BACKGROUND: Patients receiving oral anticoagulation should be tested often enough to optimize control, but excessive testing increases burden and cost. We examined the relationship between the follow-up interval after an in-range (2.0-3.0) INR and anticoagulation control.

METHODS: We studied 104,451 patients receiving anticoagulation from 100 anticoagulation clinics in the Veterans Health Administration (VA). Most patients (98,877) recorded at least one in-range INR followed by another INR within 56 days. For each such patient, we selected the last in-range INR and characterized the interval between this "index value" and the next INR. The independent variable was site mean follow-up interval after an in-range INR. The dependent variable was site mean risk-adjusted percent time in therapeutic range (TTR).

RESULTS: Site mean interval varied from 25-38 days. As the mean site follow-up interval became longer, risk-adjusted TTR was worse (-0.51% per day, p = 0.004). This relationship persisted when the index value was the first consecutive in-range INR (-0.63%, p < 0.001), or the second (-0.58%, p < 0.001), but not the third or greater (-0.12%, p = 0.46).

CONCLUSIONS: Sites varied widely regarding follow-up intervals after in-range INR (25-38 days). Shorter intervals were generally associated with better anticoagulation control, but after a third consecutive in-range value, this relationship was greatly attenuated and no longer statistically significant. Our results suggest that a maximum interval of 28 days after the first or second in-range value and consideration of a longer interval after the third or greater consecutive in-range value may be appropriate.