Characteristics of an Albumin Cobalt Binding Test for assessment of acute coronary syndrome patients: a multicenter study.

BACKGROUND The ability of the N-terminal region of human albumin to bind cobalt is diminished by myocardial ischemia. The characteristics of an assay based on albumin cobalt binding were assessed in suspected acute coronary syndrome patients and in a control reference population. The ability of the Albumin Cobalt Binding (ACB) Test measurement at presentation to predict troponin-positive or -negative results 6-24 h later was also examined. METHODS We enrolled 256 acute coronary syndrome patients at four medical centers. Blood specimens were collected at presentation and then 6-24 h later. The dichotomous decision limit and performance characteristics of the ACB Test for predicting troponin-positive or -negative status 6 h-24 h later were determined using ROC curve analysis. Results for 32 patients could not be used because the time of onset of ischemia appeared to have been >3 h before presentation or was uncertain. The reference interval was determined by parametric analysis to estimate the upper 95th percentile of a reference population (n = 109) of ostensibly healthy individuals. RESULTS Increased cTnI was found in 35 of 224 patients. The ROC curve area for the ACB Test was 0.78 [95% confidence interval (CI), 0.70-0.86]. At the optimum decision point of 75 units/mL, the sensitivity and specificity of the ACB Test were 83% (95% CI, 66-93%) and 69% (95% CI, 62-76%). The negative predictive value was 96% (95% CI, 91-98%), and the positive predictive value was 33% (95% CI, 24-44%). The within-run CV of the ACB Test was 7.3%. Results for the reference population were normally distributed; the one-sided parametric 95th percentile was 80.2 units/mL. CONCLUSIONS This exploratory study suggests that the ACB Test has high negative predictive value and sensitivity in the presentation sample for predicting troponin-negative or -positive results 6-24 h later.