After acute myocardial infarction (AMI), patients with reduced left ventricular ejection fraction (EF) are at increased risk of recurrent cardiovascular events, arrhythmias, heart failure, and mortality. Aldosterone antagonist therapy improves survival in patients with chronic heart failure and depressed EF ((1),(2),(3),4). The EPHESUS (Eplerenone Post-Acute Myocardial Infarction Heart Failure Efficacy and Survival Study) demonstrated an incremental mortality benefit of the use of the selective aldosterone receptor antagonist eplerenone in AMI patients with an EF ≤40% and heart failure or diabetes when added to other standard-of-care therapies (5). These benefits were independent of diuretic and potassium-sparing effects and occurred within the first 30 days of initiation including significant reductions in all-cause mortality, sudden arrhythmic death, and cardiovascular rehospitalization (6). The American College of Cardiology/American Heart Association ST-segment myocardial infarction (STEMI) guidelines published after the EPHESUS trial gave the use of aldosterone inhibitors in post-STEMI patients with an EF ≤40% and either symptomatic heart failure or diabetes without contraindications (renal dysfunction, hyperkalemia) a Class I recommendation ((7),8). Non–ST-segment myocardial infarction (NSTEMI) guidelines reflect similar recommendations (9). Numerous studies have evaluated the adoption of aldosterone antagonists in chronic heart failure ((10),(11),12); however, little is known regarding their adoption in patients after AMI. We sought to determine the use and temporal trends of aldosterone antagonist therapy during the initial hospitalization for AMI in clinical practice.

Get With the Guidelines–Coronary Artery Disease is an ongoing prospective registry and quality improvement program that has previously been described. For our analysis, patients admitted with AMI were included. Data were collected by participating hospitals without financial compensation and entered by trained personnel. All participating institutions were required to comply with local regulatory and privacy guidelines and, if required, to secure institutional review board approval. Outcome Sciences, Inc. (Cambridge, Massachusetts) served as the registry coordinating center. The Duke Clinical Research Institute (Durham, North Carolina) served as the data analysis center.

At the time of our analysis, the registry included 121,041 patients between July 1, 2006, and December 29, 2009. We excluded patients without a diagnosis of AMI, with an EF ≥40% (EF = 40% excluded to prevent inclusion of patients with an EF in the range of 40% to 45% who have no proven benefit from therapy), with contraindications to aldosterone antagonist therapy (serum potassium level >5.0 mEq/l, serum creatinine level >2.5 mg/dl, or a listed drug allergy), with a history of long-term dialysis, with a discharge destination other than home (skilled nursing facility, rehabilitation center, hospice, or transfer to another hospital) and who received comfort care measures only (Figure 19_gr1).

The primary outcome of the study was guideline-directed prescription of an aldosterone antagonist at the time of discharge. Patient and hospital characteristics associated with aldosterone antagonist use were also evaluated. Adherence to AMI performance and quality measures was assessed to compare differences in the quality of care between groups.

Categorical variables are expressed as frequencies with percentages using the Pearson chi-square test and continuous variables presented as medians with interquartile ranges using chi-square rank–based group means score test in testing for associations among the use of aldosterone antagonists, patient/hospital characteristics, and outcomes.

A multivariable logistic regression model was fitted to examine significant factors associated with the use of aldosterone antagonists. Variables with a p value <0.10 in the univariate screen were considered for the multivariate model. A backward selection algorithm was used to identify variables significant in the multivariate model. The generalized estimating equation method with exchangeable working correlation structure was used to account for within-hospital clustering in the final model. Unadjusted and adjusted results are presented as odds ratios and 95% confidence intervals. A p value <0.05 was considered significant. Temporal trends in the use of aldosterone antagonists from 2006 to 2009 were explored. p Values were based on chi-square rank–based group means score statistics. Statistical analyses were performed by the Duke Clinical Research Institute using SAS software (version 9.2, SAS Institute, Cary, North Carolina).

In all, 11,255 (13.8%) post-AMI patients from 219 hospitals were eligible for aldosterone antagonist therapy. The mean age was 66 years, 34.4% were female, and 75.9% were white. Most patients were admitted with NSTEMI with a median EF of 30% (interquartile range, 25% to 35%). Of eligible patients, 1,023 (9.1%) received an aldosterone antagonist at hospital discharge. When restricting the analysis to patients with diabetes, a history of heart failure or diabetes and/or a history of heart failure, 399 (11.2%), 349 (14.4%), and 571 (11.8%) patients were prescribed aldosterone antagonists, respectively. (Table 1) describes the study population. Hospital characteristics and adherence to performance and quality measures based on aldosterone antagonist prescription are shown in Tables (Table 2) and (Table 3), respectively.

A wide variation in prescription of aldosterone antagonists was noted (0% to 100%). When confining the analysis to 144 hospitals with ≥10 patients meeting American College of Cardiology/American Heart Association guideline criteria, aldosterone antagonist prescription by hospital remained variable (median, 7.0%; interquartile range, 1.9% to 13.4%; range, 0% to 40.0%). In these 144 hospitals, the median number of patients eligible for aldosterone antagonists was 42.5 (interquartile range: 21.5 to 88.5). Patient and hospital characteristics independently associated with prescription of aldosterone antagonists were identified (Table 4). Among eligible patients, prescription of an aldosterone antagonist increased from 6.0% to 13.4% (p < 0.001) between January 2006 and December 2009. There were similar temporal trends among the subgroups of patients with diabetes, a history of heart failure or diabetes, and/or a history of heart failure, with an increase in prescription rates from 6.9% to 15.4%, 8.5% to 20.2%, and 7.0% to 16.9%, respectively (p < 0.001) (Figure 19_gr2).

Our study shows that the majority of eligible patients are not prescribed aldosterone antagonists after AMI. This highlights one of the largest treatment gaps between evidence-based guidelines and current treatment patterns for patients after AMI. Despite the increasing trends seen in temporal analysis, aldosterone antagonist use remains extremely low. Current American College of Cardiology/American Heart Association guidelines for NSTEMI and STEMI recommend pre-discharge and long-term aldosterone antagonist therapy for patients with an EF ≤40% already receiving angiotensin-converting enzyme/angiotensin receptor blocker and beta-blocker therapy unless contraindicated ((8),9). These recommendations are largely based on the EPHESUS, which demonstrated a 15% reduction in all-cause mortality and a 17% reduction in cardiovascular mortality at a mean follow-up of 16 months (5).

Possible explanations for the disconnect between guideline recommendations and clinical implementation include concern for safety (kidney dysfunction/hyperkalemia), effectiveness of laboratory monitoring, misunderstanding of aldosterone-blocking benefits as opposed to a potassium-sparing diuretic, evidence of use based on a single randomized, multicenter trial, and plans to initiate therapy after hospital discharge once angiotensin-converting enzyme/angiotensin receptor blocker doses have been up-titrated ((10),(13),14). Although a Class IA recommendation in STEMI and NSTEMI guidelines, most patients in our analysis had a diagnosis of NSTEMI, a potential factor, as NSTEMI guidelines did not adopt this recommendation until 2007. Despite potential safety concerns, mortality benefit is well documented, and this therapy has been demonstrated to be cost-effective (15). Physicians should consider the potential risks versus benefits and, as guidelines suggest, monitor potassium and creatinine levels carefully.

Factors associated with aldosterone antagonist use were diabetes, heart failure, EF, and absence of kidney dysfunction. Additionally, we found that <1% of those with documented contraindications were prescribed this therapy, suggesting that physicians who prescribe these agents do so in a thoughtful manner. The rates of aldosterone antagonist prescription seen from 2006 to 2009 are increasing; however, they are still far too low.

Despite the many strengths of our study, including the large national sample size, limitations include the use of observational registry data, which can be subject to error during chart review, limited availability of discharge laboratory data, inability to document a contraindication to therapy unless listed in the reporting forms, and inability to identify those with clinical heart failure. However, during subgroup analyses of patients with diabetes, a history of heart failure or diabetes and/or previous heart failure (likely an underestimation of the treatment gap), prescribing rates were only slightly better and continue to be far less than those seen with other guideline recommended therapies for AMI. No information on outpatient follow-up or a safety measure of rates of post-discharge hyperkalemia was available. Get With the Guidelines–Coronary Artery Disease hospitals are self-selected, and the treatments in hospitals' patient characteristics and care patterns may differ.

Although rates of aldosterone antagonist use are increasing slightly over time, there is wide variability in its use, and the majority of AMI patients eligible for treatment fail to receive it at hospital discharge. The reason for this discrepancy between guideline-based therapy and actual prescribing patterns is unclear and should be further studied.