubMed] [Google Scholar]
- 12.Genereux P., Schwartz A., Oldemeyer B., et al. Design and rationale of the evaluation of transcatheter aortic valve replacement compared to surveillance for patients with asymptomatic severe aortic stenosis: the EARLY TAVR trial. Am Heart J. 2024;268:94–103. doi: 10.1016/j.ahj.2023.11.019. [DOI] [PubMed] [Google Scholar]
- 13.Snir A.D., Ng M.K., Strange G., Playford D., Stewart S., Celermajer D.S. Prevalence and outcomes of low-gradient severe aortic stenosis-from the national echo database of Australia. J Am Heart Assoc. 2021;10(22) doi: 10.1161/JAHA.121.021126. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Clavel M.A. Echocardiographic assessment of aortic stenosis severity: do not rely on a single parameter. J Am Heart Assoc. 2016;5(10) doi: 10.1161/JAHA.116.004680. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Clavel M.A., Magne J., Pibarot P. Low-gradient aortic stenosis. Eur Heart J. 2016;37(34):2645–2657. doi: 10.1093/eurheartj/ehw096. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Carabello B.A., Paulus W.J. Aortic stenosis. Lancet. 2009;373(9667):956–966. doi: 10.1016/S0140-6736(09)60211-7. [DOI] [PubMed] [Google Scholar]
- 17.Solomon M.D., Tabada G., Sung S.H., et al. Physician assessment of aortic stenosis severity, quantitative parameters, and long-term outcomes: results from the KP-VALVE project. Am Heart J. 2023;266:32–47. doi: 10.1016/j.ahj.2023.07.009. [DOI] [PubMed] [Google Scholar]
- 18.Genereux P., Sharma R.P., Cubeddu R.J., et al. The mortality burden of untreated aortic stenosis. J Am Coll Cardiol. 2023;82(22):2101–2109. doi: 10.1016/j.jacc.2023.09.796. [DOI] [PubMed] [Google Scholar]
- 19.Strange G., Stewart S., Celermajer D., et al. Poor long-term survival in patients with moderate aortic stenosis. J Am Coll Cardiol. 2019;74(15):1851–1863. doi: 10.1016/j.jacc.2019.08.004. [DOI] [PubMed] [Google Scholar]
- 20.Lancellotti P., Magne J., Dulgheru R., et al. Outcomes of patients with asymptomatic aortic stenosis followed up in heart valve clinics. JAMA Cardiol. 2018;3(11):1060–1068. doi: 10.1001/jamacardio.2018.3152. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Coisne A., Montaigne D., Aghezzaf S., et al. Association of mortality with aortic stenosis severity in outpatients: results from the VALVENOR study. JAMA Cardiol. 2021;6(12):1424–1431. doi: 10.1001/jamacardio.2021.3718. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Ross J., Jr., Braunwald E. Aortic stenosis. Circulation. 1968;38(1 Suppl):61–67. doi: 10.1161/01.cir.38.1s5.v-61. [DOI] [PubMed] [Google Scholar]
- 23.Strange G., Stewart S., Playford D., Strom J.B. Risk for mortality with increasingly severe aortic stenosis: an international cohort study. J Am Soc Echocardiogr. 2023;36(1):60–68.e62. doi: 10.1016/j.echo.2022.09.020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Stassen J., Ewe S.H., Butcher S.C., et al. Moderate aortic stenosis: importance of symptoms and left ventricular ejection fraction. Eur Heart J Cardiovasc Imaging. 2022;23(6):790–799. doi: 10.1093/ehjci/jeab242. [DOI] [PubMed] [Google Scholar]
- 25.Strom J.B., Playford D., Stewart S., et al. Increasing risk of mortality across the spectrum of aortic stenosis is independent of comorbidity & treatment: an international, parallel cohort study of 248,464 patients. PLoS One. 2022;17(7) doi: 10.1371/journal.pone.0268580. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Playford D., Schwarz N., Chowdhury E., et al. Comorbidities and symptom status in moderate and severe aortic stenosis: a multicenter clinical cohort study. JACC: Adv. 2023;2(4) doi: 10.1016/j.jacadv.2023.100356. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Ambrosy A.P., Parikh R.V., Sung S.H., et al. A natural language processing-based approach for identifying hospitalizations for worsening heart failure within an integrated health care delivery system. JAMA Netw Open. 2021;4(11) doi: 10.1001/jamanetworkopen.2021.35152. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Ambrosy A.P., Parikh R.V., Sung S.H., et al. Analysis of worsening heart failure events in an integrated health care system. J Am Coll Cardiol. 2022;80(2):111–122. doi: 10.1016/j.jacc.2022.04.045. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Genereux P., Pibarot P., Redfors B., et al. Staging classification of aortic stenosis based on the extent of cardiac damage. Eur Heart J. 2017;38(45):3351–3358. doi: 10.1093/eurheartj/ehx381. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Kanwar A., Thaden J.J., Nkomo V.T. Management of patients with aortic valve stenosis. Mayo Clin Proc. 2018;93(4):488–508. doi: 10.1016/j.mayocp.2018.01.020. [DOI] [PubMed] [Google Scholar]
- 31.Pellikka P.A., Sarano M.E., Nishimura R.A., et al. Outcome of 622 adults with asymptomatic, hemodynamically significant aortic stenosis during prolonged follow-up. Circulation. 2005;111(24):3290–3295. doi: 10.1161/CIRCULATIONAHA.104.495903. [DOI] [PubMed] [Google Scholar]
- 32.Genereux P., Stone G.W., O'Gara P.T., et al. Natural history, diagnostic approaches, and therapeutic strategies for patients with asymptomatic severe aortic stenosis. J Am Coll Cardiol. 2016;67(19):2263–2288. doi: 10.1016/j.jacc.2016.02.057. [DOI] [PubMed] [Google Scholar]
- 33.Zilberszac R., Lancellotti P., Gilon D., et al. Role of a heart valve clinic programme in the management of patients with aortic stenosis. Eur Heart J Cardiovasc Imaging. 2017;18(2):138–144. doi: 10.1093/ehjci/jew133. [DOI] [PubMed] [Google Scholar]
- 34.Monin J.L., Lancellotti P., Monchi M., et al. Risk score for predicting outcome in patients with asymptomatic aortic stenosis. Circulation. 2009;120(1):69–75. doi: 10.1161/CIRCULATIONAHA.108.808857. [DOI] [PubMed] [Google Scholar]
- 35.Otto C.M., Burwash I.G., Legget M.E., et al. Prospective study of asymptomatic valvular aortic stenosis. Clinical, echocardiographic, and exercise predictors of outcome. Circulation. 1997;95(9):2262–2270. doi: 10.1161/01.cir.95.9.2262. [DOI] [PubMed] [Google Scholar]
- 36.Iliceto A., Berndt S.L., Greenslade J.H., et al. Agreement between patient-reported and cardiology-adjudicated medical history in patients with possible ischemic chest pain: an observational study. Crit Pathw Cardiol. 2016;15(3):121–125. doi: 10.1097/HPC.0000000000000082. [DOI] [PubMed] [Google Scholar]
- 37.Dezman Z.D., Mattu A., Body R. Utility of the history and physical examination in the detection of acute coronary syndromes in emergency department patients. West J Emerg Med. 2017;18(4):752–760. doi: 10.5811/westjem.2017.3.32666. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Genereux P., Schwartz A., Oldemeyer J.B., et al. Transcatheter aortic-valve replacement for asymptomatic severe aortic stenosis. N Engl J Med. 2025;392(3):217–227. doi: 10.1056/NEJMoa2405880. [DOI] [PubMed] [Google Scholar]
- 39.Solomon M.D., Tabada G.H., Allen A., Sung S.H., Go A.S. Large-scale identification of aortic stenosis and its severity using natural language processing on electronic health records. Cardiovasc Dig Health J. 2021;2(3):156–163. doi: 10.1016/j.cvdhj.2021.03.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.