OBJECTIVE: To develop a scoring system for predicting in-hospital mortality among ST-elevation myocardial infarction (STEMI) patients submitted to percutaneous intervention (PCI) on intensive cardiac care unit admission by using early and readily available clinical, angiographic and laboratory data. DESIGN: Prospective monocentric observational study in which we used discriminant analysis to develop a final scoring system, with prospective validation. SETTING: Intensive cardiac care unit in Florence, a tertiary center. POPULATION: Five hundred and fifty-eight unselected patients with STEMI (group A) consecutively admitted from 1 January 2004 to 31 December 2006. A control group (group B) comprising 183 STEMI patients admitted from 1 January 2007 to 30 September 2007. MAIN OUTCOMES AND MEASURES: In-hospital death. RESULTS: In group A the discriminant variables were admission Killip class, admission lactic acid, admission ejection fraction, admission troponin I (TnI), admission hemoglobin (Hb), ST-segment reduction post-PCI, systolic blood pressure on admission and chronic renal failure. We elaborated a scoring system, the Florence admission STEMI risk score, which shows an agreement of 95.7% between the observed and the estimated outcome on a statistical basis in the survival and death subgroups. We applied this score to group B (C statistics = 0.986). CONCLUSION: The Florence admission STEMI risk score incorporates anamnestic (chronic renal failure), laboratory (lactic acid, TnI and Hb), procedural and post-procedural data (ST-segment reduction post-PCI, Killip class) as well as data strictly related to infarct size (ejection fraction, TnI). This scoring system is likely to be a simple and practical tool at the bedside for risk evaluation in patients with STEMI submitted to primary PCI.
OBJECTIVE: To develop a scoring system for predicting in-hospital mortality among ST-elevation myocardial infarction (STEMI) patients submitted to percutaneous intervention (PCI) on intensive cardiac care unit admission by using early and readily available clinical, angiographic and laboratory data. DESIGN: Prospective monocentric observational study in which we used discriminant analysis to develop a final scoring system, with prospective validation. SETTING: Intensive cardiac care unit in Florence, a tertiary center. POPULATION: Five hundred and fifty-eight unselected patients with STEMI (group A) consecutively admitted from 1 January 2004 to 31 December 2006. A control group (group B) comprising 183 STEMI patients admitted from 1 January 2007 to 30 September 2007. MAIN OUTCOMES AND MEASURES: In-hospital death. RESULTS: In group A the discriminant variables were admission Killip class, admission lactic acid, admission ejection fraction, admission troponin I (TnI), admission hemoglobin (Hb), ST-segment reduction post-PCI, systolic blood pressure on admission and chronic renal failure. We elaborated a scoring system, the Florence admission STEMI risk score, which shows an agreement of 95.7% between the observed and the estimated outcome on a statistical basis in the survival and death subgroups. We applied this score to group B (C statistics = 0.986). CONCLUSION: The Florence admission STEMI risk score incorporates anamnestic (chronic renal failure), laboratory (lactic acid, TnI and Hb), procedural and post-procedural data (ST-segment reduction post-PCI, Killip class) as well as data strictly related to infarct size (ejection fraction, TnI). This scoring system is likely to be a simple and practical tool at the bedside for risk evaluation in patients with STEMI submitted to primary PCI.