Search

Background: Although the Seattle Angina Questionnaire (SAQ) has been widely used to assess disease-specific health status in patients with ischemic heart disease, it was originally developed in a predominantly male population and its validity in women has been questioned., Methods: Using data from 8892 men and 4013 women across 2 multicenter trials and 5 registries, we assessed the construct validity, test-retest reliability, responsiveness to clinical change, and predictive validity of the SAQ Summary Score (SS) and its 5 subdomains (Physical Limitation (PL), Anginal Stability (AS), Angina Frequency (AF), Treatment Satisfaction (TS), and Quality of Life (QoL)) separately in men and women., Results: Comparable correlations of the SAQ SS with Canadian Cardiovascular Society class was demonstrated in both men and women (-0.48 for men, -0.46 for women). Similar correlations between the SAQ PL scale with treadmill exercise duration and Short Form-12 (SF-12) Physical Component Summary were observed in women and men (0.34-0.63 and 0.40-0.63, respectively). SAQ AS scores were significantly lower for both men and women with acute syndromes compared with 1 month later. The SAQ AF scale was strongly correlated with daily angina diaries (0.62 for men and 0.66 for women). The SAQ QoL scores were moderately correlated with the EQ5D visual analog scale and SF-12 general health question in men (0.43-0.50) and women (0.33-0.39). All SAQ scales demonstrated excellent reliability (intraclass correlation ≥0.78) in both men and women with stable CAD and were very sensitive to change after percutaneous coronary intervention (≥15-point difference in scores, standardized response mean ≥ 0.67). The SAQ SS was similarly predictive of 1-year mortality and cardiac re-hospitalizations for both men and women., Conclusion: The SAQ demonstrates similar psychometric properties in men and women with CAD. These findings provide evidence for validity of the SAQ in assessing women with IHD., (Copyright © 2018 Elsevier Inc. All rights reserved.)

Background: Patients with diabetes mellitus (DM) experience higher rates of in-stent restenosis and greater benefit from drug-eluting stents implant at the time of percutaneous coronary intervention (PCI), necessitating prolonged dual anti-platelet therapy (DAPT). While DAPT reduces risk of ischemic events post-PCI, it also increases risk of bleeding. Whether bleeding rates differ among patients with and without DM, receiving long-term DAPT is unknown., Methods: Among patients who underwent PCI and were maintained on DAPT for 1 year in a multicenter US registry, we assessed patient-reported bleeding over one year following PCI in patients with and without DM. Multivariable, hierarchical Poisson regression was used to evaluate the association of DM with bleeding during follow-up., Results: Among 2334 PCI patients from 10 US hospitals (mean age 64, 54% ACS), 32.6% had DM. In unadjusted analyses, patients with DM had fewer bleeding events over the year following PCI (DM vs no DM: BARC = 1: 78.0% vs 87.7%, P < .001; BARC ≥2: 4.3% vs 5.3%, P = .33). Following adjustment, patients with (vs without DM) had a lower risk of BARC ≥1 bleeding during follow-up (relative risk [RR] 0.89, 95% CI 0.83-0.96). This decreased bleeding risk persisted after removing bruising from the endpoint definition., Conclusions: In a real-world PCI registry, patients with DM experienced lower risk of bleeding risk on DAPT. As patients with DM also derive greater ischemic benefit from drug-eluting stents, which requires prolonged DAPT, our findings suggest that the balance between benefit and risk of this therapeutic approach may be even more favorable in patients with DM than previously considered., (Copyright © 2016 Elsevier Inc. All rights reserved.)

Background: A principal goal of treating patients with coronary artery disease (CAD) is to minimize angina and optimize quality of life. For this, physicians must accurately assess presence and frequency of patients' angina. The accuracy with which cardiologists estimate their patients' angina in contemporary, busy outpatient clinics across the United States (US) is unknown., Methods: We enrolled patients with CAD across 25 US cardiology outpatient practices. Patients completed the Seattle Angina Questionnaire before their visit, which assessed their angina and quality of life over the prior 4 weeks. The Seattle Angina Questionnaire angina frequency domain categorized patients' angina as none, daily/weekly, or monthly. After the visit, cardiologists estimated the frequency of their patients' angina using the same categories. Kappa statistic helped to assess agreement between patient-reported and cardiologist-estimated angina., Results: Among 1,257 outpatients with CAD, 67% reported no angina, 25% reported monthly angina, and 8% reported daily/weekly angina. When patients reported no angina, cardiologists accurately estimated this 93% of the time, but when patients reported monthly or daily/weekly angina symptoms, cardiologists agreed 17% and 69% of the time, respectively. Among patients with daily/weekly angina, 26% were noted as having no angina by their physicians. Agreement between patients' and cardiologists' reports (assessed by the kappa statistic) was 0.48 (95% CI 0.44-0.53), indicating moderate agreement., Conclusions: Among outpatients with stable CAD, there is substantial discordance between patient-reported and cardiologist-estimated burden of angina. Inclusion of patient-reported health status measures in routine clinical care may support better recognition of patients' symptoms by physicians., (Copyright © 2016 Elsevier Inc. All rights reserved.)

Background: Drug-eluting stents (DES) reduce restenosis, as compared with bare-metal stents (BMS); however, the relationship between stent type and health status is unknown. We examined whether stent type was associated with health status outcomes in patients undergoing percutaneous coronary intervention (PCI)., Methods and Results: We evaluated 6- and 12-month health status in 2,694 patients with acute myocardial infarction (AMI) enrolled in the TRIUMPH and PREMIER registries who underwent PCI with DES (n = 1,361) or BMS (n = 1,333). Health status was assessed with the Seattle Angina Questionnaire, Medical Outcomes Study Short Form-12, and Patient Health Questionnaire depression scale. Propensity matching was performed to account for baseline differences in patient characteristics, resulting in a comparison cohort of 784 patients treated with DES and 784 patients treated with BMS. Both groups experienced significant improvements in health status at 6 and 12 months after PCI. Drug-eluting stent use was associated with a small improvement in Seattle Angina Questionnaire quality of life and functional limitation scores at 6 months (3.6 [95% CI 0.96-6.21], P = .007, and 3.8 [1.55-6.01], P < .001, respectively), but not at 12 months (2.3 [-0.46 to 5.03], P = .10, and 0.3 [-2.04 to 2.48], P = .85, respectively)., Conclusions: In patients with AMI undergoing PCI, DES use was associated with transient but unsustained health status benefits over 12 months after AMI., (Copyright © 2015 Elsevier Inc. All rights reserved.)

Background: Renin-angiotensin system (RAS) inhibitor use after acute myocardial infarction (AMI) is a quality indicator, but there may also be reasons not to use this therapy. We sought to determine how chronic kidney disease (CKD) and acute kidney injury (AKI) affected RAS inhibitor prescription after AMI in patients with and without decreased ejection fraction (EF)., Methods: Participants from the TRIUMPH registry were categorized by admission estimated glomerular filtration rate (eGFR in mL/min per 1.73 m(2); severe [<30], moderate [30-59], mild [60-89], and no [≥90] CKD) and occurrence of AKI (an increase in creatinine ≥0.3 mg/dL or ≥50%). Renin-angiotensin system inhibitor prescriptions at discharge were compared across categories of CKD, AKI, and decreased EF (<40% vs ≥40%) using a hierarchical modified Poisson model., Results: Among 4,223 AMI patients (mean age 59.0 years, 67.0% male, 67.3% white), RAS inhibitor use decreased significantly with lower eGFR (P < .001), but there was no effect of decreased EF on this relationship (interaction P = .40). Without AKI, severe and moderate CKD were associated with significantly less RAS inhibitor use: relative risks (RRs) 0.67 (95% CIs, 0.58-0.78) and 0.94 (0.90-0.99), respectively. When AKI occurred, CKD was associated with less RAS inhibitor use: RRs 0.84 (0.76-0.93) for mild CKD, 0.78 (0.68-0.88) for moderate CKD, and 0.50 (0.42-0.61) for severe CKD. Ejection fraction <40% was associated with use (RR 1.11, 1.03-1.18), independent of renal function., Conclusions: Chronic kidney disease and AKI are associated with fewer RAS inhibitor prescriptions at discharge, but in both AKI and non-AKI patients, eGFR was more strongly associated with use than EF., (Copyright © 2015 Elsevier Inc. All rights reserved.)

Background: While the process of informed consent is designed to transfer knowledge of the risks and benefits of treatment and to engage patients in shared medical decision-making, this is poorly done in routine clinical care. We assessed the impact of a novel informed consent form for percutaneous coronary intervention (PCI) that is more simply written, includes images of the procedure, and embeds individualized estimates of outcomes on multiple domains of successful informed consent and shared decision-making., Methods: We interviewed 590 PCI patients receiving traditional consent documents and 527 patients receiving novel ePRISM consents at 9 US centers and compared patients' perceptions, knowledge transfer, and engagement in medical decision-making. Heterogeneity across sites was assessed and adjusted for using hierarchical models., Results: Site-adjusted analyses revealed more frequent review (72% for ePRISM vs 45% for original consents) and better understanding of the ePRISM consents (ORs=1.8-3.0, depending upon the outcome) with marked heterogeneity across sites (median relative difference [MRD] in the ORs of ePRISM's effect =2-3.2). Patients receiving ePRISM consents better understood the purposes and risks of the procedure (ORs=1.9-3.9, MRDs=1.1-6.2), engaged more in shared decision-making (proportional OR=2.1 [95% CI=1.02-4.4], MRD=2.2) and discussed stent options with their physicians (58% vs. 31%; site-adjusted odds ratio=2.7 [95% CI=1.2, 6.3], MRD=2.6) more often., Conclusions: A personalized consent document improved the process of informed consent and shared decision-making. Marked heterogeneity across hospitals highlights that consent documents are but one aspect of engaging patients in understanding and participating in treatment., (Copyright © 2014 The Authors. Published by Elsevier Inc. All rights reserved.)

Objective: Pay-for-performance measures, part of the Affordable Care Act, aim to reduce health care costs by linking value with Medicare payments, but until now the concept of value has not been applied to specific procedures. We sought to define value in coronary artery bypass grafting (CABG) and provide a framework to identify high-value centers., Methods: In a multiinstitutional statewide database, clinical patient-level data from 42,839 patients undergoing CABG were matched with cost data. Hierarchical models adjusting for relevant preoperative patient characteristics and comorbidities were used to estimate center-specific risk-adjusted costs and risk-adjusted postoperative length of stay. Variation in value across centers was assessed by the correlation between risk-adjusted measures of quality (mortality, morbidity/mortality) and resource use (costs and length of stay)., Results: There were no significant correlations between risk-adjusted costs and risk-adjusted mortality (r = 0.20, P = .45) or morbidity/mortality (r = 0.15, P = .57) across centers. Risk-adjusted costs and length of stay were not significantly associated (r = 0.23, P = .37) because of cost accounting differences across centers. This may explain the lack of correlation between risk-adjusted quality and risk-adjusted cost measures. When risk-adjusted length of stay and morbidity/mortality were used for the framework, there was a strong positive correlation (r = 0.67, P = .003), indicating that higher risk-adjusted quality is associated with shorter risk-adjusted length of stay., Conclusions: Risk-adjusted length of stay and risk-adjusted combined morbidity/mortality are important outcome measures for assessing value in cardiac surgery. The proposed framework can be used to define value in CABG and identify high-value centers, thereby providing information for quality improvement and pay-for-performance initiatives., (Copyright © 2014 The American Association for Thoracic Surgery. Published by Elsevier Inc. All rights reserved.)

Background: Ranolazine reduces the frequency of angina and use of sublingual nitroglycerin (SL NTG) in stable angina patients with type 2 diabetes (T2DM). Because pre-clinical data suggest that myocardial late sodium current (INaL), the target of ranolazine, is increased by hyperglycemia, we investigated whether the efficacy of ranolazine was influenced by glycemic control., Methods: TERISA was a multinational, randomized, double-blind trial of ranolazine vs. placebo in patients with T2DM and stable angina. Anginal episodes and SL NTG use were recorded daily in an electronic diary. Health status was evaluated at baseline and 8weeks post-randomization using the Seattle Angina Questionnaire (SAQ). The interaction between baseline HbA1c and treatment effect was tested across endpoints using analysis of covariance models, with HbA1c as a continuous variable with restricted cubic splines., Results: The study included 913 patients, with mean age 63.6years, 39% women, mean T2DM duration 7.4years, and mean HbA1c of 7.3%. Heterogeneity of efficacy by HbA1c was observed for the primary endpoint of angina frequency (Pinteraction = .027), the key secondary endpoint of SL NTG use (Pinteraction = .030), SAQ angina frequency (Pinteraction = .001), and SAQ treatment satisfaction (Pinteraction = .025) with greater efficacy of ranolazine in those with higher HbA1c values, increasing continuously from HbA1c levels >6.5%., Conclusion: Among patients with T2DM and stable angina, the therapeutic benefits of ranolazine were greater in those with higher HbA1c values. These data suggest that ranolazine is particularly beneficial in patients with stable angina who have suboptimally controlled T2DM., (Copyright © 2014 Mosby, Inc. All rights reserved.)

Background: Elevated blood glucose is associated with higher mortality in patients with acute myocardial infarction (AMI). Although clinical guidelines recommend targeted glucose control in this group, clinical trials have yielded inconclusive results. Our objective was to understand how this lack of evidence impacts the management of severe hyperglycemia in routine practice., Methods: We examined insulin use among 4,297 AMI admissions with a mean hospitalization blood glucose of ≥200 mg/dL across 55 US hospitals from 2000 to 2008. Temporal trends and interhospital variation in 2 measures of insulin use during hospitalization-any (subcutaneous, intravenous [IV], short acting, long acting) and IV insulin-were examined using hierarchical Poisson regression models., Results: Of the 4,297 admissions, 2,618 (61%) received any insulin and 538 (13%) received IV insulin. After multivariable adjustment, a slight increase in insulin use was observed per admission year (relative risk [RR] 1.06, 95% CI 1.01-1.11). There was a modest (albeit nonsignificant) increase in IV insulin use seen before May 2004 (RR 1.18, 95% CI 0.96-1.47), with no significant change thereafter (RR 0.99, 95% CI 0.92-1.09). Marked variability in insulin use was observed across hospitals (median rate ratio 1.5 [any insulin] and 1.8 [IV insulin]), which did not change over time., Conclusions: Insulin use among patients with AMI and severe hyperglycemia has remained low over the past decade, with substantial and persistent interhospital variation. These observations reflect marked clinical uncertainty with regard to glucose management in AMI, underscoring the imperative for a definitive clinical trial in this field., (Copyright © 2013 Mosby, Inc. All rights reserved.)

Background: Both angina and dyspnea are prevalent symptoms among post-myocardial infarction (MI) patients. Given their frequent overlap, little is known about the degree to which these symptoms provide independent information in this population., Methods: Using the multicenter Prospective Registry Evaluating Myocardial Infarction: Events and Recovery (PREMIER), 1,835 patients were administered the Rose Dyspnea Scale (scores range 0-4; higher scores indicate worse dyspnea), Seattle Angina Questionnaire (SAQ; lower scores indicate worse angina), and 12-item Short Form physical component (PCS; lower scores indicate worse physical health status) at 1 and 12 months post-MI. Multivariable regression was used to examine the association between dyspnea and quality of life (QOL) in both cross-sectional and longitudinal analyses as well as its association with rehospitalizations and mortality., Results: At 1-month follow-up, 863 patients (47%) reported dyspnea, and 340 (19%) noted moderate to severe dyspnea. After adjusting for sociodemographic and clinical factors including SAQ angina frequency, higher dyspnea scores remained strongly associated with worse QOL in both cross-sectional (1-U dyspnea increase = 2.5-point PCS decrease and 2.4-point SAQ QOL decrease) and longitudinal analyses (1-U dyspnea increase from 1 month to 1 year = 3.8-point PCS decrease and 3.5-point SAQ QOL decrease), and with increased risk of 1-year rehospitalization (hazard ratio 1.15/1 U of dyspnea, 95% CI 1.07-1.24) and 3-year mortality (hazard ratio 1.34/1 U of dyspnea, 95% CI 1.19-1.51; P < .001 for all analyses)., Conclusions: Among post-MI patients, dyspnea is common and strongly associated with impaired QOL, more frequent rehospitalization, and reduced survival--independent of both clinical factors and angina. These findings suggest that dyspnea is an important component of disease-specific health status for post-MI patients, and its assessment should be strongly considered in both research studies and clinical practice.

Background: Providing patients with documented discharge instructions is a performance measure of health care quality. It is not well known how often cardiac patients comply with the list of instructions or what their association is with health status outcomes after an acute myocardial infarction., Methods: Acute myocardial infarction patients (N = 2,498) were prospectively enrolled into a 19-center study and asked, at 1 month, if they had recalled receiving instructions at discharge on any of the 13 secondary prevention behaviors (eg, exercise, medications, diet, and smoking). Adherence, defined as the percentage of relevant activities patients reported adhering to at 1 month, was grouped into 4 categories: poor (0%-49%), partial (50%-74%), careful (75%-99%), and very careful (100%)., Results: A total of 2,046 patients completed 1-month interviews and received instruction on at least 1 risk factor management (RFM) behavior. Very careful adherence at 1-month was reported most frequently with "taking medications as prescribed" (94%). In multivariable-adjusted models, patients who reported being poorly adherent were 58% more likely to report angina at 1 year as compared with those who very carefully followed RFM (relative risk 1.58, 95% CI 1.05-2.37). There was no independent association between RFM behavior and quality of life, physical functioning, rehospitalization, or mortality., Conclusions: There is substantial variation in the types of RFM to which acute myocardial infarction patients adhere. In aggregate, stronger adherence was associated with less angina at 1 year. More research is needed to understand adherence patterns and its association with outcomes.

Background: Older age is a risk factor for higher mortality after acute myocardial infarction (AMI), but the association with health status outcomes is largely unexplored., Methods: In a prospective cohort of 2498 patients in the PREMIER study, we compared health-related quality of life (HRQL) and burden of angina symptoms among survivors of AMI by age strata (age groups > or = 75, 65-74, 50-64, and 19-49 years) using the Seattle Angina Questionnaire. Multivariable analyses assessed the relationship between age and 1-year HRQL and angina burden, adjusting for differences in clinical characteristics, treatment, and baseline health status., Results: Older patients comprised a majority: 20.1% were > or = 75 years of age, 41.7% were 65 to 74 years of age, 20.7% were 50 to 64 years of age, and 17.4% were < 50 years of age. At 12 months, older patients had higher mortality (17.0% vs 8.7% vs 6.1% vs 3.2% for age groups > or = 75, 65-74, 50-64, 19-49; P < .001). Among survivors of AMI, increasing age was associated with less angina and better HRQL. By 12 months, older patients reported less angina (10.9% vs 12.7% vs 19.3% vs 23.4% for age groups > or = 75, 65-74, 50-64, 19-49; P < .0001) and better HRQL (scores 89.1 vs 88.1 vs 82.5 vs 80.0, respectively; P < .0001), which persisted after adjustment for baseline angina, HRQL, and other demographic, clinical, disease severity, and treatment differences., Conclusions: Although older patients have higher mortality after AMI, those who survive experience fewer symptoms and better HRQL at 1 year than younger patients. Angina remains present in a number of patients across the spectrum of age, supporting strategies to systematically assess and treat symptoms after AMI.

Background: GJA4 1019 C > T, MMP3 -1171delA, and SERPINE1 -668delG genotypes have been associated with the risk of incident myocardial infarction. We tested the hypothesis that these genotypes would predict long-term mortality after an acute coronary syndrome (ACS)., Methods: We assembled a prospective cohort study on 726 patients with ACS admitted between March 2000 and October 2001. Kaplan-Meier estimates and Cox proportional hazards models of 3-year mortality adjusted for age, race, ACS type, prior heart failure, diabetes, and revascularization were used to compare groups., Results: The GJA4 1019 C > T genotype was significantly related to mortality over 3 years (8.3% vs 14%, for the C/C vs T allele carriers; P = .02), with an adjusted hazard ratio of 1.7 (95% confidence interval 1.05-2.8, P = .03). This finding was consistent in both men and women (hazard ratio = 1.9 and 1.7, respectively) with no significant sex interaction (P = .8). The MMP3 -1171delA and SERPINE1 -668delG genotypes were not significantly related to mortality in the overall population (all P > .4)., Conclusions: GJA4 1019 C > T genotype predicted risk of death after an ACS, whereas the MMP3 and SERPINE1 genotypes did not. The GJA4 1019 C > T polymorphism may warrant integration into comprehensive risk stratification algorithms for patients with ACS.

Background: Patients of low socioeconomic status (SES) have higher mortality after acute myocardial infarction (AMI). Little is known about the underlying mechanisms or the relationship between SES and rehospitalization after AMI., Methods: We analyzed data from the PREMIER observational study, which included 2142 patients hospitalized with AMI from 18 US hospitals. Socioeconomic status was measured by self-reported household income and education level. Sequential multivariable modeling assessed the relationship of socioeconomic factors with 1-year all-cause mortality and all-cause rehospitalization after adjustment for demographics, clinical factors, and quality-of-care measures., Results: Both household income and education level were associated with higher risk of mortality (hazard ratio 2.80, 95% CI 1.37-5.72, lowest to highest income group) and rehospitalization after AMI (hazard ratio 1.55, 95% CI 1.17-2.05). Patients with low SES had worse clinical status at admission and received poorer quality of care. In multivariable modeling, the relationship between household income and mortality was attenuated by adjustment for demographic and clinical factors (hazard ratio 1.19, 95% CI 0.54-2.62), with a further small decrement in the hazard ratio after adjustment for quality of care. The relationship between income and rehospitalization was only partly attenuated by demographic and clinical factors (hazard ratio 1.38, 95% CI 1.01-1.89) and was not influenced by adjustment for quality of care., Conclusions: Patients' baseline clinical status largely explained the relationship between SES and mortality, but not rehospitalization, among patients with AMI.

Background: Diabetes mellitus (DM) is a common comorbidity among patients with acute coronary syndrome (ACS). The frequency with which physicians assess diabetic patients' glycemic control during an ACS hospitalization is not known and may represent an opportunity for quality improvement., Methods: This study describes the proportion of diabetic patients who had an assessment of their glycemic control (HbA1c) at the time of an ACS hospitalization. Secondary analyses examined characteristics associated with HbA1c assessment and physicians' responses to poor glycemic control., Results: Among 968 enrolled patients with ACS, 235 (24%) had DM. HbA1c values were known or obtained in 162 (69%) patients; 60% were poorly controlled (HbA1c > 7). Older patients were less likely to have an HbA1c assessment (relative risk [RR] = 0.81 [95% CI 0.64-1.01] for patients 60-69 years and RR = 0.71 [95% CI 0.58-0.88] for those > or = 70 years compared to patients < 60 years, P = .004). Among patients without an HbA1c, only consultation by an endocrinologist was independently associated with obtaining a subsequent assessment (RR 1.60, 95% CI 1.33-1.92, P < .001). Among those with an elevated HbA1c, 42% with an HbA1c of 7 to 9 and 69% of those with HbA1c > 9 had their diabetic regimen increased., Conclusions: Almost one third of diabetic patients with ACS do not have HbA1c assessment at discharge; particularly older patients and those not evaluated by an endocrinologist. Although > 60% of those assessed had poor control, many did not have adjustments of their diabetic therapy. Assessment of diabetes represents an opportunity to improve the quality of care for diabetic patients with ACS.

Background: More information on the longitudinal care and outcomes of patients after myocardial infarction (MI) is needed to further improve the quality of MI care. The PREMIER study was designed to meet this need., Methods: Patients with MI were prospectively screened and enrolled from 19 US centers between January 1, 2003, and June 28, 2004. Consenting patients had detailed chart abstractions of their medical history and processes of inpatient care, supplemented with a detailed, patient-centered interview. Centralized follow-up at 1, 6, and 12 months is being conducted to quantify patients' postdischarge care and outcomes, with a focus on their health status (symptoms, function, and quality of life). In 2003, detailed chart abstractions, devoid of all personal health information, were collected for patients eligible but not enrolled in PREMIER., Results: Of 10,911 patients screened, 3953 were eligible and 2498 enrolled into PREMIER. Few clinically significant differences between the total MI population and those enrolled into PREMIER were observed. Adherence to accepted processes of quality care, such as aspirin and beta-blockers on admission (96% and 91%) or discharge (96% and 93%), was high. One-month follow-up rates were high, with only 9% of patients being lost to follow-up., Conclusion: PREMIER is a novel registry with detailed insights into patients' sociodemographic, clinical, and health status characteristics, as well as detailed monitoring of their inpatient and outpatient processes of care. Ultimately, PREMIER will describe patients' health status outcomes and identify determinants of these outcomes as an important step toward improving MI care.

Background: To evaluate whether depressive symptoms are independently predictive of mortality and hospitalization among patients with acute myocardial infarction (AMI) complicated by heart failure., Methods: The EPHESUS trial enrolled patients with AMI complicated by heart failure. Patients from Canada, the UK, and the United States completed a Medical Outcomes Study-Depression questionnaire at baseline in addition to a comprehensive clinical examination. Cox proportional hazards regression was used to determine the relationship between depressive symptoms and outcomes, including 2-year all-cause mortality and cardiovascular death or hospitalization, adjusting for baseline clinical variables., Results: Overall, 143 of 634 patients (22.6%) had significant depressive symptoms at baseline (Medical Outcomes Study-Depression score > or = 0.06). Depressed patients had higher 2-year mortality (29% vs 18%; P = .004) and cardiovascular death or hospitalization (42% vs 33%; P = .016). After risk adjustment, depressive symptoms remained significantly associated with mortality (hazard ratio 1.75, 95% CI 1.15-2.68, P = .01) and cardiovascular death or hospitalization (hazard ratio 1.41, 95% CI 1.03-1.93, P = .03). Results were consistent across demographic and clinical subgroups., Conclusions: Depression is an independent predictor of all-cause mortality and cardiovascular death or hospitalization after AMI complicated by heart failure. Although many factors may mediate outcomes in patients with AMI, studies are warranted to evaluate whether a depression intervention can improve survival and/or reduce hospitalizations.