Clinical and Quality-of-Life Outcomes Following Invasive vs Conservative Treatment of Patients With Chronic Coronary Disease Across the Spectrum of Kidney Function

Key Points

Question  Do clinical and quality-of-life (QoL) outcomes differ following invasive vs conservative management of chronic coronary disease based on kidney function?

Findings  In this secondary analysis of the International Study of Comparative Health Effectiveness With Medical and Invasive Approaches (ISCHEMIA) and ISCHEMIA–Chronic Kidney Disease trials, invasive management was associated with an increase in stroke and procedural myocardial infarction (MI) and a decrease in spontaneous MI with no difference in other outcomes across kidney function. Invasive management was associated with an improvement in QoL only in those with kidney disease stage 3 or lower.

Meaning  The findings indicate that QoL outcomes may be more likely to improve following invasive management in those with less advanced kidney disease.

Importance  Prior trials of invasive vs conservative management of chronic coronary disease (CCD) have not enrolled patients with severe chronic kidney disease (CKD). As such, outcomes across kidney function are not well characterized.

Objectives  To evaluate clinical and quality-of-life (QoL) outcomes across the spectrum of CKD following conservative and invasive treatment strategies.

Design, Setting, and Participants  Participants from the International Study of Comparative Health Effectiveness With Medical and Invasive Approaches (ISCHEMIA) and ISCHEMIA–Chronic Kidney Disease (CKD) trials were categorized by CKD stage: stage 1 (estimated glomerular filtration rate [eGFR] 90 mL/min/1.73m² or greater), stage 2 (eGFR 60-89 mL/min/1.73m²), stage 3 (eGFR 30-59 mL/min/1.73m²), stage 4 (eGFR 15-29 mL/min/1.73m²), or stage 5 (eGFR less than 15 mL/min/1.73m² or receiving dialysis). Enrollment took place from July 26, 2012, through January 31, 2018, with a median follow-up of 3.1 years. Data were analyzed from January 2020 to May 2021.

Interventions  Initial invasive management of coronary angiography and revascularization with guideline-directed medical therapy (GDMT) vs initial conservative management of GDMT alone.

Main Outcomes and Measures  The primary clinical outcome was a composite of death or nonfatal myocardial infarction (MI). The primary QoL outcome was the Seattle Angina Questionnaire (SAQ) summary score.

Results  Among the 5956 participants included in this analysis (mean [SD] age, 64 [10] years; 1410 [24%] female and 4546 [76%] male), 1889 (32%), 2551 (43%), 738 (12%), 311 (5%), and 467 (8%) were in CKD stages 1, 2, 3, 4, and 5, respectively. By self-report, 18 participants (<1%) were American Indian or Alaska Native; 1676 (29%), Asian; 267 (5%), Black; 861 (16%), Hispanic or Latino; 18 (<1%), Native Hawaiian or Other Pacific Islander; 3884 (66%), White; and 13 (<1%), multiple races or ethnicities. There was a monotonic increase in risk of the primary composite end point (3-year rates, 9.52%, 10.72%, 18.42%, 34.21%, and 38.01% respectively), death, cardiovascular death, MI, and stroke in individuals with higher CKD stages. Invasive management was associated with an increase in stroke (3-year event rate difference, 1%; 95% CI, 0.3 to 1.7) and procedural MI (1.6%; 95% CI, 0.9 to 2.3) and a decrease in spontaneous MI (−2.5%; 95% CI, −3.9 to −1.1) with no difference in other outcomes; the effect was similar across CKD stages. There was heterogeneity of treatment effect for QoL outcomes such that invasive management was associated with an improvement in angina-related QoL in individuals with CKD stages 1 to 3 and not in those with CKD stages 4 to 5.

Conclusions and Relevance  Among participants with CCD, event rates were inversely proportional to kidney function. Invasive management was associated with an increase in stroke and procedural MI and a reduced risk in spontaneous MI, and the effect was similar across CKD stages with no difference in other outcomes, including death. The benefit for QoL with invasive management was not observed in individuals with poorer kidney function.

Prior strategy trials of invasive vs conservative management in participants with chronic coronary disease (CCD) have routinely excluded patients with severe kidney disease or, at best, enrolled few such patients. For example, the Bypass Angioplasty Revascularization Investigation 2 Diabetes (BARI 2D) trial¹ excluded patients with serum creatinine greater than 2 mg/dL, the Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation (COURAGE) trial² enrolled only 16 individuals with estimated glomerular filtration rate (eGFR) less than 30 mL/min/1.73m², and the Fractional Flow Reserve vs Angiography for Multivessel Evaluation-2 (FAME-2) trial³ enrolled only 20 patients with serum creatinine greater than 2 mg/dL. Guideline recommendations on the treatment of patients with CCD are therefore not predicated according to underlying kidney function or chronic kidney disease (CKD) stage.^4,5

Prior studies have demonstrated that cardiovascular event rates and risks of invasive procedures are inversely proportional to kidney function.^6,7 As such, kidney function often influences treatment strategies. Moreover, patients with impaired kidney function generally exhibit lower rates of medication adherence and are less likely to achieve medical therapy targets.⁸ As such, it is important to assess whether there is heterogeneity of treatment benefit based on kidney function. In contrast to prior studies, the International Study of Comparative Health Effectiveness with Medical and Invasive Approaches (ISCHEMIA) trials^9,10 enrolled patients across the entire spectrum of kidney function. Our present objective was to compare clinical and quality-of-life (QoL) outcomes across baseline kidney function in patients with CCD and to evaluate associations between these outcomes and initial invasive vs conservative management.

This is a post hoc analysis of the ISCHEMIA and ISCHEMIA-CKD trials, the design and primary results of which have been published previously.^9-12 In brief, participants with site-determined moderate or severe ischemia on stress testing, with and without advanced CKD (eGFR less than 30 mL/min/1.73m² or receiving dialysis), were randomized in the ISCHEMIA-CKD and ISCHEMIA trials, respectively. The original Modification of Diet in Renal Disease (MDRD) formula was used to calculate eGFR.¹³ Participants were categorized based on baseline kidney function into CKD stage 1 (eGFR 90 mL/min/1.73m² or greater), stage 2 (eGFR 60-89 mL/min/1.73m²), stage 3 (eGFR 30-59 mL/min/1.73m²), stage 4 (eGFR 15-29 mL/min/1.73m²), or stage 5 (eGFR less than 15 mL/min/1.73m² or receiving dialysis). The study was approved by the local institutional review boards or ethics committees of each institution and written informed consent was obtained from all patients. Race and ethnicity were self-reported by participants. Reporting race and ethnicity in this study was mandated by the US National Institutes of Health, consistent with the Inclusion of Women, Minorities, and Children policy. Individuals were categorized as Black, Hispanic, or White based on the Department of Health and Human Services’ Implementation Guidance on Data Collection Standards for Race, Ethnicity, Sex, Primary Language, and Disability Status.

Participants were randomized 1:1 to either an initial invasive strategy of coronary angiography followed by revascularization with either percutaneous coronary intervention (PCI) or coronary artery bypass graft surgery when feasible, plus guideline-directed medical therapy (GDMT), or to an initial conservative strategy of GDMT alone, with coronary angiography and revascularization reserved for failure of GDMT. Both groups received secondary prevention that included lifestyle and pharmacologic interventions. In participants with CKD that did not require dialysis who underwent coronary angiography, recommended risk-mitigation measures included use of left ventricular end diastolic pressure-guided hydration strategy and ultra-low- or zero-contrast PCI techniques in which sites were trained.¹⁴

Patients with CKD stage 1 or 2 kidney function were expected to undergo blinded coronary computed tomography angiography to confirm anatomic eligibility for randomization. This was not required for patients with CKD stages 3 to 5. Stress-induced ischemia severity using exercise or pharmacologic stress was rated by independent core laboratories in ISCHEMIA and by sites in ISCHEMIA-CKD.

Self-reported medication-taking behavior was assessed using the modified 4-item Morisky-Green-Levine adherence scale.¹⁵ This is a modification of the original yes/no response scale to allow responses of strongly disagree, disagree, agree, strongly agree, don’t know, and refused for the following questions: “I sometimes forget to take my medicines,” “I am sometimes careless about taking my medicines,” “When I feel better, I sometimes stop taking my medicines,” and “If I feel worse when I take my medicine, sometimes I stop taking it.” Responses of disagree or strongly disagree to all 4 items were classified as being adherent, and all other patterns of response were classified as nonadherent.

The primary clinical end point for this analysis was a composite of death or nonfatal myocardial infarction (MI). The main secondary end point included a composite of death, MI, or hospitalization for unstable angina, heart failure, or resuscitated cardiac arrest. Other secondary and safety clinical end points included the composite of primary and major secondary outcomes using cardiovascular rather than all-cause mortality, composite of death or new dialysis, and the individual components of the composite outcomes. The primary QoL outcome was the Seattle Angina Questionnaire (SAQ) summary score at 1 year following randomization. Individual components of the summary score (angina frequency, QoL, and physical limitation scores) as well as the Rose Dyspnea scale and the EuroQoL Questionnaire (EQ-5D) visual analog scales are also reported.

We summarized categorical variables as counts with percentages and assessed differences between groups using Mantel-Haenszel or Cochrane Armitage trend tests. We present continuous variables as medians with IQRs and compared differences between groups using Spearman correlations. We fitted Cox proportional hazards regression models for the primary end point and each of the secondary end points. We included the following covariates: CKD stage, degree of ischemia, age, sex, smoking, history of hypertension, diabetes, MI, PCI or coronary artery bypass graft surgery, heart failure, ejection fraction, and randomized treatment arm. To test for heterogeneity of treatment effect, we included a CKD stage–by–randomized treatment arm interaction. Given the violation of the proportional hazard assumptions observed in the published primary analysis for ISCHEMIA, we computed cumulative end point event rates at 3 years according to CKD stages and randomized treatment along with the difference in cumulative incidence for the treatment groups. We derived rates from the cumulative incidence function to account for competing risk of death. We conducted companion analyses using eGFR as a continuous variable rather than CKD stage as the primary exposure of interest. The adjustment process accounted for nonlinearity by modeling eGFR with restricted cubic splines. P values for the treatment-by-eGFR interaction were derived by combining the results from 100 imputed data sets. Of the variables used in the models, precise LVEF was imputed for 700 (only an ejection fraction range was provided), degree of ischemia for 82, smoking status for 5, hypertension for 22, prior MI for 18, prior PCI for 4, and history of heart failure for 4. To minimize the impact of nonproportional hazards for some end points, plots were created with the cumulative incidence rates at 3 years across eGFR values separately for each treatment. We derived these estimated rates from separate proportional hazards regression models for each treatment where a restricted cubic spline for eGFR was the only independent variable.

We analyzed 1-year QoL outcomes using proportional odds models. Models included fixed effects for randomized treatment, eGFR, treatment-by-eGFR interaction, and adjustment for baseline QoL and other covariates described above. Analyses were run treating eGFR both as continuous and as categorized by CKD stage. Missing data were imputed using multiple imputation by chained equations with predictive mean matching; models were fit on 100 randomly imputed data sets and pooled using the Rubin method to obtain final parameter estimates. Model results are presented as the association of eGFR with QoL within treatment group (vs a reference eGFR of 103, corresponding to the median value among patients at stage 1) and outcomes following invasive vs conservative treatment as a function of eGFR and CKD stage. Model estimates denote the relative increase in odds of better QoL vs the referent. For the outcome of SAQ-7 angina frequency score, an additional analysis was conducted examining the 3-way association of eGFR, treatment, and baseline angina (ie, whether the association between eGFR and treatment varies with frequency of angina symptoms). We performed all analyses using SAS version 9.4 (SAS Institute). Statistical significance for all analyses was defined by a 2-tailed P < .05.

Among the 5956 participants included in this analysis (mean [SD] age, 64 [10] years; 1410 [24%] female and 4546 [76%] male), 1889 (32%), 2551 (43%), 738 (12%), 311 (5%), and 467 (8%) were in CKD stages 1, 2, 3, 4, and 5, respectively (eFigure 1 in the Supplement). By self-report, 18 participants (<1%) were American Indian or Alaska Native; 1676 (29%), Asian; 267 (5%), Black; 861 (16%), Hispanic or Latino; 18 (<1%), Native Hawaiian or Other Pacific Islander; 3884 (66%), White; and 13 (<1%), multiple races or ethnicities. Compared with patients with lower CKD stages, those with higher CKD stages were more likely to have diabetes, hypertension, heart failure, stroke, or TIA and were less likely to be men, current smokers, or, in the case of CKD stages 3 to 4, White. In addition, patients with higher CKD stages had less angina (higher SAQ angina frequency scores) and higher baseline systolic pressure and were less likely to have severe myocardial ischemia. There were no significant differences between those randomized to invasive vs conservative management within each CKD stage. Select baseline characteristics of ISCHEMIA and ISCHEMIA-CKD participants are outlined in eTable 1 in the Supplement.

Among patients randomized to invasive treatment who underwent coronary angiography, patients with poorer kidney function were more likely to have no epicardial coronary disease and were less likely to have 3-vessel disease or proximal left anterior descending artery disease (Table 1). In those assigned to invasive treatment, coronary angiography and revascularization were performed less frequently in those with advanced CKD (Table 1). In addition, attainment of GDMT goals (systolic BP less than 140 mm Hg, low-density lipoprotein cholesterol less than 70 mg/dL (to convert to mmol/L, multiply by 0.0259), high-intensity statin, aspirin, or aspirin alternative) and adherence to medication regimen were lower in participants with poorer kidney function (Table 2).

There were 921 primary end point events, 1016 main secondary end point events, 481 deaths, and 545 MI events during a median (IQR) follow-up of 3.1 (2.0-4.2) years. Patients with CKD stages 3 to 5 contributed 57% of all deaths despite comprising only 25% of the total cohort. An increase in risk of primary end point (3-year rates 9.52%, 10.72%, 18.42%, 34.21% and 38.01% respectively), main secondary end point, death, cardiovascular death, MI, and stroke was associated with higher CKD stages, both unadjusted and adjusted for baseline variables (Figure 1; eTable 2 in the Supplement). Patients with CKD stage 5 had a 9-fold increase in death (adjusted hazard ratio [aHR], 9.46; 95% CI, 6.94-12.88), 10-fold increase in cardiovascular death (aHR, 10.02; 95% CI, 6.99-14.35), 2-fold increase in MI (aHR, 2.44; 95% CI, 1.79-3.33), and 3-fold increase in stroke (aHR, 3.45; 95% CI, 1.73-6.88) compared with participants with CKD stage 1 (Figure 1; eTable 2 in the Supplement). In addition, an increase in risk of spontaneous MI and heart failure with no difference in the risk of resuscitated cardiac arrest or unstable angina (both low frequency events) was associated with higher CKD stages (eTable 2 in the Supplement).

In participants with higher CKD stages who underwent coronary angiography or revascularization (PCI or coronary artery bypass graft surgery) in either treatment arm, there was an increased risk of periprocedural death, acute kidney injury, and dialysis with no difference in periprocedural stroke or MI (eFigure 2 in the Supplement). There was an increased risk of bleeding associated with higher CKD stages throughout the follow-up period.

There was no significant difference between invasive and conservative management for the primary outcome (eFigure 3 in the Supplement), main secondary outcome, death, cardiovascular death, MI, or the composite of death and new dialysis with no heterogeneity of treatment effect across CKD stages (Figure 2). Invasive management was associated with an increase in stroke (3-year event rate difference, 1% [95% CI, 0.3%-1.7%]) and procedural MI (difference, 1.6% [95% CI, 0.9-2.3]) but a decrease in spontaneous MI (difference, −2.5% [95% CI, −3.9 to −1.1]) with similar association across CKD stages (Figure 2; eFigure 4 in the Supplement). The results were consistent when eGFR was used as a continuous variable (eFigures 5-13 in the Supplement). Periprocedural strokes were infrequent and similar between the treatment strategies and across CKD stages (eTable 3 in the Supplement).

Heterogeneity of QoL outcomes following both treatment approaches based on CKD stages was observed for SAQ-7 summary score; SAQ-7 angina frequency score; SAQ-7 QoL score, and SAQ-7 physical limitation score such that invasive management was associated with a significant improvement in QoL parameters in those with CKD stages 1 to 3 but not in those with CKD stages 4 to 5 (Figure 3). The results were largely similar when eGFR was used as a continuous variable with heterogeneity of treatment for SAQ-7 summary score; SAQ-7 angina frequency score; and SAQ-7 physical limitation score such that invasive management was associated with an improvement in QoL parameters in participants with eGFR 30 mL/min/1.73m² or greater but not in those with lower eGFR (eFigure 15 in the Supplement).

In this secondary analysis of the ISCHEMIA and ISCHEMIA-CKD trials, increased risk of cardiovascular events was found with higher CKD stages over the 3.1-year median duration of follow-up. However, the effect of an initial invasive strategy compared with a conservative strategy was consistent across CKD stages without evidence of significant heterogeneity in clinical outcomes between treatments. The initial invasive strategy was associated with an increase in risk of stroke and procedural MI but a decrease in risk of spontaneous MI with no difference in other outcomes. These associations were consistent across CKD stages. However, there was heterogeneity in QoL outcomes such that invasive management was associated with an improvement in angina-related QoL in those with CKD stage 3 or lower but not in those with poorer kidney function.

More than 2 decades ago, Foley et al¹⁶ observed in the US Renal Data System (USRDS) cohort that cardiovascular disease was the cause of death in more than half of patients with end-stage kidney disease, a proportion 10 to 20 times higher compared with the general population after stratification for age, race, and sex. However, other studies have shown that the mortality from noncardiovascular causes (such as sepsis and pulmonary infections) is also increased in patients with end-stage kidney disease.¹⁷ The results from the present analysis of the ISCHEMIA trials reinforces the finding that cardiovascular cause is the leading cause of death in patients across all stages of CKD in this cohort with abnormal stress test. Cardiovascular mortality and nonfatal MI were much higher in patients with CKD stages 3 to 5 than in those with CKD stages 1 to 2. The increase in nonfatal MI could be the result of greater burden of atherosclerotic risk factors and comorbidities together with challenges introduced by troponin thresholds for the diagnosis of MI in patients with CKD. In addition, the proportion of cardiovascular death compared with noncardiovascular death increased with higher CKD stage (CKD stage 1 to 5: 67.9% to 82.8%) (eFigure 14 in the Supplement). The study reinforces the evidence that cardiovascular rather than end-stage kidney disease is the leading cause of death in this high-risk population. This finding challenges the practice of excluding patients with CKD from cardiovascular clinical trials owing to a misconception that most deaths will likely be noncardiovascular and hence nonmodifiable with cardiovascular therapies. The increase in risk of cardiovascular morbidity and mortality in those with CKD has been attributed to increased prevalence of traditional risk factors (as was seen in this analysis) as well as increased inflammation and metabolic derangements resulting from CKD, leading to a state of accelerated aging of the cardiovascular system.¹⁸ Moreover, patients with poorer kidney function had lower attainment of GDMT goals and had lower adherence to medical therapy, potentially further contributing to worse outcomes.

Whether invasive management can reduce the risk of cardiovascular events and whether the effect of invasive management varies according to kidney function is not well known. Among the 2287 patients randomized in the COURAGE trial,¹⁹ 14% had CKD (defined as eGFR less than 60 mL/min/1.73 m²), which was an independent predictor of death or MI (HR 1.48; 95% CI, 1.15-1.90). However, the PCI group in COURAGE had similar rates of death or MI and similar proportions free of angina compared with persons receiving GDMT alone.¹⁹ Scores for angina frequency improved with PCI for the first 3 months but thereafter were no different compared with GDMT alone in the CKD cohort, whereas the scores remained higher with PCI for at least 2 years and converged at the 3-year time point in the non-CKD cohort.²⁰ Consistent with these findings, in the present analysis, the beneficial effect with invasive management on health status outcomes was not observed in those with advanced CKD. The reasons for this lack of association in patients with lower kidney function is not clear. Patients with poorer kidney function are more likely to have autonomic dysfunction and are therefore less likely to present with angina. In the present study, the proportion of patients with no angina at baseline increased from 33.3% to 57.5% from CKD stage 1 to stage 5. In addition, angina is likely multifactorial in origin, including contribution from microvascular causes that are less likely to respond to revascularization. Moreover, patients with advanced CKD were less likely to be revascularized and more likely to have nonobstructive coronary artery disease, and this may have contributed to the lesser association with QoL among participants with advanced kidney disease. Other trials of an invasive vs conservative strategy have contributed little to our understanding of the management of patients with CKD. In the BARI-2D trial,²¹ 21% of randomized participants had eGFR less than 60 mL/min/1.73 m², but their outcomes were not reported separately. FAME-2²² enrolled only 20 patients with serum creatinine greater than 2 mg/dL, and their outcomes were not presented separately.

Limitations of our study include lower rates of coronary angiography and revascularization with higher CKD stage. There were notable differences in baseline characteristics, coronary computed tomography angiography requirements, and method of ascertainment of ischemia severity (site reading vs core laboratory reading) between the ISCHEMIA¹⁰ and ISCHEMIA-CKD⁹ trials. However, the analysis was stratified by baseline kidney function without overlap of the 2 cohorts. Moreover, the differences in the characteristics of the 2 trials should not differentially affect the randomized treatment groups (invasive vs conservative). In addition, most participants in CKD stages 3 to 5 did not undergo coronary computed tomography angiography to confirm the presence of obstructive coronary artery disease. Moreover, although the sites were trained to use ultra-low volume contrast protocol for coronary angiography and intervention for patients with advanced CKD, the trials did not collect the amount of contrast used. The sizes of individual CKD stages were not adequately powered for any end point. In contrast to MI, stroke rates were low, and procedural stroke rates were not higher in patients with advanced CKD receiving invasive treatment. The increased risk of stroke after invasive management and the association between CKD stage, treatment strategy, and risk of stroke should therefore be considered with caution. The results do not apply to patients with acute coronary syndromes, left main coronary artery disease, heart failure, or ejection fraction less than 35% or those who are highly symptomatic. Additionally, although participants were categorized based on eGFR, other measures of kidney dysfunction, such as albuminuria, were not collected.

In conclusion, in patients with CCD and moderate or severe myocardial ischemia, a higher CKD stage was associated with a worse prognosis with high risk of cardiovascular morbidity and mortality. Invasive management was associated with higher rates of stroke and procedural MI but lower risk of spontaneous MI with no difference in other clinical outcomes, including the primary outcome and death. Consistent associations were observed across CKD stages. Invasive management was associated with an improvement in angina-related QoL but not in those with poorer kidney function.

Accepted for Publication: May 9, 2022.

Published Online: June 29, 2022. doi:10.1001/jamacardio.2022.1763

Corresponding Author: Sripal Bangalore, MD, MHA, New York University Grossman School of Medicine, New York, NY 10016 (sripalbangalore@gmail.com).

Author Contributions: Ms Stevens and Mr Jones had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Bangalore, Hochman, Spertus, Fleg, Williams, Stone, Sidhu, Maron.

Acquisition, analysis, or interpretation of data: Bangalore, Hochman, Stevens, Jones, O’Brien, Reynolds, Boden, Fleg, Williams, Mathew, Chertow.

Drafting of the manuscript: Bangalore.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Stevens, Jones, O’Brien.

Obtained funding: Hochman, Williams.

Administrative, technical, or material support: Reynolds, Fleg, Sidhu.

Supervision: Bangalore, Hochman, Spertus, Williams, Stone, Chertow.

Conflict of Interest Disclosures: Dr Bangalore reports grants from National Heart, Lung, and Blood Institute during the conduct of the study and grants and personal fees from Abbott Vascular and personal fees from Biotronik, Pfizer, Amgen, and Reata outside the submitted work. Dr Hochman is principal investigator for the International Study of Comparative Health Effectiveness with Medical and Invasive Approaches (ISCHEMIA) trial for which, in addition to support by National Heart, Lung, and Blood Institute grant, devices and medications were provided by Abbott Vascular, Medtronic, St Jude Medical, Volcano Corporation, Arbor Pharmaceuticals, AstraZeneca Pharmaceuticals, Merck Sharp & Dohme Corp, and Omron Healthcare, as well as financial donations from Arbor Pharmaceuticals and AstraZeneca Pharmaceuticals. Ms Stevens reports grants from National Heart, Lung, and Blood Institute during the conduct of the study. Mr Jones reports employment by the National Heart, Lung, and Blood Institute during the conduct of the study. Dr Spertus reports grants from National Heart, Lung, and Blood Institute during the conduct of the study and personal fees from Bayer, Novartis, AstraZeneca, Amgen, Janssen, United Healthcare, and Blue Cross Blue Shield of Kansas City and grants from American College of Cardiology outside the submitted work; in addition, Dr Spertus has a patent copyright to Seattle Angina Questionnaire with royalties paid and equity in Health Outcomes Sciences. Dr O’Brien reports grants from National Heart, Lung, and Blood Institute during the conduct of the study. Dr Reynolds reports grants from National Heart, Lung, and Blood Institute during the conduct of the study and nonfinancial support from Abbott Vascular, Siemens, and BioTelemetry outside the submitted work. Dr Boden reports grants from National Heart, Lung, and Blood Institute during the conduct of the study and grants from AbbVie, Amarin, and Amgen and personal fees from Amgen, Cleveland Clinic Clinical Coordinating Center, and Janssen outside the submitted work. Dr Fleg reports employment by the National Heart, Lung, and Blood Institute during the conduct of the study. Dr David O. Williams reports grants from National Heart, Lung, and Blood Institute during the conduct of the study. Dr Gregg W. Stone reports grants and personal fees from the National Heart, Lung, and Blood Institute during the conduct of the study and personal fees from Terumo, Amaranth, Shockwave, Valfix; TherOx, Revascularization, Vascular Dynamics, Robocath, HeartFlow, Gore, Ablative Solutions, Matrizyme, Miracor, Neovasc, V-wave, Abiomed, Claret, Sirtex, Ancora, Qool Therapeutics, SpectraWave, MAIA Pharmaceuticals, Orchestra Biomed, and Vectorious and other from Cagent, Applied Therapeutics, Biostar, MedFocus, Aria, and Cardiac Success outside the submitted work. Dr Sidhu reports grants from National Heart, Lung, and Blood Institute during the conduct of the study and personal fees from Astra Zeneca and Sanofi-Regeneron outside the submitted work. Dr Mathew reports grants from National Heart, Lung, and Blood Institute during the conduct of the study. Dr Chertow reports grants from National Heart, Lung, and Blood Institute during the conduct of the study and grants from National Institute of Diabetes and Digestive and Kidney Diseases and Amgen; personal fees from Akebia, Ardelyx, AstraZeneca, Baxter, Cricket, DiaMedica, Gilead, Reata, Sanifit, Vertex, Satellite Healthcare, Angion, Bayer, and ReCor; and other from CloudCath, Durect, and Outset outside the submitted work. Dr Maron reports grants from National Heart, Lung, and Blood Institute during the conduct of the study. No other disclosures were reported.

Funding/Support: This work was supported by National Institutes of Health (grants U01HL105907, U01HL105462, U01HL105561, U01HL105565, U01HL117904, and U01HL1179050), clinical translational science awards (11UL1 TR001445 and UL1 TR002243) from the National Center for Advancing Translational Sciences, and by grants from Arbor Pharmaceuticals and AstraZeneca Pharmaceuticals. Devices or medications were provided by Abbott Vascular, Medtronic, Phillips, and Omron Healthcare; medications were provided by Amgen, Arbor Pharmaceuticals, AstraZeneca Pharmaceuticals, Espero Pharmaceuticals, Merck, Sharp & Dohme Corp, and Sunovion Pharmaceuticals.

Role of the Funder/Sponsor: The trial design was peer-reviewed by National Heart, Lung, and Blood Institute–appointed external reviewers. National Heart, Lung, and Blood Institute program staff had a role in oversight of the conduct of the study; they did not have a role in the data collection, management, analysis and interpretation of the data; preparation, review or approval of the manuscript, or decision to submit the manuscript for publication.

Disclaimer: The contents of this manuscript are solely the responsibility of the authors and do not necessarily represent official views of the National Center for Advancing Translational Sciences, the National Heart, Lung, and Blood Institute, the National Institutes of Health, or the US Department of Health and Human Services.

Additional Contributions: We thank the investigators at the International Study of Comparative Health Effectiveness with Medical and Invasive Approaches (ISCHEMIA) and International Study of Comparative Health Effectiveness With Medical and Invasive Approaches–Chronic Kidney Disease (ISCHEMIA-CKD) sites and to the participants who made this study possible.