Determinants and measures of quality in oral anticoagulation therapy.

BACKGROUND: Anticoagulation management services or clinics have been recommended as the preferred method in the long-term management of oral anticoagulation with vitamin K antagonists and have been shown to increase the time patients spend in the therapeutic range. This surrogate marker of the quality of anticoagulation control is a well accepted predictor of bleeding and thromboembolic events and is generally used as a quality measure. However, the method of calculating the time in the therapeutic range can give different results and there is no consensus on the methodology that should be used or the benchmark targets that should be aimed for. Additionally, the expected rates of bleeding and thromboembolic complications are dependent on the indications for anticoagulation in the patient population being evaluated. These issues need to be taken into account when setting quality standards for anticoagulation clinics.
METHODS: An informal survey and group discussion with anticoagulation clinic personnel attending a workshop at the 9th National Conference on Anticoagulant Therapy was used to generate a list of pragmatic barriers to measuring these quality indicators and to share ideas on other quality markers. A narrative review of selected literature was used throughout the workshop to exemplify potential benchmark rates for therapeutic time in range, bleeding, and thromboembolic complication rates.
RESULTS: Approximately 65% of the workshop attendees measure time in range in their anticoagulation clinics, however, only 15% used the linear interpolation method which has a quality measurement target of 65%. Less than half of the attendees measure bleeding or complication rates and very few adjust these rates based on the indication for anticoagulation. There was strong agreement regarding pragmatic barriers to collect this information and difficulties in extrapolating standards from the literature. Several clinics also measure the percent of extremely high International Normalized Ratios (INR) and also track late patients.
CONCLUSIONS: Using clinical trial bleeding and thromboembolic complication rates to set quality measurement targets for anticoagulation clinics may not be appropriate, given the inherent difference in these patient populations. Additionally, there are pragmatic issues affecting the completeness and accuracy of adverse event gathering outside of a trial scenario that could be misleading. The time in the therapeutic range, however, is relatively easy to calculate and is a well substantiated surrogate marker for complication rates and should be a standard quality indicator. Benchmark targets for time in range are dependent on the methodology used in the calculation and should be adjusted accordingly.