The main purpose of our study was to evaluate the reliability of self-management (with all of its theoretical advantages, including ease for weekly testing at home) in comparison with the actual management of anticoagulant therapy in Spanish clinics. In most Spanish clinics, the number of patients each clinic sees makes weekly monitoring almost impossible. For example, in our clinic, we monitor more than 6000 patients each month. Maybe the ideal design of our study would have included the same interval between tests in both arms, but in our opinion the results would not have been applicable to real-life practice.

In contrast to Dr. Wheeler's experience, our results were accurate and consistent using the portable monitor (CoaguChek S, Roche Diagnostics, Mannheim, Germany) in a previous pilot study. We compared the INR results of 150 patients obtained simultaneously with the portable coagulometer and through our laboratory (KC 10 coagulometer, Amelung, Lemgo, Germany). The correlation coefficient was 0.95 (unpublished data). This excellent agreement between the use of CoaguChek devices and standard laboratory coagulometers has been reported elsewhere (1). In our trial, we found the mean percentage of in-range INRs (± SD) to be statistically significantly higher in the self-management group (58.6% ± 14.3%) than in the conventional management group (55.6% ± 19.6%), although we considered this difference irrelevant from a clinical point of view. Of course, the intended INR target ranges were the same width in both comparison groups.

We believe that the most important result from our study relates to the safety of self-management, not the efficacy of testing (which was at least as good in the self-management group as in the conventional group). We observed an impressive reduction (70%) in major complications and in minor hemorrhages in the self-management group, as well as a trend toward reduced mortality. For these reasons, we agree with Beyth (2) that a patient–professional partnership in long-term anticoagulation can reduce the incidence of very serious related complications and that this model of care requires a shift in focus and resources by health care systems and providers.

The new thrombin inhibitor ximelagatran is a very promising drug, with clinical outcomes similar to those seen with conventionally managed warfarin (3). It would be interesting to compare the use of ximelagatran with patient self-management using warfarin in appropriate clinical trials, considering that self-management by trained patients taking warfarin can result in severe complications in 1% or less of patient-years (4).