RESUMO
OBJECTIVES: Point-of-care (POC) anticoagulation testing devices can potentially improve warfarin therapeutics in long-term care; however, there is variable accuracy reported for these devices and scant data when used in older adults. Accordingly, we undertook this study to determine the accuracy of a POC device (Hemosense INratio) in long-term care settings and examine factors associated with discrepant results. DESIGN: Case series. SETTING: Two, demographically comparable continuing-care retirement communities in the Southeastern United States. PARTICIPANTS: Long-term (nursing home and assisted living) residents with atrial fibrillation, venous thromboembolism, or prior cerebrovascular accident (16 at site 1; 8 at site 2). MEASUREMENTS: INR results calculated by the Hemosense device were compared with those determined by venipuncture-collected samples run in commercial laboratories. Patient demographic and clinical data were collected, as was performance by tester. RESULTS: Correlation varied between sites and, at site 1, between testers. Accuracy at site 1 was comparable to published reports for 2 of the 3 testers, with rather disconcerting discrepancy rates of 17.8% and 23.1%. However, correlation for the third tester was much better, with only a 7% discrepancy rate based on clinician rating of Hemosense-Reference lab differences. Correlation at site 2 was considerably worse than site 1, to the point that the Hemosense could not be safely adopted. CONCLUSION: POC devices may not be appropriate for commercial laboratory tests substitution without prior performance evaluation. Running POC INRs concurrent with laboratory-determined INRS can determine test reliability. Timing of Hemosense testing in relation to when laboratory INRs were drawn is one likely explanation for our results, although user differences may also contribute significantly. Further research in larger, more diverse populations, using a variety of POC devices, and with direct comparison of older and younger patients is needed.