After a careful reading of the letter by Alfaro-Lara et al, we would like to make a number of comments. First, several studies have analyzed the degree of control of the international normalized ratio (INR) in patients with atrial fibrillation receiving anticoagulation therapy with vitamin K antagonists in Spain. Although each study has its own particular methodology, all have shown similar results, namely, that in clinical practice about 35% to 45% of patients receiving vitamin K antagonists have poor control of the INR.1–4 Crucially, these results are in accordance with those of international studies, including the decisive clinical trials of the direct anticoagulants. Recruitment of the investigators by convenience—rather than randomized—sampling could indeed cause a bias. However, the INR control is probably even worse than that found in these studies, because the investigators selected tend to be the most motivated and their results are thus likely to be significantly better than those of general clinical practice.
In addition, although Alfaro-Lara et al consider that terms like “patients with poor control” should be avoided, we believe that this term properly reflects the high-risk situation of many patients whose INR values are often outside the guideline-recommended therapeutic range.
Finally, the authors note that the results of the CHRONOS-TAO study1 have also been calculated according to an “adjusted range” INR of 1.8 to 3.2, due to the possible margin of error of the coagulometer (±0.2). We believe this statement to be incorrect because it could be misinterpreted. According to the authors, we understood that if a patient has an INR of 3.2, it is specifically because the error has been -0.2, whereas if a patient has an INR of 1.8, the error has been +0.2. However, if the error is ± 0.2, the INR interval to be reached would be 2.2 to 2.8. This approach would ensure an estimated INR control of 2.0 to 3.0, which is the appropriate range, as was first shown more than 10 years ago5 and as included in all clinical practice guidelines.
Conflicts of interestThe PAULA trial was sponsored by Bayer Hispanis S.L., without this sponsorship influencing in any way the performance of the study, its results, or their interpretation.