UNLABELLED: In a pharmacist-managed anticoagulation clinic, portable point-of-care coagulation devices may facilitate patient monitoring by providing rapid INR measurement. Few studies, however, have validated this type of device. OBJECTIVE: To evaluate the reliability, validity and ease of use of the CoaguChek S, a new portable coagulation device. METHODS: A total of 100 patients followed at a pharmacist-managed anticoagulation clinic attended two study visits. INRs were measured using the CoaguChek S and the standard laboratory technique. RESULTS: Reliability: The test-retest reliability (precision) of the CoaguChek S, estimated by the intraclass correlation coefficient (ICC) and a 95% confidence interval (95% CI), was high (0.98 (0.98-0.99)) and comparable to the standard laboratory technique (0.99 (0.98-0.99)). Interrater reliability was also high (0.97 (0.95-0.98)). Reliability coefficients did not vary with the test-strip lot number nor the CoaguChek S operator. VALIDITY: When compared with standard laboratory procedure, the ICC (95% CI) was equal to 0.93 (0.91-0.95). The mean difference (95% CI) between INR measured by the laboratory and the CoaguChek S was equal to -0.02 units (-0.06-0.03). The mean absolute and relative absolute differences (95% CI) were equal to 0.24 units (0.21-0.27) and 9% (8%-10%), respectively. Differences tended to increase for INRs greater than 3 units as seen by a mean difference (95% CI) of -0.17 units (-0.35-0.02). This represented a mean absolute difference (95% CI) of 0.44 units (0.33-0.55) and a mean relative absolute difference of 12% (9%-15%). Concordance between therapeutic decisions based on CoaguChek S and laboratory results was high (Kappa = 0.68). In 34 cases (18%), the therapeutic decision would have been different. However, in 15 of these discordant observations, the difference between the CoaguCheck S and laboratory INR was <or=0.25 units. Ease of use: In 3% of cases, no INR could be measured by the CoaguChek S. The percentage of extra finger pricks and extra test-strips were equal to 25.8% and 23.7%, respectively. CONCLUSIONS: When used by health professionals in a pharmacist-managed anticoagulation clinic, the CoaguChek S is reliable, valid and easy to use. However, its validity tends to decrease as the INR increases, possibly due to the low sensitivity of the thromboplastin. If the CoaguChek S INR is supratherapeutic, we would therefore recommend confirming the results with a standard laboratory measurement.
UNLABELLED: In a pharmacist-managed anticoagulation clinic, portable point-of-care coagulation devices may facilitate patient monitoring by providing rapid INR measurement. Few studies, however, have validated this type of device. OBJECTIVE: To evaluate the reliability, validity and ease of use of the CoaguChek S, a new portable coagulation device. METHODS: A total of 100 patients followed at a pharmacist-managed anticoagulation clinic attended two study visits. INRs were measured using the CoaguChek S and the standard laboratory technique. RESULTS: Reliability: The test-retest reliability (precision) of the CoaguChek S, estimated by the intraclass correlation coefficient (ICC) and a 95% confidence interval (95% CI), was high (0.98 (0.98-0.99)) and comparable to the standard laboratory technique (0.99 (0.98-0.99)). Interrater reliability was also high (0.97 (0.95-0.98)). Reliability coefficients did not vary with the test-strip lot number nor the CoaguChek S operator. VALIDITY: When compared with standard laboratory procedure, the ICC (95% CI) was equal to 0.93 (0.91-0.95). The mean difference (95% CI) between INR measured by the laboratory and the CoaguChek S was equal to -0.02 units (-0.06-0.03). The mean absolute and relative absolute differences (95% CI) were equal to 0.24 units (0.21-0.27) and 9% (8%-10%), respectively. Differences tended to increase for INRs greater than 3 units as seen by a mean difference (95% CI) of -0.17 units (-0.35-0.02). This represented a mean absolute difference (95% CI) of 0.44 units (0.33-0.55) and a mean relative absolute difference of 12% (9%-15%). Concordance between therapeutic decisions based on CoaguChek S and laboratory results was high (Kappa = 0.68). In 34 cases (18%), the therapeutic decision would have been different. However, in 15 of these discordant observations, the difference between the CoaguCheck S and laboratory INR was <or=0.25 units. Ease of use: In 3% of cases, no INR could be measured by the CoaguChek S. The percentage of extra finger pricks and extra test-strips were equal to 25.8% and 23.7%, respectively. CONCLUSIONS: When used by health professionals in a pharmacist-managed anticoagulation clinic, the CoaguChek S is reliable, valid and easy to use. However, its validity tends to decrease as the INR increases, possibly due to the low sensitivity of the thromboplastin. If the CoaguChek S INR is supratherapeutic, we would therefore recommend confirming the results with a standard laboratory measurement.