Prognostic value of ST segment depression in acute coronary syndromes: Insights from PARAGON-A applied to GUSTO-IIb

OBJECTIVES Our objectives were to develop a risk-stratification model addre ssing the importance of the magnitude and distribution of ST segment depres sion in predicting long-term outcomes and to validate the model in an analo gous patient population. BACKGROUND Although patients without ST segment elevation presenting with a cute coronary syndromes represent an increasingly frequent population admit ted to coronary care units, little attention has been paid to quantifying t heir ST segment abnormalities. METHODS ST segment depression was categorized into three groups: 1) no ST s egment depression; 2) l-mm ST segment depression in two contiguous leads; a nd 3) ST segment depression greater than or equal to2 mm in two contiguous leads. A logistic regression model was developed using Platelet IIb/IIIa An tagonism for the Reduction of Acute coronary syndrome events in a Global Or ganization Network (PARAGON-A) data to assess the prognostic value of the e xtent and distribution of ST segment depression in predicting one-year mort ality. The model was validated using the non-ST segment elevation populatio n in Global Use of Strategies To Open occluded arteries in acute coronary s yndromes (GUSTO-IIb). RESULTS ST segment depression was the strongest predictor of one-year morta lity, accounting for 35% of the model's predictive power. Patients with ST segment depression greater than or equal to2 mm were similar to6 times (odd s ratio [OR] 5.73, 95% confidence interval [CI] 2.8 to 11.6) more likely to die within one year than patients with no ST segment depression. On valida tion, the model showed good discriminatory power (c-index = 0.75). Patients with ST segment depression greater than or equal to2 mm in more than one r egion were almost 10 times more likely tc, die within one year than patient s with no ST segment depression. CONCLUSIONS These data provide new evidence supporting the powerful prognos tic value of the baseline electrocardiogram and, in particular, the magnitu de and distribution of ST segment depression in predicting unfavorable even ts. (J Am Coll Cardiol 2001;38:64-71) (C) 2001 by the American College of C ardiology.