Determinants of Time in Therapeutic Range in Patients Receiving Oral Anticoagulants (A Substudy of IMPACT)

Abstract

Implanted cardiac arrhythmia devices can detect atrial tachyarrhythmias (atrial high-rate episodes [AHREs]) that are considered to correlate with atrial fibrillation and risk of stroke. In the IMPACT trial, oral anticoagulation was initiated when AHREs were detected by implanted cardioverter-defibrillators and withdrawn when they abated, according to a protocol accounting both for AHRE duration as detected by remote device monitoring and stroke risk assessment. In this analysis, we ascertained determinants of time in therapeutic range (TTR) among protocol-determined vitamin K antagonist-treated patients during the trial. We enrolled 2,718 patients with at least 1 additional stroke risk factor (CHADS2 score ≥1) at 104 arrhythmia centers. The sex, age <60, medical history, treatments interacting with VKA, tobacco use (2 points) and race (2 points for non-Caucasian) (SAMe-TT2R2) score is a simple clinical-derived score designed to aid decision-making on whether a patient is likely to achieve good anticoagulation control on vitamin K antagonist (e.g., warfarin), which was calculated and related to TTR achieved using the Rosendaal method. We analyzed 229 patients (mean age 66.7years; mean CHADS2 score 2.85 [SD 1.1]) with mean TTR of 0.536 (SD 0.23) overall. Univariate analysis identified 5 variables associated with differences in mean TTR. Mean TTR was lower in those who were women (p= 0.031), of black race (p= 0.005) and in New York Heart Association class IV (p= 0.014), whereas hemoglobin >13.5g/dl (p= 0.010) and New York Heart Association class I (p= 0.037) were associated with higher mean TTR. There was a significant difference in mean TTR value between US and non-US sites (Canada and Germany) (mean TTR for US: 0.513 vs non-US: 0.686; p <0.0001). Mean TTR was significantly lower (Δ= 0.1382, 95% CI 0.0382 to 0.2382) for patients with SAMe-TT2R2 scores of 4 (p= 0.007) and higher (Δ= 0.0612, 95% CI 0.0005 to 0.1219) for patients with SAMe-TT2R2 scores of 1 (p= 0.048). Linear regression confirmed a significant association between lower SAMe-TT2R2 score and improved anticoagulation control (p= 0.0021) with a 1-unit decrease in SAMe-TT2R2 score associated with an increase in TTR of 0.0404 (95% CI 0.0149 to 0.0659). In conclusion, clinical, geographical, and demographic factors were associated with the quality of anticoagulation control as reflected by TTR. Although overall TTR in this population was poor, lower SAMe-TT2R2 scores were associated with better TTR.