Predicting in-hospital mortality in acute myocardial infarction: Impact ofthrombolytic therapy on APACHE II performance

Objective: To study the usefulness of the Second Acute Physiology and Chron ic Health Evaluation (APACHE II) scoring system for prognostication of in-h ospital mortality in patients with acute myocardial infarction treated with thrombolysis. Design: A prospective validation study was conducted at a medical intensive care unit at a university hospital. Over a 3-year period, 1714 patients wi th acute myocardial infarction were studied (mean age 72 +/- 10 years). Thr ombolytic therapy was the prescribed treatment for 316 patients and total h ospital mortality was 16%. Results: The patients who received thrombolysis were younger, had higher bl ood pressure, lower heart and respiratory rates and higher Glasgow Coma Sca le scores. Total in-hospital mortality was 9.5% in patients treated with th rombolysis and 17.1% in untreated patients (p < 0.01). Corresponding APACHE II predictions of mortality were 11.8 and 15.8% (p < 0.01). There was no s ignificant difference between observed and predicted mortality. When a deci sion rule of 50% predicted risk of death was employed, sensitivity was 20% and specificity 99% in the thrombolytic group, while the corresponding figu res in the nonthrombolytic group were 31% and 97%, respectively. Conclusions: In-hospital mortality in groups of patients treated with or wi thout thrombolysis for acute myocardial infarction could be predicted with the APACHE II scoring system. Prognostication in individual patients is not possible with the APACHE II system.