The 131st Annual Meeting (November 15-19, 2003) of APHA

Abstract #70224

Michael A Posner, MS¹, Arlene Ash, PhD², and Eric Green, MD². (1) Department of Biostatistics, Boston University, 720 Harrison Ave #1108, Boston, MA 02118, 617-638-8151, mposner@bu.edu, (2) Health Care Research Unit, Boston University School of Medicine, 720 Harrison Ave., Suite 1108, Boston, MA 02118

Acute Myocardial Infarction (AMI) is a leading cause of death in the United States, with over 395,000 AMI hospitalizations among Medicare beneficiaries annually, costing the health care system $3.6 billion. After more than a decade of decline, one year mortality post-AMI has begun to increase. To what extent are people coming in for second heart attacks that previously would have died on the first one a factor? The incidence of heart attacks during the year prior to hospital admission has, in fact, been increasing. We sought to characterize the association between the presence of a previous AMI and one-year mortality (after the subsequent AMI admission), using 1995 fee-for-service Medicare data. A case is defined as a Medicare hospital admission with a principal diagnosis code of AMI (ICD9-CM code 410). Admissions within one day of discharge were rolled up into one record. Only those with complete part A and part B coverage for the entire year prior to their subsequent AMI were included. 18.6% of our cases were hospitalized with a primary diagnosis of AMI in the previous year. The crude odds of dying was 24.7% lower for those that had an AMI in the previous year relative to those that did not have one. Controlling for age, gender, original reason for entitlement, and DCG Score (a measure of comorbidity burden) through multiple logistic regression, the odds of dying increased to 44.4% less. In order to test the sensitivity of our results to our definition of previous AMI, a broader definition was examined. 31.4% of all cases had any diagnosis of AMI recorded in the previous 365 days - an ICD9-CM code of 410 (AMI) or 412 (previous MI) in any inpatient, outpatient, or physician/supplier record. The effect sizes were smaller than in the previous model, but still quite significant. Patients may change their diet, habits, and medications after surviving the first AMI or are more likely to identify the onset of an AMI and get to the hospital quicker. Alternatively, recurrent AMIs may be systematically different than first-time AMI or may in fact be a side-effect of interventions. Patients with previous AMI often develop collateral circulation to bypass stenotic coronary arteries, which may reduce the impact of coronary artery stenosis. Finally, it may also be a type of “hardy” person that is less susceptible to death following an AMI.

Learning Objectives:

• Understand the association between having a previous AMI and mortality following subsequent AMI
• Evaluate the methodology used to assess this association.

Keywords: Medical Care, Heart Disease

Presenting author's disclosure statement:
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.