Abstract | PURPOSE: To determine the potential impact of estrogen status on the pretest and postexercise test diagnostic accuracy of exercise testing. PATIENTS AND METHODS: The study comprised a total of 234 women and 326 men who underwent exercise testing followed by coronary angiography. We performed incremental logistic regression analysis of pretest (age, symptoms, smoking, diabetes, cholesterol level) with and without estrogen status (defined according to menopausal and oral estrogen status) and exercise test (two ST-segment and three non-ST-segment) variables separately for men and women. Outcomes were assessed by receiver operating characteristic (ROC) curve area analysis. RESULTS:
Estrogen status was an independent pretest predictor of angiographic coronary disease. Pretest ROC curve areas: women without estrogen status = 0.79, women with estrogen status = 0.85, men = 0.78 (women with estrogen status versus other groups, p < 0.001). Postexercise test ROC curve areas: women without estrogen status = 0.83, women with estrogen status = 0.87, men = 0.88 (women without estrogen status versus other groups, p < 0.001). CONCLUSION: Consideration of estrogen status allowed for a significant improvement in the pretest clinical diagnosis of coronary disease in women. When these improvements were added to the results of exercise testing, the diagnostic accuracy of the combined clinical and exercise test data was similar for men and women. Estrogen status may be an important diagnostic clinical variable in women with suspected coronary disease.
|
Authors | Anthony P Morise, Jyotsna N Dalal, Robert D Duva |
Journal | The American journal of medicine
(Am J Med)
Vol. 94
Issue 5
Pg. 491-496
(May 1993)
ISSN: 0002-9343 [Print] United States |
PMID | 8498394
(Publication Type: Journal Article, Research Support, Non-U.S. Gov't)
|
Chemical References |
|
Topics |
- Adult
- Aged
- Coronary Angiography
- Coronary Disease
(diagnosis)
- Estrogens
- Exercise Test
- Female
- Humans
- Male
- Middle Aged
- Multivariate Analysis
- ROC Curve
- Regression Analysis
- Retrospective Studies
|