Patient Self-Testing of Prothrombin Time After Hip Arthroplasty

C. Anderson Engh, Jr., MD, William J. Culpepper II, MA, Patricia A. Charette, RN and Rachel Brown, from the Anderson Orthopaedic Research Institute, Alexandria, Va.

J South Orthop Assoc. 2001;10(3) 

In This Article


Of the 46 patients who agreed to participate in the study, 7 (15%) could not be trained because they were apprehensive about sticking their own finger or were unable to accurately place a drop of blood in the device. Of the 39 patients who completed the training, 9 (23%) were withdrawn before completion of the 6-week period for reasons unrelated to the device -- 5 (13%) because their Coumadin therapy needed to be stopped and 4 (10%) because they were discharged out of the service area. Thirty patients were discharged from the hospital with the device. One patient was unable to obtain an unsupervised result after training. Thus, 29 patients completed the 6 weeks of Coumadin therapy and the associated testing protocol.

Table 1 compares the reliability of the tests done by the nurse with the self-testing device, the patient self-tests, and the laboratory tests. Results from patient self-tests were significantly and positively correlated with the nurse's CoaguChek test results (Pearson correlation 0.86, P < .01) with no difference in mean INR values between these two methods (P = .45). Similarly, the patient self-test results were significantly and positively correlated with the venipuncture laboratory value (Pearson correlation 0.82, P < .01). The mean laboratory result was slightly higher than the mean patient self-testing result (INR 1.88 ± 0.060 versus 1.67 ± 0.46, P < .05).

In comparing test results between the study group (patient self-testing twice a week) and the control group (laboratory testing once a week), we found that the self-testing group stayed within the target therapeutic range (INR of 1.5 to 2.5) significantly more often than the control group. As shown in Table 2, 59.4% of the test results in the self-testing group were within range, versus 39.1% of the control group's results (Fisher's exact test, P < .01). Further analysis showed the self-testing group had significantly fewer tests that fell below an INR value of 1.5 than the control group (33.3% versus 49.4%; chi-square, P < .01).

We have reported the DVTs and bleeding complications only for completeness, since it was not the study's purpose to compare DVT rates. Two patients in the study group had a DVT (one distal and one proximal) after operation. One patient in the study group and one patient in the control group had minor bleeding episodes.

Because some patients needed to repeat finger sticks to obtain a result, we evaluated the effect of the number of finger sticks required early in the study to the number required 6 weeks postoperatively. (It had been necessary for some patients to repeat finger sticks, because the CoaguChek gives no value if an inadequate amount of blood is obtained or if the blood is not accurately placed on it.) We divided the study-group patients into three groups according to the number of attempts required for a valid result during the first week of self-testing. Group 1 consisted of patients who obtained a valid result on the first attempt; Group 2 patients required 2 to 3 attempts; and Group 3 patients required more than 3 attempts. No difference was found among the 3 groups in the average number of attempts after the second week of home testing. By the fifth test (midway through the study), the patients were consistently obtaining a valid result on the first or second attempt. This improvement indicated that there was a two- to five-test learning curve for these patients.

To assess whether the number of finger sticks needed to obtain a value influenced reliability, we compared among the three groups the absolute difference between the patient's selt-test result and the laboratory result. No difference was found.

Costs associated with each prothrombin time monitoring protocol are summarized in Table 3. In factoring overall costs for the CoaguChek protocol, we included two early home health visits during which the nurse assessed the patient's home for assistive devices and checked the surgical incision. These visits also could be used to ensure that the patient is doing the tests properly and is consistently obtaining a valid result on the first or second attempt before relying on only the patients' CoaguChek results for dosage adjustments. Comparing the self-testing protocol (12 prothrombin time tests) to the standard monitoring protocol (6 prothrombin time tests), the self-testing protocol had an average saving of $260 per patient. If 12 prothrombin time tests were done, according to the traditional prothrombin time monitoring protocol, the cost saving would total $896 per patient.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: