Abstract 3480: Comparison of Albumin Excretion Rate and TIMI Risk Score to Predict In-Hospital Mortality After Acute Myocardial Infarction
The TIMI risk score is a well known strong clinical tool for predicting mortality among patients with ST-elevation myocardial infarction. Aim of the present study was to compare the predictive power of TIMI risk score and Albumin Excretion Rate (expressed as Albumin to Creatinine ratio, ACR) towards in-hospital mortality in 505 consecutive, unselected patients with definite acute myocardial infarction (AMI) admitted to three coronary care units (median age 67; IQ 58–75 years, 29% females, CK-MB peak median 126; IQ 69–236 IU/L). The TIMI risk score ranges from 0 to >8. Accordind to TIMI risk score assessment, the following variables were considered: age (in our study 60% of the patients were older than 65 years), history of diabetes mellitus (24%), hypertension (47%), angina pectoris (20%), systolic blood pressure (5%<100 mmHg), heart rate (7%>100 bpm), Killip class (33%>1), weight (28%<67 Kg), anterior site (33%), time fron onset of symptoms to intensive care unit (42%>4 hours). ACR was measured by radioimmunoassay, in 24-hour urine collection samples, on the 3rd day after admission. Median ACR was 8.7; IQ 3.8–30.8 mg/g. Forty patients (7.9%) died before discharge (all but one because of cardiovascular causes). C-statistic (area under the ROC curve) was 0.840 for TIMI risk score and 0.885 for ACR. At bivariate logistic regression analysis TIMI risk score (OR 1.8; 95%CI 1.5–2.2; z-zcore 6.5) and ACR (OR for 1-SD increase 6.6; 95%CI 3.9–11.0; z-score 7.2) resulted associated to outcome. At multivariable analysis, both TIMI risk score (OR 1.5; 95%CI 1.2–1.8; z-zcore 3.8) and ACR (OR 4.3; 95%CI 2.5–7.4; z-score 5.3) still resulted independently associated to the outcome and showed a negative interaction (p=0.03). At the Likelihood ratio Chi-square statistic (to evaluate the incremental prognostic power of a variable added to a standard model) ACR gave a higher increase (Chi sq=25.4, p<0.0001) than TIMI risk score (Chi sq=4.9, p=0.02) to the standard model. In conclusion, ACR in patients with AMI is more closely associated to, and predictive of in-hospital mortality than TIMI risk score. ACR can be a reliable simple tool to discriminate patients for early hospital discharge.