Killian, Jill M.

Background— Although age-adjusted cardiovascular disease (CVD) mortality has declined over the past decades, controversies remain about whether this trend was similar across locations of death and disease categories and about the existence of age and sex disparities. Methods and Results— We examined CVD mortality trends in Olmsted County, Minnesota, between 1979 and 2003 using the categories defined by the American Heart Association, including coronary heart disease (CHD), non-CHD diseases of the heart, and noncardiac circulatory diseases. Data on demographics, cause, and location of death of all 6378 residents who died of CVD were analyzed. Although decreases in the age-adjusted rates occurred in all groups, the magnitude of the decline varied widely. Lesser annual declines were noted in out-of-hospital than in-hospital deaths (1.8% versus 4.8%; P <0.001), in older than in younger persons (1.5% at age ≥85 years versus 3.9% for those ≤74 years of age; P <0.001), and in women relative to men (2.5% versus 3.3%; P =0.007). Furthermore, although CHD showed a marked annual decrease (3.3%), more modest decrements were found for non-CHD diseases of the heart (2.1%) and noncardiac circulatory diseases (2.4%) ( P =0.02 and P =0.04 for the comparison with CHD decline, respectively). Conclusions— Over the past 25 years, CVD mortality declined markedly in the community, but there were large disparities in the magnitude of the decline, resulting in a shift in the distribution toward out-of-hospital and non-CHD deaths. Further reduction in CVD mortality will require strategies directed at elderly persons and women, in whom out-of-hospital rates have improved only minimally.

Background— Several scores, most of which were derived from clinical trials, have been proposed for stratifying risk after myocardial infarctions (MIs). Little is known about their generalizability to the community, their respective advantages, and whether the ejection fraction (EF) adds prognostic information to the scores. The purpose of this study is to evaluate the Thrombolysis in Myocardial Infarction (TIMI) and Predicting Risk of Death in Cardiac Disease Tool (PREDICT) scores in a geographically defined MI cohort and determine the incremental value of EF for risk stratification. Methods and Results— MIs occurring in Olmsted County were validated with the use of standardized criteria and stratified with the ECG into ST-segment elevation (STEMI) and non–ST-segment elevation (NSTEMI) MI. Logistic regression examined the discriminant accuracy of the TIMI and PREDICT scores to predict death and recurrent MI and assessed the incremental value of the EF. After 6.3±4.7 years, survival was similar for the 562 STEMIs and 717 NSTEMIs. The discriminant accuracy of the TIMI score was good in STEMI but only fair in NSTEMI. Across time and end points, irrespective of reperfusion therapy, the discriminant accuracy of the PREDICT score was consistently superior to that of the TIMI scores, largely because PREDICT includes comorbidity; EF provided incremental information over that provided by the scores and comorbidity. Conclusion— In the community, comorbidity and EF convey important prognostic information and should be included in approaches for stratifying risk after MI.