Validation of the i-STAT Handheld Activated Clotting Time for Use With Bivalirudin

ducibility. Estimates of the ICC and exact confidence intervals were derived by a 1-way random effects model. 2,3 The ICC and corresponding 95% confidence inter- val for the i-STAT to Hemochron and i-STAT to i-STAT comparisons in the 123 patients are shown in Table 1. Applying the criteria for acceptable values of ICC as suggested by Landis and Koch, 4 these results indicate " near perfect"(ICC 0.80) i-STAT to Hemo- chron agreement. Moreover, the lower bound for each i-STAT to Hemochron ICC is near or 0.75, further demonstrating the equivalence of these tests. 5 The i-STAT to i-STAT comparison (an evaluation of re- producibility) yielded an even higher ICC. Agreement between both i-STATs and Hemochron as well as i-STAT reproducibility was evaluated graphically. Figures 1 and 2 show scatter plots for the i-STAT measurements compared with the Hemo- chron, along with the y x line as a reference to perfect agreement. In Figures 1 and 2, the points are tightly clustered around the y x line, indicating good agreement. Figure 3 displays the same plot for