Skip to main content Back to Top


Dutch Study Supports INR Lowering

Kate Traynor

A recent report from clinicians in the Netherlands supports the country's 1996 decision to lower the target international normalized ratio (INR) for the prevention of arterial thrombosis in certain patients.

According to the March 22 Archives of Internal Medicine report, lowering the target INR range from 3.6–4.8 to 3.0–4.0 resulted in a 30 percent overall decrease in the occurrence of major bleeding events in 2,341 patients with atrial fibrillation, a mechanical heart valve, or an ischemic stroke who were taking a coumarin drug for anticoagulation.

Most of the decrease in major bleeding was attributed to a decrease in intracranial bleeding in patients undergoing anticoagulation because of atrial fibrillation or a previous ischemic stroke, according to the report. In all, 27 of 2,341 patients suffered intracranial bleeding during the higher-INR period, compared with 11 of 2,260 after the target range was lowered—a 46 percent overall decrease in prevalence.

The report also states that lowering the INR range cut in half the risk of thromboembolism, a finding that the authors described as "surprising."

An accompanying editorial by clinicians Martin O'Donnell and Jack Hirsch of the Henderson Research Centre in Hamilton, Ontario, disputes the reliability of the data that were used to calculate the risk of thromboembolism. The problem, according to O'Donnell and Hirsch, is that many patients who were in the higher-INR group continued treatment in 1996 as members of the lower-INR group.

The editorialists stated that patients who had survived treatment at the higher INR "would be expected to be at lower risk of bleeding and thrombosis than a newly assembled cohort." Thus, the O'Donnell and Hirsch argued, the higher- and lower-INR groups were not really comparable at baseline, and the primary analysis was faulty.

"From a research standpoint, it would have been cleaner had they not mixed the patients in both groups," said pharmacist Maumi Villarreal, chief of patient services at the Texas Tech University Health Sciences Center.

"But when you're in the midst of a clinic," she noted, "you want to grab as many patients and as many lab values and as many numbers as you can get" to try to validate your procedures.

Villarreal said she was struck by the use of INR target ranges in the Netherlands that, even with the 1996 change, are higher than those described in the American College of Chest Physicians (ACCP) guidelines that are widely used in the United States.

ACCP recommends an INR range of 2.0–3.0 for most patients receiving antithrombotic treatment, including those with atrial fibrillation and patients who have recently suffered a stroke.

Villarreal said that "there is still some debate as to what the INR ranges should be" for patients with mechanical heart valves—a view shared by O'Donnell and Hirsch in their editorial.

She said that universal agreement on an optimal INR range for these patients would be particularly useful at her clinic, which is on the Texas–Mexico border and treats many young patients whose heart valves were apparently replaced after complications from inadequately treated rheumatic fever.

Students who come to the clinic, Villarreal said, "are always amazed that we see so many people here with valves [who]...are 30, 40, as opposed to 60 and 70 years old."

She pointed out that a young person whose heart valve is replaced can expect to face decades of anticoagulation therapy and all of its attendant difficulties.

Villareal said she thought the Archives article contained "good, good information" despite the criticism about the study design.

Overall, she said the study's finding of improved outcomes after lowering INR targets "sort of validates what we're doing in this country" with regards to anticoagulation therapy.

"Their outcomes told us that what we're doing in regards to our [ACCP] guidelines are probably right on target," she added.