论文部分内容阅读
Aims: To demonstrate the feasibility and clinical utility of developing dynamic risk assessment models for ST- segment elevation myocardial infarction(STEMI) patients. Methods and results: In 6066 STEMI patients enrolled in the Assessment of the Safety and Efficacy of a New Thrombolytic- 3(ASSENT- 3) trial with complete electrocardiographic data, we assessed the probability of 30- day mortality over the following forecasting periods beginning at day 0(baseline), 3 h, day 2, and day 5 using multiple- logistic regression. These models were validated and simplified in independent samples of 1622 similar fibrinolytic- treated p atients from the ASSENT- 3 PLUS trial and in 814 STEMI patients undergoing primary percutaneous coronary intervention in the COMplement inhibition in Myocardial infarction treated with Angioplasty(COMMA)trial. The discriminatory power of these predictive models, from baseline to day 5, was excellent(c- statistics 0.80 to 0.87); and their predictive ability was supported by strong gradients in mortality outcomes as the risk score increased. Dynamic modelling also provided information on the change in prognosis over time which may be used to advise more appropriate therapeutic decisions, e.g. the identification of high- risk patients for possible co- interventions. Conclusion: Dynamic modelling for STEMI patients enhances the risk assessment and stratification and should provide valuable ongoing guidance for their management.
Methods and results: In 6066 STEMI patients enrolled in the Assessment of the Safety and Efficacy of a New Thrombolytic-3 (ASSENT-3) trial with complete electrocardiographic data, we assessed the probability of 30-day mortality over the following forecast period beginning at day 0 (baseline), 3 h, day 2, and day 5 using multiple-logistic regression. These models were validated and simplified in independent samples of 1622 similar fibrinolytic-treated patients from the ASSENT-3 PLUS trial and in 814 STEMI patients undergoing primary percutaneous coronary intervention in the COMplement inhibition in Myocardial infarction treated with Angioplasty (COMMA) trial. The discriminatory power of these predictive models, from baseline to day 5, was excellent (c-statistics 0.80 to 0.87); and their predictive ability was suppo rted by strong gradients in mortality outcomes as the risk score increased. Dynamic modelling also provided information on the change in prognosis over time which may be used to advise more appropriate therapeutic decisions, eg the identification of high- risk patients for possible co-interventions. Conclusion: Dynamic modeling for STEMI patients enhances the risk assessment and stratification and should provide valuable ongoing guidance for their management.