Quality of Care and Outcomes Among Medicare Advantage vs Fee-for-Service Medicare Patients Hospitalized With Heart Failure

Key Points

Question  Are there significant differences in the quality of care received by patients hospitalized with heart failure who are enrolled in Medicare Advantage vs fee-for-service Medicare?

Findings  In this cohort study, among 262 626 patients hospitalized with heart failure, no significant differences were noted in the quality of care received or in-hospital mortality between those enrolled in Medicare Advantage vs fee-for-service Medicare. Medicare Advantage patients had lower use of post–acute care facilities after discharge.

Meaning  Unlike care in the ambulatory setting, Medicare Advantage does not appear to be associated with the quality of care that Medicare patients receive when they are hospitalized with heart failure.

Importance  Medicare Advantage (MA), a private insurance plan option, now covers one-third of all Medicare beneficiaries. Although patients with cardiovascular disease enrolled in MA have been reported to receive higher quality of care in the ambulatory setting than patients enrolled in fee-for-service (FFS) Medicare, it is unclear whether MA is associated with higher quality in patients hospitalized with heart failure, or alternatively, if incentives to reduce utilization under MA plans may be associated with worse care.

Objective  To determine whether there are differences in quality of care received and in-hospital outcomes among patients enrolled in MA vs FFS Medicare.

Design, Setting, and Participants  Observational, retrospective cohort study of patients hospitalized with heart failure in hospitals participating in the Get With the Guidelines—Heart Failure registry.

Exposures  Medicare Advantage enrollment.

Main Outcomes and Measures  In-hospital mortality, discharge disposition, length of stay, and 4 heart failure achievement measures.

Results  Of 262 626 patients hospitalized with heart failure, 93 549 (35.6%) were enrolled in MA and 169 077 (64.4%) in FFS Medicare. The median (interquartile range) age was 78 (70-85) years for patients enrolled in MA and 78 (69-86) years for patients enrolled in FFS Medicare. Standard mean differences in age, sex, prevalence of comorbidities, or objective measures on admission, including vital signs and laboratory values, were less than 10%. After adjustment, there were no statistically significant differences in receipt of evidence-based β-blockers when indicated; angiotensin-converting enzyme inhibitor, angiotensin II receptor blockers, or angiotensin receptor-neprilysin inhibitors at discharge; measurement of left ventricular function; and postdischarge appointments by Medicare insurance type. Patients enrolled in MA, however, had higher odds of being discharged directly home (adjusted odds ratio [AOR], 1.16; 95% CI, 1.13-1.19; P < .001) relative to patients enrolled in FFS Medicare and lower odds of being discharged within 4 days (AOR, 0.97; 95% CI, 0.93-1.00; P = .04). There was no significant difference in in-hospital mortality between patients with MA and patients with FFS Medicare (AOR, 0.98; 95% CI, 0.92-1.03; P = .42).

Conclusions and Relevance  Among patients hospitalized with heart failure, no observable benefit was noted in quality of care or in-hospital mortality between those enrolled in MA vs FFS Medicare, except lower use of post–acute care facilities. As MA continues to grow, it will be important to ensure that participating private plans provide an added value to the patients they cover to justify the higher administrative costs compared with traditional FFS Medicare.

Medicare Advantage (MA), the private insurance option for Medicare beneficiaries, has grown substantially in recent years, with 1 in 3 Medicare beneficiaries now enrolled in the program.¹ Medicare Advantage comprises private health plans that manage health care for patients mostly through health maintenance organizations or preferred provider organizations. As enrollment in MA has grown, there have been increased efforts to understand the association of MA with the quality of care received by patients with chronic diseases relative to fee-for-service (FFS) Medicare.

Much of the prior literature examining the quality of care delivered to patients enrolled in MA has been limited to evaluations of administrative data, which often lack the granular clinical detail needed for robust risk adjustment and are subject to potential issues of upcoding.^2-5 One 2019 study using an outpatient registry of cardiology practices, however, found that patients with coronary artery disease enrolled in MA were more likely to receive evidence-based therapies in the ambulatory setting compared with those enrolled in FFS Medicare.⁶ This may be because MA plans directly incentivize ambulatory clinicians to adhere to guideline-recommend treatments for patients with cardiovascular disease, such as optimizing blood pressure and cholesterol control.^7-9 In addition, MA plans offer additional services, such as disease management programs and navigator services, that assist with care coordination across multiple clinicians.^10-12 The MA plans that achieve high quality scores on outpatient-based clinical process measures are then eligible for bonus payments from the federal government.^1,13

However, to date, it is unclear whether MA is associated with the quality of care that patients receive when they are hospitalized. It is possible that MA plans are associated with improved care of hospitalized patients. In recent years, MA plans are increasingly introducing value-based contracts that may include quality measures for patients with heart failure (HF), thereby providing an incentive to improve delivery of care.^14-16 Alternatively, MA plans might be adversely associated with care during hospitalization, given that these plans use several utilization management strategies that limit what clinicians can do in the inpatient setting. For example, they may require prior authorizations for specific procedures or medications, or they may limit utilization of post–acute care facilities, such as skilled nursing homes or rehabilitative care. It is also possible, however, that MA plans have no association with the quality of care that patients receive in the hospital. Inpatient physicians are often unaware of the specific insurance plans in which patients are enrolled and therefore may not be influenced to change the way they practice for their patients with MA.^17,18 As enrollment in the MA market continues to grow, it is important to ensure that MA is not negatively associated with the quality of care of Medicare beneficiaries when they are hospitalized. This is particularly relevant to understand given the recent evidence on other Medicare programs, which often target improvements in quality of care for patients with cardiovascular conditions, that showed no clinical benefit or, in some cases, potential harm.¹⁹

The Get With the Guidelines—Heart Failure (GWTG-HF) registry presents a unique opportunity to investigate whether MA is associated with quality of care for patients hospitalized with HF. Using this registry, we sought to answer 3 key questions. First, are there significant differences in the characteristics of patients hospitalized with HF between those enrolled in MA vs those enrolled in FFS Medicare? Second, are there significant differences in the quality of care related to clinical process measures between the 2 Medicare insurance types? Finally, do patients enrolled in MA experience better or worse in-hospital outcomes compared with patients enrolled in FFS Medicare?

The GWTG-HF registry is a national quality improvement program for hospitals and is overseen by the American Heart Association.²⁰ It enrolls patients with HF admitted to participating hospitals. Trained clinical personnel populate registry data using standardized definitions for patient demographic characteristics, clinical comorbidities, inpatient laboratory data, treatment therapies at admission and discharge, and in-hospital outcomes. This process allows for information to be more reliably collected in a similar fashion across hospitals. All participating institutions were required to comply with local regulatory and privacy guidelines and, if required, to secure institutional review board approval. Because data were used primarily at the local site for quality improvement, sites were granted a waiver of informed patient consent under the Common Rule. IQVIA (Parsippany, New Jersey) served as the registry coordination center. IQVIA is also the data collection coordination center for the American Heart Association/American Stroke Association Get With the Guidelines programs. The Duke Clinical Research Institute (Durham, North Carolina) served as the data analysis center, and institutional review board approval was granted to analyze aggregate deidentified data for research. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

We included all patients 18 years or older hospitalized with HF with Medicare insurance from January 1, 2014, through December 31, 2018. We limited analysis to patients who had recorded an insurance type of FFS Medicare or MA as their only source of payment. We excluded patients who were transferred during hospitalization, left against medical advice, or had disposition missing from the medical record. A flowchart of our selection criteria is shown in eFigure 1 in the Supplement. Of note, it is possible that a patient may appear more than once in the data set given that the registry does not have unique identifiers for patients.

We examined 3 outcome measures in the GWTG-HF registry: in-hospital mortality, discharge disposition, and length of stay of 4 or fewer days. We also examined 4 HF achievement measures, including receipt of angiotensin-converting enzyme inhibitors (ACEIs), angiotensin II receptor blockers (ARBs), or angiotensin receptor-neprilysin inhibitors (ARNIs) at discharge; use of evidence-based specific β-blockers; measurement of left ventricular function; and postdischarge appointment for patients with HF.

For secondary outcomes, we examined 18 additional quality measures: aldosterone antagonist at discharge, hydralazine and nitrates at discharge, cardiac resynchronization therapy placed or prescribed at discharge, implantable cardioverter defibrillator (ICD) counseling/ICD placed or prescribed at discharge, anticoagulation therapy for atrial fibrillation/flutter, deep vein thrombosis prophylaxis, influenza vaccination during flu season, pneumococcal vaccination, and follow-up visit within 7 or fewer days. We also examined the following reporting measures: prescription of β-blocker at discharge, ICD placed or prescribed at discharge, smoking cessation, receipt of discharge instructions, advance care plan documented, weight instructions, QRS duration documented, HF disease management program referral, and referral to HF patient education/interactive workbook.

We first examined differences in patient demographic characteristics, clinical comorbidities, vital signs on presentation, and laboratory values by insurance type. We also examined differences in counts across each outcome and quality measure between patients with MA and FFS Medicare. We then used Pearson χ² tests to compare differences in categorical variables and Wilcoxon rank sum tests to compare continuous variables. P values less than .05 were considered statistically significant. Standardized differences were calculated by dividing the differences between the treatment group and the control group by the within-group standard deviation; values greater than 0.1 were considered relevant differences.

Next, we examined unadjusted associations between Medicare insurance type and each of the primary and secondary outcomes using univariate logistic regression models. Then, we used multivariable logistic regression models, along with the generalized estimating equations procedure, to examine associations between each of the in-hospital outcomes and the quality measures with Medicare insurance type. The generalized estimating equations procedure was used to account for the clustering of patients within hospitals. The following covariates were used for adjustment: patient demographic characteristics (age, sex, and race/ethnicity), comorbidities (anemia, ischemic etiology, cerebrovascular accident/transient ischemic attack, diabetes, hyperlipidemia, hypertension, chronic obstructive pulmonary disease or asthma, peripheral vascular disease, kidney insufficiency, and smoking), hospital characteristics (including geographic region, teaching status, number of beds, and rural location), laboratory values (serum sodium and creatinine) and vitals on admission (heart rate and systolic blood pressure), and year of admission. A separate model was performed for each primary and secondary outcome. As a sensitivity analysis, we performed a mixed-effects model that included hospital random effects. We also included county-level variables obtained from the Health Resources and Services Administration Area Health Resources Files, which were based on the hospital’s zip code (including median home value, proportion of people with a high school diploma or above, proportion of people with college degrees, median household income, and unemployment rate).

Variables with missing data were not imputed for univariate tables. Model covariates with missing data were imputed using multiple imputation methods with 10 data sets, a standard process for the GWTG-HF studies.²¹ Missing medical history was imputed to no. All statistical analyses were performed at the Duke Clinical Research Institute using SAS software, version 9.4 (SAS Institute).

From January 1, 2014, through December 31, 2018, we identified 567 405 patients admitted with HF across 569 sites. Of these, 262 626 patients had Medicare insurance, of which 93 549 (35.6%) were enrolled in MA, and the remaining 169 077 (64.4%) were enrolled in FFS Medicare. The mean (SD) number of hospitalizations per site was 461.6 (575.3), and the median (interquartile range) was 266 (88-642) patients. Distribution of hospitalized patients across all sites is shown in eFigure 2 in the Supplement.

On average, there were no substantial differences in patient demographic characteristics and burden of clinical comorbidities between patients with MA and patients with FFS Medicare (Table 1). The median (interquartile range) age was 78 (70-85) years for patients enrolled in MA and 78 (69-86) years for patients enrolled in FFS Medicare. Patients enrolled in MA vs FFS Medicare were slightly less likely to be female (49.7% vs 50.2%) and White (71.2% vs 72.4%). Patients enrolled in MA were also more likely to have certain comorbidities, such as diabetes (47.4% vs 45.4%), hyperlipidemia (59.8% vs 58.3%), peripheral vascular disease (13.0% vs 12.5%), and coronary artery disease (52.0% vs 51.5%), though less likely to be receiving long-term dialysis (3.4% vs 5.0%). All standardized differences were less than 0.1 for each demographic and clinical variable examined except for geographic location: patients enrolled in MA were more often admitted to hospitals in the Midwest compared with patients enrolled in FFS Medicare (26.4% vs 21.2%; P < .001). There were also few differences in objective measures on admission, including vital signs and laboratory values, between MA and FFS Medicare beneficiaries (Table 2). Missing variables of key baseline characteristics are reported in eTable 1 in the Supplement.

There were no differences between eligible patients enrolled in MA vs FFS Medicare in the receipt of ACEIs/ARBs or ARNIs at discharge or use of β-blockers when indicated. Patients enrolled in MA were less likely to have left ventricular function measured or postdischarge HF appointments scheduled, although standardized differences were less than 0.1. There were also no significant differences in other quality measures between patients enrolled in MA vs FFS Medicare except for advance care planning, with patients enrolled in MA slightly more likely to receive this (eTable 2 in the Supplement). There were no differences between eligible patients enrolled in MA vs FFS Medicare in receiving an aldosterone antagonist at discharge, hydralazine and nitrates at discharge, deep vein thrombosis prophylaxis, and smoking cessation advice. Patients enrolled in MA were slightly less likely to receive influenza or pneumococcal vaccinations, although standardized differences were less than 0.1.

We also examined differences in discharge disposition and in-hospital outcomes. Patients enrolled in MA and FFS Medicare had a median length of stay of 4 days each (Table 3). Patients enrolled in MA were more likely to be discharged home (72.8% vs 69.8%; P < .001) and less likely to be discharged to skilled nursing facilities (16.1% vs 17.5%; P < .001) or inpatient rehabilitation facilities (1.9% vs 3.2%; P < .001) compared with patients enrolled in FFS Medicare (Table 3). There was no statistically significant difference in in-hospital mortality between patients enrolled in MA vs FFS Medicare (2.9% vs 3.0%; P = .05).

In adjusted analysis, we found no differences across 4 of the HF achievement measures, including receiving ACEIs/ARBs or ARNIs at discharge, use of β-blockers when indicated, measurement of left ventricular function, and scheduling of postdischarge appointment (Table 4). We did find that patients enrolled in MA had higher odds of being discharged directly home (adjusted odds ratio [AOR], 1.16; 95% CI, 1.13-1.19; P < .001) relative to patients enrolled in FFS Medicare (Table 4). They also had lower odds of being discharged within 4 or fewer days (AOR, 0.97; 95% CI, 0.93-1.00; P = .04). There were no differences in in-hospital mortality between patients enrolled in MA vs FFS Medicare.

We also examined the association of Medicare insurance type with other quality measures (eTable 3 in the Supplement). Of 18 process-based quality measures, only 2 were found to have a statistically significant association with Medicare insurance type. Patients enrolled in MA were more likely to receive anticoagulation therapy for atrial fibrillation or atrial flutter (AOR, 1.13; 95% CI, 1.06-1.21; P < .001) and more likely to receive discharge instructions (AOR, 1.11; 95% CI, 1.03-1.20; P = .008) relative to patients enrolled in FFS Medicare.

In our sensitivity analysis, we performed a mixed-effects model that included patient-level characteristics, county-level social determinant variables, and hospital random effects. County-level variables were similar across hospitals visited by patients enrolled in MA and FFS Medicare (eTable 4 in the Supplement). Overall, our results were consistent. Patients enrolled in MA were still more likely to be discharged home relative to patients enrolled in FFS Medicare (eTable 5 in the Supplement). We also did not find any differences in in-hospital mortality or across the majority of HF quality measures between the 2 groups of patients.

Using a national registry of patients hospitalized with HF, we found few meaningful differences in the characteristics of patients or the quality of care received between those enrolled in MA vs those enrolled in FFS Medicare. Patients with HF enrolled in MA and FFS Medicare were observed to have similar demographic characteristics, clinical comorbidities, and objective measures on admission, including laboratory values and vital signs. Additionally, there were no statistically significant differences by Medicare insurance type across most quality measures or in-hospital mortality. Patients enrolled in MA were, however, less likely to be discharged to a post–acute care facility and more likely to be discharged directly home compared with patients enrolled in FFS Medicare. Taken together, unlike in the ambulatory setting, we found little evidence that MA is associated with the quality of care for patients hospitalized with HF beyond possibly dictating discharge disposition.

Our work highlights important findings. First, unlike earlier reports that show MA plans attract healthier patients,^8,22 our work uses granular clinical data obtained from medical records to show that there are no substantial differences in the demographic characteristics or clinical comorbidities of hospitalized patients with HF by Medicare insurance type given that standard mean differences were less than 10%. Even among objective measures, such as vital signs and laboratory values, which are not susceptible to upcoding practices, we found no substantial difference between patients enrolled in MA and patients enrolled in FFS Medicare. These results suggest that hospitalized patients with HF enrolled in MA are just as sick as those enrolled in FFS Medicare. The only major difference is geographic variation, with more patients enrolled in MA residing in the Midwest, which is a reflection of MA penetration being relatively higher there than in other regions.¹ However, as the cohort studied was confined to those hospitalized with HF, there may still be important differences in severity among patients with HF overall for those enrolled in MA vs FFS Medicare.

Next, we did not find statistically significant differences in the quality of care that patients received by Medicare insurance type, with the exception of a few measures. Importantly, there was no difference in in-hospital mortality. There was also no difference in the use of most guideline-recommended therapies, such as ACEI/ARB/ARNI use or specific β-blockers among eligible patients. These findings are in contrast to prior literature that found that MA plans outperform FFS Medicare across a broad set of process-based clinical measures in the ambulatory care setting. A 2019 study,⁶ for example, found that patients enrolled in MA with coronary artery disease were much more likely to receive guideline-recommended treatment when eligible compared with patients enrolled in FFS Medicare, including β-blockers, ACEIs/ARBs, and statins.

What could explain the difference in findings in the inpatient vs ambulatory setting? One important reason may be the nature of how MA negotiates contracts with ambulatory clinicians compared with inpatient clinicians. Unless part of a larger health system, ambulatory clinicians and their practices negotiate directly with MA plans or use an intermediary, such as an independent practice association. As part of this process, MA plans often set quality benchmarks for ambulatory clinicians, which may also influence their reimbursement rate.²³ As such, ambulatory clinicians have a strong incentive to manage the care of their patients enrolled in MA and ensure a certain level of quality, which may help explain why quality of care for patients enrolled in MA in the ambulatory setting tends to be better than that for patients enrolled in FFS Medicare. In the hospital setting, however, although similar quality contracts exist, inpatient physicians are often unaware of the specifics of the contract, and it is thus much harder to influence their practice. Therefore, it is likely that the complex organizational and financial structures of hospitals may limit the influence that MA has in the inpatient setting. Also, unlike their ability to select physicians to be within their network in the ambulatory setting, MA plans likely do not have the ability to select the physicians who work in the hospital.

However, there were some important differences observed by Medicare insurance type. One main difference was the use of post–acute care facilities. Patients enrolled in MA vs FFS Medicare were less likely to be discharged to a skilled nursing facility or inpatient rehabilitative center, which is more costly than being sent directly home. This finding is consistent with the well-known practice of utilization management by MA plans to keep costs down.⁴ Medicare Advantage plans negotiate contracts with post–acute care clinicians and pay them from their monthly capitated allotment they receive from Medicare. As such, MA plans have a strong incentive to control costs of their patients, since any difference between the capitation payment and their expenditures are kept as a profit or incurred as a loss. To achieve lower utilization of post–acute care facilities, MA plans often require a prior authorization or limit coverage of post–acute care services to specific clinicians.²⁴ However, we cannot determine if these differences in discharge destination were appropriate or not.

Another difference was that patients enrolled in MA were more likely to receive anticoagulation therapy for atrial fibrillation or atrial flutter. This finding may reflect potential differences in drug coverage by insurance type. Almost all patients enrolled in MA have drug coverage through a Part D plan, and they also have out-of-pocket caps in spending.²⁵ In contrast, patients enrolled in FFS Medicare do not have out-of-pocket caps, and only about 60% are enrolled in a Part D drug plan.²⁶ Therefore, unless they have supplemental drug insurance from a former employer or a union, it is possible that patients enrolled in FFS Medicare may be less able to afford some of the newer high-cost oral anticoagulant medications.

This study has several limitations. First, the GWTG-HF is a voluntary registry, so hospitals that participate in the program may differ from those that do not. There may also be important differences in the demographic characteristics of the patients served in these hospitals compared with patients at nonparticipating hospitals. Therefore, the generalizability of our results may be limited to similar types of hospitals. However, it is important to note that a prior study found that patients in hospitals that participate in the HF registry are nationally representative.²⁷ Second, the GWTG-HF registry does not contain patient-level social determinants of health that may affect health; however, our sensitivity analysis that included county-level characteristics did not alter our findings. Third, our observational study can only report associations and cannot prove a causal relationship between Medicare insurance type and quality. Third, our sample does not include all patients enrolled in MA and FFS Medicare; therefore, we are also unable to definitively exclude small but potentially important differences in quality performance and in-hospital outcomes. Fourth, we were unable to evaluate postdischarge quality metrics, including 30-day mortality and readmission rates, because linked data on patients enrolled in MA are not available for research purposes at this time. Understanding how patient outcomes differ will be important to determine the appropriateness of fewer patients being discharged home under MA relative to FFS Medicare.

Using a national registry of hospitalized patients with HF, we found no meaningful differences in the characteristics of patients or the quality of care received between those enrolled in MA vs those enrolled in FFS Medicare. The main influence of MA appears to be limiting the use of post–acute care facilities, which is likely a utilization management strategy aimed at reducing costs. As MA continues to grow, it will be important to continue to ensure that utilization management strategies do not adversely affect care and also consider changes in the program that may eventually improve quality of care among hospitalized patients.

Accepted for Publication: May 12, 2020.

Corresponding Author: Jose F. Figueroa, MD, MPH, Harvard Medical School, Brigham and Women’s Hospital, 42 Church St, Cambridge, MA 02138 (jffigueroa@partners.org).

Published Online: September 2, 2020. doi:10.1001/jamacardio.2020.3638

Author Contributions: Ms Xu had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Figueroa, Wadhera, Fonarow, Xu, DeVore, Yancy, Joynt Maddox.

Acquisition, analysis, or interpretation of data: Figueroa, Wadhera, Frakt, Fonarow, Heidenreich, Xu, Lytle, DeVore, Matsouaka, Bhatt, Joynt Maddox.

Drafting of the manuscript: Figueroa, Lytle, Joynt Maddox.

Critical revision of the manuscript for important intellectual content: Figueroa, Wadhera, Frakt, Fonarow, Heidenreich, Xu, DeVore, Matsouaka, Yancy, Bhatt, Joynt Maddox.

Statistical analysis: Xu, DeVore, Matsouaka.

Obtained funding: Figueroa, Fonarow.

Administrative, technical, or material support: Figueroa, Lytle.

Study supervision: Figueroa, Frakt, Fonarow, Yancy.

Conflict of Interest Disclosures: Dr Figueroa reported receiving grants from the National Center for Advancing Translational Sciences (KL2 TR002542) and the Commonwealth Fund during the conduct of the study. Dr Wadhera reported receiving grants from the National Heart, Lung, and Blood Institute (K23HL148525-1) and personal fees from Regeneron outside the submitted work. Dr Frakt reported receiving grant support from VA Health Services Research and Development Service, VA Quality Enhancement Research Initiative, The Laura and John Arnold Foundation, and The Robert Wood Johnson Foundation outside the submitted work; the contents do not represent the views of the US Department of Veterans Affairs or the US Government. Dr Fonarow reported receiving personal fees from Abbott, Amgen, Bayer, CHF Solutions, Janssen, Merck, Medtronic, and Novartis outside the submitted work. Dr Lytle reported having a contract from American Heart Association during the conduct of the study. Dr DeVore reported receiving grants from the American Heart Association during the conduct of the study; grants from Amgen, Bayer, Intra-Cellular Therapies, Luitpold Pharmaceuticals, Medtronic, the National Heart, Lung, and Blood Institute, Novartis, and the Patient-Centered Outcomes Research Institute; other from Amgen, AstraZeneca, Bayer, InnaMed, LivaNova, Mardil Medical, Novartis, Procyrion, scPharmaceuticals, and Zoll; and personal fees and other from Abbott outside the submitted work. Dr Bhatt reported receiving grants from Abbott, Afimmune, Amarin, Amgen, AstraZeneca, Bristol-Myers Squibb, Chiesi, Eisai, Ethicon, Forest Laboratories/AstraZeneca, Fractyl, Idorsia, Ischemix, Ironwood, Lexicon, Lilly, Medtronic, Pfizer, Regeneron, Roche, Sanofi Aventis, Synaptic, and The Medicines Company; grants and other from Cardax, PhaseBio, and PLx Pharma; personal fees from Bayer, Belvoir Publications, Cleveland Clinic, CSL Behring, Duke Clinical Research Institute, Elsevier, Ferring Pharmaceuticals, Harvard Clinical Research Institute (now Baim Institute for Clinical Research), HMP Global, Journal of the American College of Cardiology, Mayo Clinic, Medtelligence/ReachMD, Mount Sinai School of Medicine, Population Health Research Institute, Slack Publications, TobeSoft, VA, and WebMD; personal fees and other from Boehringer Ingelheim; personal fees, nonfinancial support, and other from American College of Cardiology; personal fees and nonfinancial support from Society of Cardiovascular Patient Care; nonfinancial support from American Heart Association; and other from Biotronik, Boston Scientific, Boston VA Research Institute, Cereno Scientific, Clinical Cardiology, CSI, FlowCo, Medscape Cardiology, Merck, Novo Nordisk, Regado Biosciences, St Jude Medical (now Abbott), Svelte, and Takeda outside the submitted work. Dr Joynt Maddox reported receiving grants from the National Heart, Lung, and Blood Institute (R01HL143421), the National Institute on Aging (R01AG060935), and the Commonwealth Fund and performing contract work for the US Department of Health and Human Services outside the submitted work. No other disclosures were reported.

Disclaimer: Dr Fonarow is Section Editor and Dr Yancy is Deputy Editor of JAMA Cardiology, but they were not involved in any of the decisions regarding review of the manuscript or its acceptance.