OBJECTIVES: To design and validate a simple, reliable and easy to use questionnaire to identify patients under vitamin K antagonists (VKA) treatment with sub-optimal time in therapeutic range (TTR; without needing the TTR values), in order to allow a closer follow-up of the actual anticoagulant treatment or guiding a possible treatment changes.

METHODS: Non-interventional, cross-sectional, multicenter design. A cardiologist expert panel identified those clinical variables capable of forecasting a good TTR control. The 5 proposed indicators (ICUSI) were included in an electronic clinical record form along with thromboembolic (CHADS & CHADS-VASc) and hemorrhagic (HAS-BLED) risk indexes. A patient diary collecting last 6 month INR measures was also included. Discriminant capabilities were assessed using binary logistic regression respect to TTR control (Rosendaal & direct methods), and discriminant validity using the area under the curve (AUC) of the ROC curve. A total of 870 patients with documented AF diagnosis, receiving VKA, and meeting the selection criteria were included. Patients were recruited at 54 Cardiology departments covering all the Spanish territory.

RESULTS: The final sample contained 777 (88.9%) of assessable cases: 45% women, 13% lived alone, and 67% did not have caregiver. Mean age was 75.3±9.2 years, mean TTR=66.0±19.8% and 46.7% attained INR control between 2-3 (Rosendaal TTR≥65%). Individually, ICUSI indicators were statistically significant (p<0.05) respect to Rosendaal or Direct non-optimal control, but one item was discarded in the overall estimation. The ICUSI index was statistically discriminant with respect to 6 months TTR<65% with AUC=0.705 (SE=0.018). Using the cut-off value ICUSI≥1 the risk of TTR<65%: sensitivity=80.9%, specificity=44.9%, positive predictive value=62.6%, and negative predictive value=67.4%.

CONCLUSIONS: Initial evidences on the ICUSI validity support a good predictive behavior on screening patients with bad anticoagulant control.