ABC | Volume 114, Nº3, March 2020

Original Article Anticoagulation Therapy in Patients with Non-valvular Atrial Fibrillation in a Private Setting in Brazil: A Real‑World Study Pedro Gabriel Melo de Barros e Silva, 1, 2 Henry Sznejder, 2 Rafael Vasconcellos, 3 Georgette M. Charles, 4 Hugo Tannus F. Mendonca-Filho, 2 Jack Mardekian, 5 Rodrigo Nascimento, 6 Stephen Dukacz, 5 Manuela Di Fusco 5 Hospital Samaritano Paulista, 1 São Paulo, SP – Brazil United Health Group-Brasil, 2 Rio de Janeiro, RJ – Brazil United Health Group, 3 Minnetonka, Minnesota – USA Optum Life Sciences, 4 Eden Prairie, Minnesota – USA Pfizer Inc., 5 New York – USA Pfizer Inc., 6 São Paulo, SP – Brazil Mailing Address: Pedro Gabriel Melo de Barros e Silva • Hospital Samaritano Paulista - R. Dr. Fausto Ferraz, 204-232. Postal Code 01333-030, Bela Vista, SP – Brazil E-mail: pgabriel@prestadores.samaritanopaulista.com.br Manuscript received July 29, 2018, revised manuscript April 04, 2019, accepted May 15, 2019 DOI: https://doi.org/10.36660/abc.20180076 Abstract Background: The safety and effectiveness of warfarin depend on anticoagulation control quality. Observational studies associate poor control with increased morbidity, mortality and healthcare costs. Objectives: To develop a profile of non-valvular atrial fibrillation (NVAF) patients treated with warfarin in a Brazilian private ambulatory and hospital setting, evaluate the quality of anticoagulation control, and its association with clinical and economic outcomes. Methods: This retrospective study, through a private health insurance dataset in Brazil, identified NVAF patients treated with warfarin between 01 MAY 2014 to 30 APRIL 2016, described their anticoagulation management, and quantified disease-related costs. Data on demographics, clinical history, concomitant medication and time in therapeutic range (TTR) of international normalized ratio (INR) values were retrieved. Patients were grouped into TTR quartiles, with good control defined as TTR ≥ 65% (Rosendaal method). Major bleeds and all-cause direct medical costs were calculated and compared between good and poor control subgroups. P-values < 0.05 were considered statistically significant. Results: The analysis included 1220 patients (median follow-up: 1.5 years; IQR: 0.5–2.0). On average, each patient received 0.95 monthly INR measurements (mean INR: 2.60 ± 0.88, with 26.1% of values < 2 and 24.8% > 3), (median TTR: 58%; IQR: 47–68%), (mean TTR: 56.6% ± 18.9%). Only 31% of patients were well-controlled (mean TTR: 78% ± 10%), with 1.6% having major bleeds within median follow‑up, and direct medical costs per member per year (PMPY) of R$25,352(± R$ 37,762). Poorly controlled patients (69%) were associated with 3.3 times more major bleeds (5.3% vs. 1.6%; p < 0.01) and 40% higher costs (R$35,384 vs. R$25,352; p < 0.01). Conclusions: More than 60% of the patients were below the desired target and the associated costs were higher. (Arq Bras Cardiol. 2020; 114(3):457-466) Keywords: Warfarine/therapeutic use; Anticoagulants/adverse effects; Atrial Fibrillation/comlications; Hospitals, Private/ economics; Health Care Quality, Access and Evaluation. Introduction Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia that affects more than 33 million people worldwide. Most cases are non-valvular AF (NVAF) patients. 1-3 Epidemiology data for AF in Latin America is limited and a significant proportion of patients has poor control of key risk factors and does not receive appropriate anticoagulation treatment (18.3% – 24.6%). 4,5 Clinical guidelines recommend the use of an oral anticoagulant (OAC) in NVAF to reduce the risk of stroke. 2,3 For decades, vitamin K antagonists (VKAs), the most commonly used of which is warfarin, have been the cornerstone of OAC therapy for NVAF. However, the safety and efficacy of warfarin have limitations and depend on the tight quality of anticoagulation control. 2 This is achieved using a standardized measure of clotting time known as the international normalized ratio (INR), which is desired to be between 2 and 3. 6 Frequent INR monitoring and dose adjustment are needed to maintain target INR levels. 2,3 However, monitoring can increase the medical and economic burden. 7 Time in therapeutic range (TTR) is the standard means of assessing the long-term quality of anticoagulation control and the risk–benefit profile of warfarin. 6 TTR represents the proportion of time that a patient’s INR values are between 2 and 3, having the maximum benefit when the TTR is 60% to 70% or higher. 2 In Latin America, the median TTR was at the lower 457