Determinants and Prognostic Impact of Heart Failure Complicating Acute Coronary Syndromes
Observations From the Global Registry of Acute Coronary Events (GRACE)
Background— Few data are available on the impact of heart failure (HF) across all types of acute coronary syndromes (ACS).
Methods and Results— The Global Registry of Acute Coronary Events (GRACE) is a prospective study of patients hospitalized with ACS. Data from 16 166 patients were analyzed: 13 707 patients without prior HF or cardiogenic shock at presentation were identified. Of these, 1778 (13%) had an admission diagnosis of HF (Killip class II or III). HF on admission was associated with a marked increase in mortality rates during hospitalization (12.0% versus 2.9% [with versus without HF], P<0.0001) and at 6 months after discharge (8.5% versus 2.8%, P<0.0001). Of note, HF increased mortality rates in patients with unstable angina (defined as ACS with normal biochemical markers of necrosis; mortality rates: 6.7% with versus 1.6% without HF at admission, P<0.0001). By logistic regression analysis, admission HF was an independent predictor of hospital death (odds ratio, 2.2; P<0.0001). Admission HF was associated with longer hospital stay and higher readmission rates. Patients with HF had lower rates of catheterization and percutaneous cardiac intervention, and fewer received β-blockers and statins. Hospital development of HF (versus HF on presentation) was associated with an even higher in-hospital mortality rate (17.8% versus 12.0%, P<0.0001). In patients with HF, in-hospital revascularization was associated with lower 6-month death rates (14.0% versus 23.7%, P<0.0001; adjusted hazard ratio, 0.5; 95% CI, 0.37 to 0.68, P<0.0001).
Conclusions— In this observational registry, heart failure was associated with reduced hospital and 6-month survival across all ACS subsets, including patients with normal markers of necrosis. More aggressive treatment of these patients may be warranted to improve prognosis.
Received March 25, 2003; de novo received July 16, 2003; revision received October 23, 2003; accepted October 24, 2003.
Coronary artery disease, including the acute coronary syndromes (ACS) of unstable angina, non–ST-segment elevation myocardial infarction (NSTEMI) and ST-segment elevation myocardial infarction (STEMI), is the most common cause of heart failure (HF).1 Conversely, heart failure is a frequent complication of ACS2 and significantly worsens the prognosis of patients with ischemic heart disease.3 Owing to the strong association between ACS and HF, it is important to understand the determinants of HF in patients hospitalized with ACS, particularly those with unstable angina, and the impact of HF on outcomes. Although the determinants and impact of HF in patients with acute STEMI have been studied,4,5 little information is available for other ACS subsets, particularly unstable angina.
See p 440
The objective of the GRACE (Global Registry of Acute Coronary Events) registry is to provide data on the treatment, practice patterns, and outcomes of patients with ACS.2,6 The aim of this report is to evaluate the determinants and impact on outcomes of HF complicating ACS in a large unselected population from the GRACE registry.
Full details of the GRACE methodology have been published.2,6,7 Currently, 94 hospitals located in 14 countries are participating in this observational study. Patients must be ≥18 years of age, be admitted for ACS as a presumptive diagnosis, and have at least one of the following: ECG changes consistent with ACS, serial increases in serum biochemical markers of cardiac necrosis, and/or documentation of coronary artery disease.7 The qualifying ACS must not have been precipitated or accompanied by a significant comorbidity, trauma, or surgery. At 6 months after hospital discharge, patients are followed up by telephone, clinic visits, or calls to the primary care physician. Where required, study investigators received approval from their local hospital ethics or institutional review board.
Standardized definitions of all patient-related variables and clinical diagnoses were used as well as hospital complications and outcomes.7 All cases were assigned to one of the following categories: STEMI, NSTEMI, unstable angina, and other cardiac/noncardiac diagnoses. These definitions take into account clinical presentation, ECG findings, and the results of serum biochemical markers of necrosis. Specifically, unstable angina was defined as ACS with normal biochemical markers of necrosis. Patients were categorized at the time of hospital admission according to the classification of Killip and Kimball8 for signs of HF. Because patients with Killip class IV represent a small subset (≈1% of the total cohort) with a dismal prognosis, they were excluded from the data set. To distinguish HF complication on admission for ACS from chronic HF, patients with prior HF were also excluded.
Continuous variables in respective comparison groups are summarized by medians (interquartile range) and were analyzed by means of the Wilcoxon rank sum test. Categorical variables are reported as frequencies and percentages and tested by χ2 test. Ordinal categorical variables were analyzed by a χ2 trend test. Independent predictors of HF were identified by means of stepwise, multivariable, logistic regression. The impact of HF on hospital and postdischarge to 6-month follow-up mortality rates was examined by means of multiple logistic, Kaplan-Meier, and Cox regression analysis, respectively. Predetermined clinically relevant variables “age, gender, diastolic blood pressure, systolic blood pressure, initial creatinine, medical history (angina, smoking, stroke, diabetes, hypertension, coronary artery disease, myocardial infarction, peripheral vascular disease, hyperlipidemia, atrial fibrillation, PCI, CABG, bleeding, and renal dysfunction), cardiac arrest at admission, any significant Q wave at admission, left bundle-branch block, ventricular pacing rhythm, atrioventricular block, atrial fibrillation or flutter, prior use of statins, diuretics, ACE inhibitors, and β-blockers, unstable angina, STEMI, and NSTEMI” were entered into the multivariable backward stepwise logistic and Cox regression analyses. Only significant variables at a value of P<0.05 were retained in the final models.
Between April 1999 and September 2001, 16 166 patients with suspected ACS were enrolled. Of these, 13 707 had a confirmed diagnosis of ACS without prior HF and Killip class IV at the time of hospital presentation, and represent the study sample. Among these, 1778 patients (13%) had HF (Killip class II or III) at hospital admission.
Patients with HF were significantly older than patients without HF and were less likely to be male or smokers (Table 1). A history of comorbidities was more common in patients with HF compared with those without HF. The incidence of HF was similar in patients with STEMI (15.6%) or NSTEMI (15.7%) and was half as frequent in patients with unstable angina (8.2%). Heart rate was higher in patients with versus those without HF. Of the patients with HF, 46.2% had an anterior myocardial infarction compared with 33.6% in those without HF (P<0.0001). Q waves were present in a higher proportion of patients with HF compared with those with uncomplicated ACS (30.2% versus 24%, respectively; P<0.0001) Left bundle-branch block was present in 7.7% of patients with HF versus 3.1% of patients without (P<0.0001). Atrial fibrillation/flutter was recorded in 9.9% of patients with HF compared with 5.2% in those without (P<0.0001).
Patients with HF were less likely to undergo cardiac catheterization procedures than patients without HF (Table 1). ACE inhibitors, diuretics, digoxin, warfarin, antiarrhythmic agents, and inotropic drugs were more frequently prescribed during the acute hospitalization in patients who had HF before admission. The use of β-blockers and statins was significantly lower among patients with HF at the time of hospital admission.
At discharge, patients with HF were more frequently prescribed digoxin (12.0% versus 3.3%, P<0.0001), diuretics (41.4% versus 14.4%, P<0.0001), and ACE inhibitors (63.7% versus 49.5%, P<0.0001) than those without HF. The trend to reduced prescribing of β-blockers and statins in patients with HF continued at discharge (65.1% versus 73.4%, P<0.0001, and (45.7% versus 52.1%, P<0.0001, respectively).
HF at admission was associated with a 4-fold increase in crude hospital mortality rates (12.0% versus 2.9%; OR, 4.6; 95% CI, 3.85 to 5.40). This was true across all ACS subsets, including unstable angina (16.5% versus 4.1% in STEMI; 10.3% versus 3% in NSTEMI; 6.7% versus 1.6% in unstable angina; P<0.0001 for each comparison and for trend). The mortality rates were 2.9%, 9.9%, and 20.4%, for patients in Killip classes I, II, and III, respectively (P<0.0001). In patients with myocardial infarction, the median duration of hospitalization was ≈2 days longer in patients with compared with those without HF (9 versus 7 days, P<0.0001 for STEMI; 8 versus 6 days, P<0.0001 for NSTEMI; 5 days for both in unstable angina, P=0.3174).
When admission predictors of hospital death were analyzed by multivariable logistic regression analysis, an admission diagnosis of HF was associated with a significantly increased risk of dying (adjusted OR, 2.2; 95% CI, 1.75 to 2.67). Patients ≥75 years of age had the highest mortality rates across all ACS subgroups. However, the interaction between age and HF was statistically significant (P=0.0022). Patients <55 years of age had the highest relative increase in mortality rate related to HF, and this was consistent across ACS subsets (Figure 1). When the group of patients with a diagnosis of myocardial infarction was broken down according to the presence or absence of Q waves at discharge, it was apparent that HF on admission had a very strong negative prognostic impact not only on patients with Q-wave myocardial infarction (in-hospital mortality rate, 16.4% versus 5.2%, P<0.0001) but also in those with non–Q-wave myocardial infarction (in-hospital mortality rate, 12.4% versus 3.0%, P<0.0001).
Postdischarge data were available from 11 348 (83.0%) of the patients eligible for follow-up at 6 months; 12.0% of these patients had HF at hospital admission, and HF developed in a further 5.6% during hospitalization. Across all ACS subsets, patients with an admission diagnosis of HF were more likely to be rehospitalized than those with uncomplicated ACS (STEMI, 25.0% versus 14.7%, P<0.0001; NSTEMI, 24.7% versus 15.8%, P<0.0001; unstable angina, 23.1% versus 17.7%, P=0.0200). Compared with similar patients without HF, those with STEMI (19.7% versus 2.8%; P<0.0001) or NSTEMI (11.0% versus 3.6%; P<0.0001) complicated by HF also had 3-fold higher postdischarge mortality rates.
Patients with HF at admission had an approximately 3-fold increase in 6-month postdischarge mortality rate (8.5% versus 2.8%, respectively; P<0.0001) and were also more likely to be rehospitalized (23.6% versus 15.7%, P<0.001) than other patients with ACS. Postdischarge mortality rates did not differ significantly between patients who presented with HF and those in whom HF developed during hospitalization (8.5% versus 7.7%, respectively, P=0.52). Postdischarge use of revascularization was slightly lower in patients with HF compared with those without HF (12.5% versus 14.8%, P=0.02).
Cumulative in-hospital and postdischarge mortality rate was 20.7% in patients with HF on admission compared with 5.9% (P<0.0001) in those without HF on admission (P<0.001; hazard ratio [HR], 3.8; 95% CI, 3.33 to 4.36) (Figure 2). Cumulative mortality rate was even higher in patients who had HF development in the hospital compared with those with HF at admission (25.3% versus 20.7%) (Figure 3). Among patients with HF, those who underwent in-hospital revascularization had a cumulative 6-month mortality rate of 14.0 versus 23.7% in those who did not (P<0.0001). After adjustment for baseline differences, mortality rate remained lower for HF patients who had undergone revascularization (adjusted HR, 0.5; 95% CI, 0.37 to 0.68, P<0.0001).
Admission Predictors of HF
Patient admission characteristics were studied to evaluate the admission predictors of HF. Increased age, raised pulse rate, STEMI, NSTEMI (compared with unstable angina), prior diuretic use, diabetes, and left bundle-branch block were among the factors found to be strong independent predictors of HF (Table 2).
Timing of Development of HF
Of the patients with HF, two thirds had it at presentation and one third had it after admission. Patients with HF on admission tended to be older and were also more likely to have a history of coronary artery disease, myocardial infarction, atrial fibrillation, and CABG (Table 3). Patients with HF on initial evaluation were less likely to undergo cardiac catheterization or CABG, to require mechanical ventilation, or to have a Swan-Ganz catheter inserted than those who had development of HF during hospitalization. They were also less likely to receive ACE inhibitors, β-blockers, diuretics, and inotropic agents. Incidence and mortality rates according to the presence and timing of CHF and to ACS subtype are displayed in Figure 4. Patients who had development of HF later on had a higher mortality rate than those with an admission diagnosis of HF (17.8% versus 12.0%, P<0.0001). This was true overall but also for each ACS subset, except for patients with STEMI, in whom there was a nonsignificant trend. Patients with later HF also had an increased duration of hospitalization compared with patients with HF at admission (10 versus 7 days, P<0.001).
Heart failure on hospital admission was associated with an approximately 3- to 4-fold increase in hospital and 6-month death rates. HF was also associated with longer hospital stay and higher readmission rates. As in previous studies, development of HF during hospital stay (as opposed to HF at admission) was associated with an even worse outcome.4,9 Although it is well known that HF is a frequent and severe complication of acute myocardial infarction,4,5,9–11 this report provides evidence that even in patients with unstable angina and normal biochemical markers of necrosis, HF is a potent predictor of poor outcomes. Using criteria as simple as Killip class, in a subset of almost 5000 patients with unstable angina, HF was associated with a 4-fold increase in mortality rates.
Importantly, there was a reduced frequency of PCI and lower β-blocker usage among patients with HF on admission. Similar findings were reported in a recent study of STEMI.4 In view of the poor prognosis of HF, aggressively identifying patients who are suitable for revascularization would appear to be justified to preserve left ventricular function, prevent left ventricular remodeling, and improve survival. Indeed, revascularization has been shown to be associated with improved survival in patients with acute myocardial infarction and HF12,13 or shock.14 In non–ST-segment elevation ACS, regardless of the presence of HF, three recent trials have provided consistent evidence of the benefit of early revascularization.15–17 Given the high mortality rate of patients with HF and ACS, this group would be expected to derive an even greater benefit from revascularization. Importantly, patients with HF who underwent revascularization had lower cumulative 6-month mortality rates that those who did not, even after adjustment for baseline differences. However, the present study shows that cardiac catheterization and revascularization are underused in patients with HF at admission. Among patients who had development of congestive HF after admission, procedures were performed more frequently, possibly because the development of HF despite therapy was precisely considered ominous.
Limitations of the Study
In this multinational registry, the need to obtain, at certain sites, informed consent from patients to gather follow-up data may have prevented the inclusion of dying patients and biased the sample toward a lower risk group. This may explain the lower crude mortality rates observed among patients with myocardial infarction compared with NRMI 1 and 2.4,5 However, the 3- to 4-fold increase in hospital mortality rates related to HF is remarkably consistent across all these studies.5,18 In GRACE, HF at admission was assessed by clinical examination, which has limitations in the acute setting, as symptoms may be nonspecific and physical examination may lack sensitivity.19,20 Indeed, there may be underrepresentation of patients with atypical symptoms in GRACE, especially when the presenting symptoms are related to HF. However, such low sensitivity would be expected to underestimate the prevalence of HF among patients with ACS and therefore does not detract from our finding of the ominous prognostic implications of HF. In addition, our data suggest that simple clinical signs of HF, as assessed in the Killip classification, are directly linked to higher in-hospital and 6-month mortality rates. These findings extend prior reports, which confirmed the persistent value of the Killip classification in assessing prognosis after acute myocardial infarction21–24 to all ACS subsets. Finally, analyses of postadmission variables related to physician discretion (eg, medications, interventions) may be strongly influenced by unmeasured confounders.
This analysis shows that HF is a common and ominous complication associated with all forms of ACS, including unstable angina. It also suggests that patients with ACS and HF are undertreated. A more aggressive treatment should be recommended in this group of patients to improve outcomes.18
GRACE is supported by an unrestricted educational grant from Aventis Pharma. The authors express their gratitude to all GRACE participants. Further information is available at www.outcomes.org/grace.
Dr Steg has been a speaker and consultant for Aventis Pharma. GRACE is funded by an unrestricted educational grant from Aventis Pharma.
Gheorghiade M, Bonow RO. Chronic heart failure in the United States: a manifestation of coronary artery disease. Circulation. 1998; 97: 282–289.
Spencer FA, Meyer TE, Gore JM, et al. Heterogeneity in the management and outcomes of patients with acute myocardial infarction complicated by heart failure: the National Registry of Myocardial Infarction. Circulation. 2002; 105: 2605–2610.
Fox KAA, Goodman SG, Klein W, et al. Management of acute coronary syndromes: variations in practice and outcome; findings from the Global Registry of Acute Coronary Events (GRACE). Eur Heart J. 2002; 23: 1177–1189.
Emanuelsson H, Karlson BW, Herlitz J. Characteristics and prognosis of patients with acute myocardial infarction in relation to occurrence of congestive heart failure. Eur Heart J. 1994; 15: 761–768.
Wallentin L, Lagerqvist B, Husted S, et al. Outcome at 1 year after an invasive compared with a non-invasive strategy in unstable coronary-artery disease: the FRISC II invasive randomised trial: FRISC II Investigators: Fast Revascularisation during Instability in Coronary artery disease. Lancet. 2000; 356: 9–16.
Fox KA, Poole-Wilson PA, Henderson RA, et al. Interventional versus conservative treatment for patients with unstable angina or non-ST-elevation myocardial infarction: the British Heart Foundation RITA 3 randomised trial: Randomized Intervention Trial of unstable Angina. Lancet. 2002; 360: 743–751.
Remes J, Miettinen H, Reunanen A, et al. Validity of clinical diagnosis of heart failure in primary health care. Eur Heart J. 1991; 12: 315–321.
Lee KL, Woodlief LH, Topol EJ, et al. Predictors of 30-day mortality in the era of reperfusion for acute myocardial infarction: results from an international trial of 41,021 patients: GUSTO-I Investigators. Circulation. 1995; 91: 1659–1668.
Rott D, Behar S, Gottlieb S, et al. Usefulness of the Killip classification for early risk stratification of patients with acute myocardial infarction in the 1990s compared with those treated in the 1980s: Israeli Thrombolytic Survey Group and the Secondary Prevention Reinfarction Israeli Nifedipine Trial (SPRINT) Study Group. Am J Cardiol. 1997; 80: 859–864.