Sex Differences in the Application of Evidence-Based Therapies for the Treatment of Acute Myocardial Infarction:  The American College of Cardiology's Guidelines Applied in Practice Projects in Michigan FREE

The American College of Cardiology (ACC) Acute Myocardial Infarction (AMI) Guidelines Applied in Practice (GAP) program has been implemented in numerous hospitals throughout Michigan.^1- 3 The GAP program incorporates ACC/American Heart Association national AMI standards into everyday care by changing the processes used to treat patients.⁴ Through the use of clinical care tools, such as standardized orders and a discharge document, and systems that reinforce the use of evidence-based AMI care that is often omitted, the GAP program has been shown to increase the use of evidence-based medicines at hospital arrival and discharge, smoking cessation counseling, and dietary counseling.^2- 8 Furthermore, recent data⁹ have revealed that the GAP program reduced mortality at 30 days and 1 year post-AMI.

Much literature has examined differences in symptom presentation and care provided to men and women. Women with AMI have presented to the hospital differently than men. Women are more likely to have nausea and vomiting, neck, back, or jaw pain, and dyspnea and correspondingly less likely to complain of anterior chest pain.^10- 12 Research^13- 15 has also shown that women with AMI are less likely to undergo primary angioplasty within 2 hours and that women presenting with AMI are treated, on average, less aggressively than men. The older age of women and the delay in presentation when seeking care for AMI have been suggested as potential causes for these findings.^16- 18

Building from these studies, we sought to understand the benefits of the GAP program and its associated discharge document on men and women. We studied use of evidence-based care, including medications at discharge, use of the discharge tool, and mortality, between men and women in the GAP program.

Following initial development at the University of Michigan Health System,¹⁹ the ACC AMI GAP projects included a 10-hospital pilot study in southeast Michigan³ in 1999; a 5-hospital project in Flint and Saginaw, Mich,² in 2001; and an 18-hospital expansion project in southeast Michigan¹ in 2002. The AMI GAP projects were conducted in partnership with MPRO, the Medicare Quality Improvement Organization for Michigan; local health coalitions; the ACC; and local hospitals. Each new project built on lessons learned from the previous GAP projects. A rapid-cycle quality improvement model was created, emphasizing a collaborative culture of learning, sharing, and problem solving among hospitals and designing care processes to ensure clinical tool use.¹ The GAP project fosters systems-based care from admission to discharge by incorporating evidence-based tools into practice and targeting patients, physicians, and nurses. Patients included Medicare and non-Medicare recipients diagnosed as having AMI (principal diagnosis code 410.xx, excluding 410.x2). The methods used in the ACC GAP projects have previously been described.^{1- 3,20- 21} The clinical care tool templates are readily available at the ACC Web site (http://www.acc.org/gap/gap.htm).

We examined the impact of GAP on Medicare beneficiaries by studying patients treated in each hospital before and immediately after implementation of GAP. Because follow-up information could only be obtained in Medicare patients, only this population was included in the study. Baseline samples were created using a 50% random sample with at least 20 cases per hospital of Medicare patients with AMI (principal diagnosis code 410.xx, excluding 410.x2) treated in the year preceding GAP implementation. The post-GAP sample included a 95% to 100% sampling of all Medicare patients with AMI in the 4 months immediately after GAP implementation at each hospital. Hospital records for each patient were copied and forwarded to DynKePRO's Centers for Medicare & Medicaid Services' Clinical Data Abstraction Center, Rockville, Md. Each record was screened to ensure that AMI was the principal diagnosis. Then, data regarding patient history, presenting symptoms and signs, comorbidities, diagnostic studies, therapies, outcomes, and use of standard orders and/or the discharge tool were extracted. For quality-of-care indicators, only ideal cases were included in the denominator and, thus, patients with contraindications to evidence-based therapies were excluded from analyses. Ideal cases refer to indicator-specific inclusion and exclusion criteria developed by the Centers for Medicare & Medicaid Services, where patients without absolute or relative contraindications to that indicator remain in the denominator.^22- 23 To ensure quality, data were reabstracted for a random sample of baseline and post-GAP records by the Clinical Data Abstraction Center. Reliability (94%) and accuracy (97%) were high. To assess for 30-day and 1-year mortality, Medicare claims were screened using each patient's unique identifier.

We compared men and women treated before (baseline) and immediately after (post) GAP implementation. By using standard statistics, demographics, presentation variables, comorbidities, diagnostic test results, treatments (including evidence-based therapies and use of AMI discharge tools), and outcomes were analyzed in the 2 cohorts. We examined 30-day and 1-year mortality in the baseline sample compared with men and women in the post-GAP sample. To assess the potential independent benefit of GAP on mortality rates of men and women, separate multivariate logistic regression models for 30-day and 1-year mortality were developed for each sex, including use of the discharge document as an explanatory variable. Patients who died in the hospital were excluded from the patient population used to develop the postdischarge multivariate mortality models. Each hospital was also included as an independent variable into each model to account for practice variation between the facilities. Age, heart rate at admission, and troponin increases were analyzed as continuous variables, while all other variables were dichotomous. Candidate variables in each model included participation in the GAP program; use of the discharge tool; age; history of MI, congestive heart failure, cerebrovascular disease or stroke, hypertension, diabetes mellitus, and chronic obstructive pulmonary disorder; current smoking; percutaneous coronary intervention; and coronary artery bypass grafting. We also assessed in-hospital variables, including heart rate at admission, congestive heart failure, MI location, atrial fibrillation, hematocrit of less than 30%, reduced left ventricular ejection fraction, percutaneous coronary intervention, coronary artery bypass grafting, elevated creatine kinase level, hypotension, shock, stroke, renal failure, transfusion, and hemorrhage.

Women represented 48.1% of the study population, 48.0% of the baseline population, and 48.2% of the post-GAP population. The mean age of men was 69 years; and of women, 75 years.

Baseline and post-GAP demographics, medication use, and unadjusted mortality information are shown in Table 1. Focusing on women, the post-GAP population had a higher rate of history of MI and prior coronary artery bypass grafting. Men in the post-GAP sample were more likely to have hypertension and less likely to have a left ventricular ejection fraction of less than 25% compared with men in the baseline sample.

When men post-GAP were compared with women post-GAP, several interesting results emerge (Table 1). Women were significantly older than men and were less likely to have had a prior MI or prior coronary artery bypass grafting. Women were more likely to have hypertension and a history of congestive heart failure compared with men.

Women in the post-GAP period were more likely to be given aspirin within 24 hours of admission. Men and women in the post-GAP population were more likely to be given aspirin and β-blockers at discharge compared with their counterparts in the baseline group. Rates of lipid-lowering drug use and angiotensin-converting enzyme inhibitor use at discharge were also significantly higher for men in the post-GAP group.

Unadjusted mortality was lower for men post-GAP in the hospital, at 30 days after discharge, and at 1 year after discharge. Women treated after GAP implementation also had significantly lower mortality at 1 year.

Use of the discharge document was substantially higher in men (34.0% vs 1.5%; P<.01) and women (27.9% vs 1.3%; P<.01) after the implementation of GAP.

Women in the post-GAP population were less likely than men to have had the discharge document used. Mortality in the post-GAP population at 30 days and 1 year was also significantly higher in women.

Table 2 focuses on medication use rates at discharge when the discharge document is used. In women, the use of lipid-lowering agents was significantly higher when the discharge document was used and there was a trend toward higher use of aspirin at discharge when the tool was used. Similar results were seen in men when the discharge document was used. Rates of aspirin and lipid-lowering agent use at discharge were significantly higher when the discharge tool was used.

Multivariate models for mortality at 30 days and 1 year showed a trend for the GAP program to lower mortality rates for women (Table 3) and men (Table 4). However, use of the GAP program's discharge tool was highly protective against death at 1 year in women and possibly at 30 days. In men, there was a nonsignificant trend toward the discharge tool being protective against death at 1 year and at 30 days.

This study suggests that use of the GAP discharge tool is important in increasing the use of evidence-based medications and preventing death after AMI at 1 year. Men and women older than 65 years had positive benefits from the GAP program. The discharge tool was correlated with substantial increases in the rates of lipid-lowering drug use at discharge in men and women and increased discharge aspirin use in men. These increases in medication use could explain the promising late outcome results found in our study. In women, use of the discharge document may reduce the odds of death at 1 year after AMI by nearly 50%. Despite these protective benefits, women were significantly less likely to have the discharge document used as part of their process of care compared with men.

Why the discharge document is less likely to be used in women may be explained by the older age of women presenting with an AMI. Several studies^16- 18 have found that as age increases, use of evidence-based treatment decreases, and it is this age × sex interaction that is responsible for the lower use of evidence-based care observed in women. This age × sex interaction may also have resulted in the lower use of the discharge document in women. The women in the post-GAP population were nearly 6 years older than the men, on average. It is possible that as patient age increases, so does uncertainty of physicians as to the potential benefits of evidence-based therapies and, thus, use of the discharge document may decline. The mean age of women in the post-GAP period, however, was nearly 75 years, indicating that much of the female patient population that derived benefits from the GAP intervention was older than 75 years. Furthermore, our mortality results are age adjusted. Thus, we believe the benefits available by using the GAP program and the discharge document should not be reserved only for younger patients, because older patients can experience the positive effects of the program as well. We believe that the GAP program can provide benefits to all patients regardless of age or sex.

Why is the discharge document so successful in reducing death at 1 year after AMI? The discharge document ensures that patients are educated about their condition and understand how to care for themselves after they leave the hospital. It is completed and explained to the patient in the presence of a physician or a nurse and includes instructions on taking medications, goals for controlling cholesterol levels, goals for smoking cessation, instructions on diet and exercise, heart disease education, and instructions for follow-up with the patient's primary care physicians. The patient and the physician or nurse must sign the document, attesting to the patient's understanding of the material included in the document.

The discharge document forces processes to take place that may easily be forgotten in the busy environment of a hospital. For the physician or nurse to explain everything on the document to the patient, the physician or nurse must go through the process to ensure that all the aspects of the document have taken place. For example, use of the GAP discharge document forces physicians to either prescribe aspirin, β-blockers, angiotensin-converting enzyme inhibitors, and cholesterol-lowering agents or document the reason these drugs were not indicated. Similarly, the discharge document ensures that diet counseling and smoking cessation counseling have occurred, that educational information about heart disease and heart attacks has been provided to patients, and that exercise and cardiac rehabilitation instructions and referrals have been given. The success of the discharge document lies in its ability to make certain that evidence-based processes occur at discharge and patients are informed about plans for their medical care after they leave the hospital.

While the discharge document is an integral component in preventing mortality, we do not believe that the results seen in this study would occur without implementation of the entire GAP program. The GAP program emphasizes far more than simply the quality indicators reported herein. The GAP intervention focuses on the patient-physician-nurse triangle of care, trying to ensure that a short list of key care priorities were emphasized in a consistent way, and with systems that supported these priorities during the care itself.

The GAP system also strongly endorses long-term adherence to evidence-based medicines among eligible patients and firmly encourages lifestyle goals in the hospital and at discharge. We suspect that success in achieving multiple pharmacological and lifestyle targets is not an additive effect, but may be multiplicative, just like the risk that is associated with multiple risk factors.²⁴ Furthermore, Fonarow et al²⁵ have shown a substantial reduction in 1-year mortality at an academic institution when a program to improve the rates of evidence-based treatment of AMI was initiated. A study²⁶ of a quality improvement program at Intermountain Health Systems in Utah also demonstrated that increasing the rates of guideline-based therapies for patients with cardiovascular disease also significantly reduced 1-year mortality rates. The discharge document ensures that all the processes of the GAP program are addressed as the patient transitions to the outpatient setting.

Unfortunately, most patients involved in this study did not receive the benefits of the discharge document. Substantial gains can be made in ensuring that all patients with AMI receive the process of care that the GAP program provides. Furthermore, similar programs are being developed for other cardiac and noncardiac conditions to ensure that evidence-based care is provided to a broader array of patients.

The GAP intervention was not a randomized controlled trial and, as such, we cannot prove that the treatment gains seen in this study are solely due to GAP. However, the post-GAP period followed immediately after the pre-GAP period, suggesting that the effect of passive diffusion of care would not be strong. Also, the post-GAP period was only 3½ to 4 months long; we believe that passive diffusion of knowledge would likely take much longer to have a major impact.

In conclusion, the GAP program and its associated discharge document were important in increasing the rates of evidence-based care use in men and women with AMI who were older than 65 years. The GAP discharge tool was independently associated with a substantial reduction in mortality after discharge at 1 year in women. Nevertheless, women were significantly less likely to have the discharge document used as a component of their AMI care. Greater use of the GAP discharge tool in women might narrow the post-MI sex mortality gap.

Correspondence: Kim A. Eagle, MD, University of Michigan Cardiovascular Center, 300 N Ingalls, Room 8B02, Ann Arbor, MI 48109-0477 (keagle@umich.edu).

Accepted for Publication: December 15, 2005.

Financial Disclosure: None.

Funding/Support: This study was supported by the Walter and Esther Hewlett Fund for Cardiovascular Research in women. The GAP projects for AMI care were supported by unrestricted grants from the national American College of Cardiology Foundation, Bethesda, Md; the Centers for Medicare & Medicaid Services, Baltimore, Md; MPRO (the Medicare Quality Improvement Organization for Michigan), Farmington Hills, Mich; Pfizer Inc, New York, NY; AstraZeneca, Wilmington, Del; the Greater Detroit Area Health Council, Detroit, Mich; the Greater Flint Health Coalition, Flint, Mich; the Mardigian Foundation, Bloomfield Hills, Mich; and the University of Michigan, Ann Arbor.

Role of the Sponsor: The funding bodies had no role in data extraction and analyses, in the writing of the manuscript, or in the decision to submit the manuscript for publication.

Disclaimer: The analyses on which this publication is based were performed under contract 500-02-MI-02, entitled “Utilization and Quality Control Peer Review Organization for the State of Michigan,” sponsored by the Centers for Medicare & Medicaid Services, Department of Health and Human Services. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the US government. The authors assume full responsibility for the accuracy and completeness of the ideas presented. This article is a direct result of the Health Care Quality Improvement Program, initiated by the Centers for Medicare & Medicaid Services, which has encouraged identification of quality improvement projects derived from analysis of patterns of care and, therefore, required no special funding on the part of this contractor.Article