THE DEMOGRAPHICS ASSOCIATED WITH THE USE OF THE CARDIAC TROPONIN TEST IN A UNIVERSITY EMERGENCY DEPARTMENT.

Abstract

Purpose To analyze the pattern of use of troponin testing in an ethnically diverse urban emergency department (ED) setting. Methods A database was constructed from clinical and demographic information along with cardiac troponin (cnT) results on 3,366 consecutive patients presenting to the University of New Mexico ED from January 2005 to November 2005 with a diversity of chief complaints. A clinically relevant “positive” is defined as > 0.1 ng/mL. Data are summarized as mean ± SD for continuous variables and counts or percentages for dichotomous variable. Sensitivity, specificity, and predictive values were computed using the discharge diagnosis (DRG) as a “gold standard.” Results There were 6,204 recorded cnT results for 3,366 patients, with 8% “positive” and 92% negative. Only 34% of patients had more than one cnT determination. Of the 3,366 patients, 1,741 (52%) were women and 1,625 (48%) were men. The distribution based on race was 41% Hispanic, 27% Caucasian, 23% other, 5% American Indian, 3% African American, and 1% Oriental. The age ranged from 8 to 104, with the average age for men being 51 (range 8-104) and the average age for woman being 56 (range 10-102). The overall rate of DRG for any type of acute myocardial infarction was 193 or 21%. Thus, the sensitivity is 61%, the specificity is 55%, the positive predictive value is 17%, and the negative predictive value is 90%. Conclusions (1) In the majority of patients only one cnT was obtained. (2) The sensitivity and specificity as well as the positive predictive value of a positive cnT in this setting were low, indicating a low pretest probability of disease in this ethnically and demographically diverse population. (3) These data suggest that cnT is a poor screening test for the detection of cardiac ischemic syndromes in similar populations.