Cost-effectiveness of Sodium-Glucose Cotransporter-2 Inhibitors for the Treatment of Heart Failure With Preserved Ejection Fraction

Key Points

Question  Is adding a sodium-glucose cotransporter-2 inhibitor (SGLT2-I) to standard therapy cost-effective in heart failure with preserved ejection fraction?

Findings  In this Markov model that simulated monthly outcomes based on results from the Empagliflozin in Heart Failure With a Preserved Ejection Fraction (EMPEROR-Preserved) and Dapagliflozin in Heart Failure With Mildly Reduced or Preserved Ejection Fraction (DELIVER) trials, adding an SGLT2-I to standard therapy is projected to cost $141 200 per quality-adjusted life-year gained. An SGLT2-I would not be cost-effective if there is no mortality benefit.

Meaning  Results suggest that adding SGLT2-I therapy is intermediate or low value in patients with heart failure with preserved ejection fraction.

Importance  Adding a sodium-glucose cotransporter-2 inhibitor (SGLT2-I) to standard-of-care treatment in patients with heart failure with preserved ejection fraction (HFpEF) reduces the risk of a composite outcome of worsening heart failure or cardiovascular mortality, but the cost-effectiveness in US patients with HFpEF is uncertain.

Objective  To evaluate the lifetime cost-effectiveness of standard therapy plus an SGLT2-I compared with standard therapy in individuals with HFpEF.

Design, Setting, and Participants  In this economic evaluation conducted from September 8, 2021, to December 12, 2022, a state-transition Markov model simulated monthly health outcomes and direct medical costs. Input parameters including hospitalization rates, mortality rates, costs, and utilities were extracted from HFpEF trials, published literature, and publicly available data sets. The base-case annual cost of SGLT2-I was $4506. A simulated cohort with similar characteristics as participants of the Empagliflozin in Heart Failure With a Preserved Ejection Fraction (EMPEROR-Preserved) and Dapagliflozin in Heart Failure With Mildly Reduced or Preserved Ejection Fraction (DELIVER) trials was used.

Exposures  Standard of care plus SGLT2-I vs standard of care.

Main Outcomes and Measures  The model simulated hospitalizations, urgent care visits, and cardiovascular and noncardiovascular death. Future medical costs and benefits were discounted by 3% per year. Main outcomes were quality-adjusted life-years (QALYs), direct medical costs (2022 US dollars), and incremental cost-effectiveness ratio (ICER) of SGLT2-I therapy from a US health care sector perspective. The ICER of SGLT2-I therapy was evaluated according to the American College of Cardiology/American Heart Association value framework (high value: <$50 000; intermediate value: $50 000 to <$150 000; and low value: ≥$150 000).

Results  The simulated cohort had a mean (SD) age of 71.7 (9.5) years and 6828 of 12 251 participants (55.7%) were male. Standard of care plus SGLT2-I increased quality-adjusted survival by 0.19 QALYs at an increased cost of $26 300 compared with standard of care. The resulting ICER was $141 200 per QALY gained, with 59.1% of 1000 probabilistic iterations indicating intermediate value and 40.9% indicating low value. The ICER was most sensitive to SGLT2-I costs and effect of SGLT2-I therapy on cardiovascular death (eg, increasing to $373 400 per QALY gained if SGLT2-I therapy was assumed to have no effect on mortality).

Conclusions and Relevance  Results of this economic evaluation suggest that at 2022 drug prices, adding an SGLT2-I to standard of care was of intermediate or low economic value compared with standard of care in US adults with HFpEF. Efforts to expand access to SGLT2-I for individuals with HFpEF should be coupled with efforts to lower the cost of SGLT2-I therapy.

Heart failure (HF) with preserved ejection fraction (HFpEF) affects 3 million US adults, yet, until recently, there were no evidence-based therapies proven to improve health outcomes.^1,2 In the Empagliflozin in Heart Failure With a Preserved Ejection Fraction (EMPEROR-Preserved) and Dapagliflozin in Heart Failure With Mildly Reduced or Preserved Ejection Fraction (DELIVER) clinical trials, sodium-glucose cotransporter-2 inhibitors (SGLT2-Is) reduced the risk of a composite end point of HF hospitalization or cardiovascular death compared with placebo.^3-5 The benefit of SGLT2-I in these trials was primarily due to a reduction in HF hospitalizations.^3-5A subsequent meta-analysis showed consistent results between the trials, suggesting a drug class effect.⁵ SGLT2-Is received a Class 2a recommendation in the 2022 American Heart Association/American College of Cardiology/Heart Failure Society of America HF guidelines, which were published prior to the release of results from the DELIVER trial.⁶

SGLT2-Is are currently proprietary medications with a relatively high annual price. A 1-year supply has an average wholesale price of approximately $8000, and the Federal Supply Schedule cost, which reflects the lower price at which the drug is available to federal government purchasers, is approximately $4500. With 3 million US adults eligible for therapy, the adoption of SGLT2-Is could substantially increase pharmaceutical spending. High total and out-of-pocket costs are key barriers to widespread uptake of effective cardiovascular therapies in the US, and cost-effectiveness analyses have previously helped spur price reductions that improved affordability and access.⁷ Quantifying the lifetime cost-effectiveness and annual budget impact of SGLT2-Is in HFpEF may guide uptake and influence US pricing of this novel therapy. Given the increasing salience of rigorous economic evaluations in clinical guidelines and their ability to inform uptake and pricing, we evaluated the lifetime cost-effectiveness of adding an SGLT2-I to standard of care in individuals with HFpEF from the US health care sector perspective.

We developed a state-transition Markov model to project the lifetime health effects and direct medical costs of adding an SGLT2-I (empagliflozin, 10 mg, once daily or dapagliflozin, 10 mg, once daily) to standard of care in patients with HFpEF who meet the eligibility criteria of the EMPEROR-Preserved or DELIVER trials.^3-5 In each monthly cycle, patients could have an HF urgent care visit, an HF hospitalization, or die of cardiovascular or noncardiovascular causes (Figure 1; eFigure 1 and eTable 1 in Supplement 1).

The simulation model was developed using TreeAge Pro 2021 (TreeAge Software Inc). Additional analyses were performed in Microsoft Excel, version 16 (Microsoft Corporation) and R, version 4.0.2 (R Core Team). Our analysis followed the best-practice recommendations from the Second Panel on Cost-Effectiveness in Health and Medicine (eTables 2 and 3 in Supplement 1).⁸ These data are deidentified, and the analysis is not considered human subject research; therefore, institutional review board approval was not required. This investigation was conducted independently from the writing groups and sponsors of the EMPEROR-Preserved or DELIVER trials.

We simulated a cohort of patients matching the characteristics of those of the EMPEROR-Preserved and DELIVER trials (eMethods 1 in Supplement 1). Meta-analyses did not show evidence of heterogeneity in treatment effects by race and ethnicity; therefore, our analyses did not consider race and ethnicity beyond the characteristics of trial participants. The pooled mean (SD) age of EMPEROR-Preserved and DELIVER participants was 71.7 (9.5) years; 5423 were female (44.3%); and 6828 were male (55.7%). Participants in the trials were from the following race and ethnicity categories: 2098 Asian (17.1%), 417 Black (3.4%), 8981 White (73.3%), and 755 other (6.2%). The pooled mean (SD) left ventricular ejection fraction of trial participants was 54.2% (8.8%) and 66.5% of participants (8152 of 12 251) had a left ventricular ejection fraction of 50% or greater. Model input parameters were derived from published data, including primary and secondary analyses of the EMPEROR-Preserved and DELIVER trials, a meta-analysis of the trials, published HF and HFpEF epidemiologic and trial data, and publicly available data sets (Table 1).^3-5,9-29

For the first 2.3 years in the model (mean of the median follow-up in the EMPEROR-Preserved and DELIVER trials), rates of cardiovascular and noncardiovascular mortality in the standard-of-care arm were derived from the pooled control arms of the 2 trials. Extrapolation beyond the trial follow-up used a nonparametric approach as in prior work (ie, comparing cardiovascular and noncardiovascular mortality rates in the final 6 months of trial follow-up to that observed in the age-matched US general population) (eMethods 2 and eFigure 2 in Supplement 1).^{3,4,17,18,21,23} The rates of first HF hospitalization and HF urgent care visits in the standard-of-care arm were based on the control arms of the 2 trials. The risk of death during a hospitalization and risk of readmission were derived from the published literature and national sources (eMethods 3 in Supplement 1).^10-14

In a meta-analysis of the EMPEROR-Preserved and DELIVER trials, SGLT2-I use was associated with a significant reduction in HF hospitalizations (hazard ratio [HR], 0.74; 95% CI, 0.67-0.83) and a nonsignificant reduction in cardiovascular death (HR, 0.88; 95% CI, 0.77-1.00).⁵ We applied these HRs, which were probabilistically sampled from distributions in the base-case analysis to reflect the uncertainty of the estimates, to the monthly rates in the standard-of-care arm to calculate the corresponding monthly rates in the SGLT2-I arm. The rate of noncardiovascular mortality during the first 2.3 years in the model was derived from the pooled intervention arms of the 2 trials (eMethods 5 in Supplement 1). In the base case, extrapolation beyond the trial period assumed that SGLT2-Is had no effect on noncardiovascular mortality (eMethods 5 in Supplement 1). These assumptions were varied in sensitivity and scenario analyses.

The base-case annual cost of SGLT2-Is was derived from the Federal Supply Schedule as recommended by the Second Panel on Cost-effectiveness in Health and Medicine and assumed to be $4506 (mean price for empagliflozin and dapagliflozin for the Department of Veterans Affairs, Department of Defense, Public Health Service, and Coast Guard).^19,20 In sensitivity analyses, we varied annual SGLT2-I costs from $1415 (mean of most discounted drug price for empagliflozin and dapagliflozin available to some US payers) to $6620 (mean of wholesale acquisition cost for empagliflozin and dapagliflozin).³⁰ As our model incorporated intention-to-treat effect estimates for SGLT2-Is, we accounted for within-trial nonpersistence by proportionally reducing medication costs by the pooled discontinuation rates reported in the trials (18.6%).^3,4 Facility costs for HF hospitalizations and 30-day readmissions were derived from the National Inpatient Sample.¹⁵ Total hospitalization costs were calculated by multiplying facility costs by a professional fee ratio (1.177).²⁴ Background medical costs, defined as age-adjusted direct medical costs, excluding HF hospitalizations, HF urgent care visits, and SGLT2-I costs, were derived from national sources as previously described.^21-23

The EMPEROR-Preserved and DELIVER trials measured health-related quality of life using the Kansas City Cardiomyopathy Questionnaire (KCCQ) (eMethods 4 in Supplement 1). As in prior analyses, we mapped the KCCQ to quality-of-life weights using a published regression formula.^3,4,21,25-27 During the first year of the analysis, we assumed the KCCQ captured all components of quality of life, including quality-of-life reductions for HF hospitalizations. During subsequent years, quality-of-life reductions due to HF hospitalizations, HF urgent care visits, and aging were modeled separately as these were not captured in the trial.^21,28,29 In the base-case analysis, we assumed that quality-of-life improvements with SGLT2-Is persisted over a lifetime.¹⁶

We examined the model performance by comparing outcomes observed in the trials, including event rates and HRs, with model outputs at 2.3 years (mean of the median follow-up in the 2 trials). Main outcomes were direct medical costs (2022 US dollar), quality-adjusted life-years (QALYs), and the lifetime incremental cost-effectiveness ratio (ICER) of an SGLT2-I plus standard-of-care therapy compared with standard-of-care therapy alone. We adopted a US health care sector perspective and a lifetime analytic horizon. All future costs and QALYs were discounted 3% annually.⁸ We applied the American College of Cardiology/American Heart Association value framework, in which interventions with ICERs less than $50 000, $50 000 to less than $150 000, and $150 000 or more per QALY gained are classified as high, intermediate, and low value, respectively.³¹ The mean and 95% uncertainty interval (UI) of key outcomes and the proportion of ICERs in each value category were derived by running 1000 iterations of the model, randomly sampling model inputs from prespecified statistical distributions. A budget impact analysis estimated the annual effect on US health care spending if an SGLT2-I was added to current treatment for eligible US adults with HFpEF (eMethods 6 in Supplement 1).

Sensitivity analyses were used to examine the effect of uncertainty in key input parameters on our findings (eTable 4 in Supplement 1). One-way deterministic analyses varied 1 parameter at a time while holding all other parameters at their base-case value (Table 1).⁹ Two-way analyses simultaneously varied 2 key inputs while holding all others at the base-case value, such as the association of SGLT2-I therapy with cardiovascular mortality rates vs SGLT2-I drug cost and time to generic entry vs the expected percent reduction in price at that time. In exploratory analyses, we examined heterogeneity in cost-effectiveness as a function of baseline age and left ventricular ejection fraction (eMethods 5 in Supplement 1).

Additional sensitivity analyses included the following: (1) varied the analytic time horizon; (2) assumed that the association with noncardiovascular mortality observed in the trial period persisted in the posttrial period; (3) assumed the association of SGLT2-Is with HF hospitalizations, HF urgent care visits, and cardiovascular death waned over time; (4) used quality-of-life inputs derived from national sources (rather than patient-reported values from the trial) with and without improvement in quality of life with SGLT2-I use; (5) assumed quality-of-life benefits waned over time; and (6) modeled trial-specific populations, drug costs, treatment outcomes, and survival projections (rather than pooled results from both trials) (eMethods 4 and eTables 4 and 5 in Supplement 1).

Because the long-term cost-effectiveness of a therapy is typically sensitive to assumptions of the effect on mortality, we examined an alternative scenario that assumed that SGLT2-I therapy was not associated with cardiovascular or noncardiovascular mortality over the entire lifetime (ie, HRs were set to 1).³² As appropriate, we repeated sensitivity analyses previously described with this assumption.

At 2.3 years of follow-up, the model reproduced the pooled event rates and HRs for HF hospitalizations, HF urgent care visits, cardiovascular death, and all-cause mortality effects observed in the trials (Figure 1; eTable 6 in Supplement 1).

Adding an SGLT2-I to standard-of-care therapy was projected to reduce the mean number of lifetime HF hospitalizations from 0.74 (95% UI, 0.65-0.87) to 0.53 (95% UI, 0.44-0.63), increase the mean undiscounted survival from 7.73 (95% UI, 7.65-7.80) years to 7.95 (95% UI, 7.68-8.20) years, and increase lifetime QALYs from 5.27 (95% UI, 5.22-5.33) to 5.46 (95% UI, 5.29-5.62) (Table 2; eFigure 3 in Supplement 1). This gain of 0.22 (95% UI, −0.03 to 0.46) life-years or 0.19 (95% UI, 0.03-0.33) QALYs was accompanied by an increase in lifetime direct health care spending of $26 312 (95% UI, $20 290-$31 722). Thus, the lifetime ICER of adding SGLT2-I therapy to standard of care in HFpEF was $141 200 per QALY gained compared with standard-of-care therapy. In 1000 probabilistic iterations, the use of an SGLT2-I was an intermediate-value intervention (ie, ICER was $50 000 to <$150 000 per QALY gained) in 591 iterations (59.1%) and low value in 409 iterations (40.9%) (Figure 2).

In 1-way analyses, SGLT2-I therapy became a low-value intervention when the HR for cardiovascular death exceeded 0.89 (Figure 3). The use of SGLT2-I was high value if the annual cost of SGLT2-I was less than $1431 and low value if the annual cost exceeded $4795. In 2-way analyses, if SGLT2-I costs were $1415 (most discounted drug price available to some US payers), then it would be an intermediate-value intervention as long as the HR for cardiovascular mortality was less than 1.00 and high value when less than 0.90 (eFigure 4 in Supplement 1). If the availability of generic formulations in 5 years were to lower the drug cost by 50%, the ICER would decrease to $112 900 per QALY gained (eTable 7 in Supplement 1).

Exploratory analyses projected that standard of care plus SGLT2-I became low value when baseline age was older than 78 years and when baseline left ventricular ejection fraction was 65% or greater (eFigure 5 in Supplement 1). In addition, the ICER increased when the analytic horizon was shortened (eg, $357 400 per QALY gained at 2.3 years) (eFigure 6 in Supplement 1). Other sensitivity analyses did not meaningfully alter the results, with standard of care plus SGLT2-I being either intermediate or low value (eTable 8 in Supplement 1). Because the annual cost of empagliflozin is lower than that of dapagliflozin and because the DELIVER trial enrolled a higher-risk population, the ICER was numerically lower in the EMPEROR-Preserved trial–specific analysis ($126 300 per QALY gained) compared with the DELIVER trial–specific analysis ($144 000 per QALY gained) (eTable 8 and eFigure 7 in Supplement 1).

Assuming SGLT2-I therapy had no effect on cardiovascular or noncardiovascular mortality, the ICER increased to $373 400 per QALY gained; in that context, the therapy was low value in 100.0% of simulations (eTable 9 and eFigure 8 in Supplement 1). Adding SGLT2-I therapy to standard of care was low value in all other sensitivity analyses repeated with this assumption (eTables 10 and 11 and eFigure 9 in Supplement 1).

If one-half of all eligible US adults with HFpEF were to receive an SGLT2-I, annual US health care spending could increase by $4.3 billion (range, $3.4-5.5 billion). This includes an annual increase in pharmaceutical spending of $4.8 billion and $0.5 billion in savings from averted cardiovascular hospitalizations and other costs (eFigure 10 in Supplement 1).

In this economic evaluation that used a simulation model of US adults with HFpEF, we projected that the addition of an SGLT2-I to standard of care would increase quality-adjusted survival by 0.19 years (equivalent to approximately 69 additional days of perfect health) and increase lifetime costs by $26 300. The resulting ICER was $141 200 per QALY gained, an intermediate-value strategy using the American College of Cardiology/American Heart Association framework.³¹ However, the economic value is sensitive to the association with cardiovascular mortality, and a small increase in cardiovascular mortality (mean HR from 0.88 in the base case to 0.89 or higher) would make SGLT2-Is a low-value strategy. Lowering the annual cost of SGLT2-Is to $1431—a 68% reduction from the current Veterans Affairs price and a 78% reduction in wholesale acquisition cost, which may be reasonable after generic formulations are available—would result in SGLT2-Is being high to intermediate value.

Our findings build on those of a recent cost-effectiveness analysis that projected that empagliflozin would be a low-value strategy when added to standard-of-care therapy in US adults with HFpEF.³² That study examined 1 SGLT2-I (empagliflozin) based on the results of 1 trial (EMPEROR-Preserved), and its base-case analysis assumed that SGLT2-I therapy would not be associated with all-cause mortality. Our model has the advantage of more precise inputs drawn from the larger sample size of pooled analyses of the 2 pivotal trials of SGLT2-Is in HFpEF. Even so, we found similar results in sensitivity analyses that assumed no mortality benefit (ICER $373 400 in our model compared with $437 400 in a prior analysis).³² Future studies should examine the effect of SGLT2-Is on death from cardiovascular and noncardiovascular causes.

Our analysis also builds on prior cost-effectiveness analyses evaluating the use of SGLT2-Is in patients with HF with reduced EF (rEF), which projected that the lifetime ICER of dapagliflozin was $68 300 to $83 650 per QALY gained.^33,34 The lower ICER in HFrEF compared with HFpEF is likely due to 2 factors. First, SGLT2-Is produced a significant associated reduction in cardiovascular mortality in HFrEF, but the reduction in cardiovascular mortality did not achieve statistical significance in the meta-analysis of HFpEF trials.^5,33 Second, cardiovascular deaths were a larger proportion of deaths in patients with HFrEF than in patients with HFpEF.^3,4,33,35

The cost-effectiveness findings should be interpreted in the context that SGLT2-Is are the first pharmaceutical class to meet primary end points in randomized clinical trials of patients with HFpEF, and the Class 2a recommendation in the recent HF guidelines is likely to be upgraded in the future given the recent positive trial results. Historically, however, evidence of effectiveness has been insufficient to motivate adoption of a novel therapy, which is influenced by 3 related but distinct economic considerations. The cost-effectiveness of the new therapy compared with the best available alternative or standard of care, a marker of economic efficiency, answers the following question: “Does the new therapy represent good value to the health care sector/society?” Our results suggest that SGLT2-Is are of intermediate or low value in HFpEF at 2022 prices. Second, the budget impact of the new therapy, which is the net change in health care spending when the drug is adopted by eligible individuals, answers the following question: “Is the new therapy affordable to the health care sector/society?” We estimated that treating one-half of all eligible US adults with HFpEF with SGLT2-Is could increase total US health care spending by $4.3 billion annually. However, the budget impact may be mitigated by the slow uptake of cardiovascular therapies in recent years and would be reduced with availability of generic formulations in the future.³⁶ Finally, examining out-of-pocket costs, which affect the patient’s ability and willingness to initiate or adhere to a new therapy, helps answer the question, “Can patients afford the new therapy?” According to 2019 Medicare data, the median annual patient cost sharing for SGLT2-Is may be as high as $1800, which may be a barrier to access.³⁷ Cost sharing does not affect all patients equally—fixed-income seniors may not be able to afford the high (and frequently variable) out-of-pocket costs of a novel therapy.^38,39 The Inflation Reduction Act of 2022 will control total drug costs by negotiating the price of top-spending drugs, imposing penalty rebates on increases in prices above inflation and capping out-of-pocket costs at $2000 per year.⁴⁰ These provisions are likely to eventually improve the cost-effectiveness and affordability of SGLT2-Is for Medicare beneficiaries, but affordability may remain a challenge for enrollees of high cost-sharing commercial plans and uninsured patients.

Our study has a few limitations. Our model inputs were derived from a synthesis of currently available data, and our results should be updated if longer-term safety or effectiveness data become available. Some patient groups may derive larger or smaller benefits from SGLT2-Is than estimated by our model, with commensurate heterogeneity in cost-effectiveness. For instance, if patients with chronic kidney disease and HFpEF derive additional benefit from the drug’s nephroprotective effects, we may have underestimated the cost-effectiveness of SGLT2-I therapy in this population. In contrast, we may have overestimated the cost-effectiveness of SGLT2-Is in older adult patients with multiple morbidities and a higher-than-average competing risk of noncardiovascular death. We used a published approach to map KCCQ scores to quality-of-life measures, but these instruments measure different aspects of quality of life. Our approach may not capture a patient’s preference for that health state, and our results were sensitive to quality-of-life benefits. We did not include treatment-related adverse events as there were no significant safety-related differences between study groups in the trials.^3,4 However, rare adverse effects (eg, euglycemic ketoacidosis or complicated genitourinary infections) may only become apparent after widespread adoption of the therapy and longer follow-up.

In this economic evaluation, which included a simulation model of US adults with HFpEF, results suggest that adding an SGLT2-I at an annual cost of $4506 to standard-of-care therapy was of intermediate or low economic value compared with standard of care alone. Studies examining the long-term effect of SGLT2-I on mortality should be coupled with efforts to lower costs of SGLT2-Is.

Accepted for Publication: December 19, 2022.

Published Online: March 4, 2023. doi:10.1001/jamacardio.2023.0077

Corresponding Authors: Dhruv S. Kazi, MD, MSc, MS, Richard A. and Susan F. Smith Center for Outcomes Research in Cardiology, Beth Israel Deaconess Medical Center, 375 Longwood Ave, 4th Floor, Boston, MA 02215 (dkazi@bidmc.harvard.edu); Brandon K. Bellows, PharmD, MS, Department of Medicine, Columbia University Irving Medical Center, 622 W 168th St, PH9-105, New York, NY 10032 (bkb2132@cumc.columbia.edu).

Author Contributions: Drs Kazi and Bellows had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Drs Kazi and Bellows contributed equally to this work as co–senior authors.

Concept and design: Cohen, Isaza, Ho, Kazi, Bellows.

Acquisition, analysis, or interpretation of data: Cohen, Isaza, Hernandez, Lewis, Fonarow, Kazi, Bellows.

Drafting of the manuscript: Cohen, Isaza, Bellows.

Critical revision of the manuscript for important intellectual content: Cohen, Hernandez, Lewis, Ho, Fonarow, Kazi, Bellows.

Statistical analysis: Cohen, Isaza, Kazi, Bellows.

Obtained funding: Kazi.

Supervision: Lewis, Kazi, Bellows.

Conflict of Interest Disclosures: Dr Hernandez reported receiving personal fees from Pfizer and BMS outside the submitted work. Dr Ho reported receiving grants from the National Institutes of Health outside the submitted work. Dr Fonarow reported receiving consulting fees from Abbott, Amgen, AstraZeneca, Bayer, Cytokinetics, Edwards, Eli Lilly, Janssen, Medtronic, Merck, Novartis, and Pfizer outside the submitted work. No other disclosures were reported.

Funding/Support: This work was supported by grant K01-HL140170 from the National Heart, Lung, and Blood Institute (Dr Bellows) and grants K24-HL153669, R01-HL134893, and R01-HL1420224 from the National Institutes of Health (Dr Ho).

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: Dr Fonarow is Associate Section Editor of JAMA Cardiology, but he was not involved in any of the decisions regarding review of the manuscript or its acceptance.

Meeting Presentation: This paper was presented at the American College of Cardiology 23rd Annual Scientific Session and Expo; March 4, 2023; New Orleans, Louisiana.

Data Sharing Statement: See Supplement 2.

Additional Information: The simulation model used to create this analysis can be requested from the research team by submitting a research proposal to Drs Bellows or Kazi. If approved, the requester will be asked to sign a Creative Commons agreement prior to release of the model.