Correlation of Point-of-Care International Normalized Ratio to Laboratory International Normalized Ratio in Hemodialysis Patients Taking Warfarin

Abstract

To determine whether point-of-care (POC) International Normalized Ratio (INR) test results correlate with plasma INR measures in intermittent hemodialysis (IHD) patients on warfarin. Anemia is thought to reduce the accuracy of POC INR assay results. Whether POC INR testing could be implemented for hemodialysis patients on chronic warfarin, who are often anemic despite hematopoietic therapy, has not been established. Thirty-seven chronic hemodialysis patients on warfarin contributed sets of three consecutive blood samples for INR comparison immediately before hemodialysis: one finger stick, two from hemodialysis access (arteriovenous graft, fistula, or catheter). POC INR testing was performed using CoaguChek S device. Anemia was defined as hematocrit < 32%. Pairwise comparison and correlation of 258 INR results showed high correlation for POC versus laboratory INR (r = 0.94; P < 0.001). Of these, 16 (6%) differed by >0.6 INR units, four (1.6%) differed by >0.8 INR units, and one differed by >1.0 INR units. Resulting pairwise correlation analyses between samples were: for anemic patients (0.96; P < 0.001), nonanemic patients (0.93; P < 0.001), and for those obtained from arteriovenous grafts (0.94; P < 0.001). POC INR samples from dialysis catheters correlated poorly with laboratory INR results. POC INR correlates well with plasma INR measures in IHD patients requiring chronic warfarin, and anemia did not influence this reliability. Blood sampling from finger stick or arteriovenous graft or fistula showed excellent correlation with laboratory INR, whereas sampling from dialysis catheters was unsatisfactory, likely from heparin contamination.