Anticoagulation Therapy by Age and Embolic Risk for Nonvalvular Atrial Fibrillation in Mexico, an Upper-Middle-Income Country: The CARMEN-AF Registry

Background: Documenting the patterns of oral anticoagulation therapy (OAT) is essential to prevent thromboembolic complications of nonvalvular atrial fibrillation (NVAF). Objective: To report the patterns of OAT according to age and thromboembolic risk in patients included in CARMEN-AF, a nationwide registry of NVAF in Mexico, an upper middle-income country. Material and methods: There were 1,423 consecutive patients ≥18 years old and with at least one thromboembolic risk factor enrolled in the CARMEN-AF Registry at their regular clinical visit during a three-year period. They were analyzed according to 1) age, 2) AF type, and 3) CHA2DS2-VASc score. Results: Overall, 16.4% of patients did not receive antithrombotic treatment, 19.4% received antiplatelet drugs (APD), 29.2% vitamin K antagonists (VKA), and 34.6% direct oral anticoagulants (DOAC). With increasing age, the proportion of subjects treated with VKA decreased significantly from 36.2% in subjects <65 years to 22.5% in those ≥75 years old (P <0.0001). Concomitantly, an increase in both APD and no antithrombotic treatment was observed with increasing age. DOAC were prescribed equally among all age groups (34.2% in <65, 36.0% in 65–74, and 33.9% in ≥75). According to the type of AF, VKA use was more common in patients with permanent AF (32.7%). A lower use of DOAC was observed in high thromboembolic risk subjects (33.6% in CHA2DS2-VASc ≥2) compared with the moderate risk group (41% in CHA2DS2-VASc = 1). Conclusions: VKA use for NVAF in Mexico decreased in relation to increasing age. The proportion of DOAC therapy was the same in all age groups. Nevertheless, elderly patients with high thromboembolic risk received a suboptimal thromboprophylaxis. These data could help to improve gaps in the implementation of global guidelines. Clinical trial registration: http://www.clinicaltrials.gov. Unique identifier: NCT02334852. Highlights: CARMEN-AF is a nationwide multi-centric registry seeking to bridge the data gap on anticoagulation therapy for NVAF in Mexico. Elderly patients are more prone to receive suboptimal OAT for NVAF. DOAC were less frequently used in high thromboembolic risk patients (CHA2DS2-VASc ≥2).


Introduction
Atrial fibrillation (AF) is the most common sustained arrhythmia encountered in clinical practice. 1,2 It affects about 1-2% of the general population 3 , its prevalence increases with age, and it has multiple causes, among which are: heart failure, diabetes mellitus type 2, and hypertension 4 . AF is an important risk factor of stroke, as it is responsible for 25% of ischemic stroke, and 50% of cardioembolic stroke. Patients with AF have a cerebral ischemia incidence of 7% anually 5 . Besides, it is demonstrated that AF increases in-hospital mortality 6 , increases the number of hospitalizations, and decreases quality of life 7 . Thus, AF represents a public health problem with an associated high economic cost 8,9 . One of the main objectives of AF treatment is the prevention of thromboembolism with oral anticoagulants (OAC). The AF therapeutic guidelines 5,10 provide recommendations about the suitability of a patient for OAC treatment according to different stroke risk stratification scores (CHADS 2 y CHA 2 DS 2 -VASc) 11 and bleeding risk (HAS-BLED) 12 . The Mexican Clinical Practice Guideline follows these same recommendations but does not emphasize the benefits of OAC in the more long-lived patients or with a high risk, which are usually treated with antiplatelet drugs 13 .
Both direct OAC (DOAC), and vitamin K antagonists (VKA), have shown to be effective in preventing stroke in patients with AF [14][15] . However, adherence to vitamin K antagonists is highly variable, with a dropout rate in the first year of treatment that can reach up to 25%, and multiple drug interactions. Non-compliance with medication is associated with a poor perception by the patient about the need of taking medications or fear of adverse effects 16 . To improve the adherence more information is needed, so the patients be aware of the risk of their illness, consider it seriously, and get involved in essential prevention and treatment behaviors.
CARMEN-AF registry aims primarily to determine the current status of thromboprophylaxis of non-valvular AF in Mexico in order to identify the characteristics of the current use of OAC in Mexico. A secondary objective is to know the morbidity and mortality associated with this arrhythmia and the efficacy and safety in thromboprophylaxis of atrial fibrillation with the use of new and specific oral anticoagulants, in at least one year of follow-up.

Methods
The CARMEN-AF registry is an observational, longitudinal, multicenter, and nationwide survey about non-valvular AF. The CARMEN-AF registry is proposed as the first registry of the treatment of thromboembolic risk in non-valvular AF in Mexico that will have national representation. The participation of major health institutions, both public and private, of each state to ensure that the registry is representative of the country is planned. It seeks for wide geographical distribution to compare the quality of care between the different geographical areas of the country.
The follow-up period of each patient will be for a minimum of two years with periodic evaluations every three months. Data will be registered in an electronic case report form (Annex 2) stored in a centralized database in which all the necessary data for the study will be included. The study variables will be (1) demographic, (2)

Inclusion and exclusion criteria
Candidates for the registry are those patients of any gender, older than 18 years who have documented by a 12-lead ECG, rhythm strip, Holter ECG monitoring or atrial electrograms of a pacemaker, at least one episode of AF in the last 6 months. All patients diagnosed with AF at the baseline visit must be eligible to receive antithrombotic prophylaxis with OAC and must have at least one risk factor for thromboembolism according to the CHA2DS2-VASc score. Female gender must be accompanied by some other risk factor, otherwise, it will not be coded for the risk score measured by CHA2DS2-VASc. Patients with AF of transient causes, AF onset in immediate postoperative or 3 months in cardiac surgery, terminal illness, mental inability to take anticoagulants, inability to fulfill the follow-up visits, already programmed for ablation of pulmonary veins, pregnant or lactating women, will be not eligible for inclusion in this registry (Table 1).

Data analysis
Descriptive and differential statistics will be used according to the observation obtained. For continuous variables, mean, standard deviation (SD), standard error of the mean (SEM), 95% confidence interval (CI 95%) minimum, percentile 25 (P25), median or percentile 50 (P50), percentile 75 (P75) and maximum, number of patients (n) and number of lost data will be obtained as descriptive statistics. For categorical variables, % will be obtained from the total and number of patients (n) in each category. Lost data will be treated as a new category.
Differences between continuous variables with normal distribution will be examined by Student's t-test. The Wilcoxon Sum-Rank Test will be used when continuous variables have failed in normality tests. To analyze categorical variables, chi-square will be used by the Fisher exact test or Yates correction. A two-tailed test with a value of p <0.05 will be considered statistically significant. Logistic regression analysis will be used to select independent predictors in those variables in which a value of p <0.01 has been obtained through univariate regression analysis.
Kaplan-Meier survival curves and the Cox proportional hazard model will be used to adjust the survival analysis. A p <0.05 will be considered statistically significant.

Results
The recruitment phase began on September 19, 2014 and will end on September 18, 2016. The participation of a total of 1,200 patients is estimated. A first preliminary analysis will be carried out with the first 1,000 patients. In September 2016, the cross-sectional analysis of all included patients will be performed (1,200 patients are calculated), and then one and two-year follow-up analysis will be performed. Currently, there are 51 participating centers confirmed by the time being to include patients in the registry (Table 2).

Discussion
As in the rest of the world, in Latin America, the increase of life expectancy of population increases the incidence of ischemic stroke, and it is feared that its mortality will increase in the coming years 17,18 . Thus, the Canadian guidelines advise the use of the CHADS 2 score to assess the risk of stroke, and the HAS-BLED score to assess the risk of bleeding, and the use of DOAC (dabigatran, rivaroxaban or apixaban) against VKA 23 . The USA guideline recommends the use of antithrombotic agents as a rule, except in contraindicated cases 24 . Finally, the European guidelines recommend the use of the CHA 2 DS 2 -VASc and the HAS-BLED scores to measure the risks of stroke and bleeding, respectively, and the use of DOAC against aspirin and VKA 10 .
All these recommendations are based on the results obtained in phase III studies such as ARISTOTLE, RE-LY or ROCKET AF [25][26][27][28] , which have shown advantages in the safety, efficacy, and management of DOAC compared to VKA 29 . The recommendations of the Mexican clinical practice guideline also favor treatment with oral anticoagulants 5 .
Despite the existing recommendations and the evidence in favor of treatments with OAC, it is estimated that only between 15% and 79% of patients receive this treatment properly 30  Adherence to VKA is very variable. The ATRIA study documented a 25% dropout in the first year in patients treated with VKA 32 , while the IN-RANGE study documented that 92% of patients missed a dose and 36% missed more than 20% of treatment 33 . This lack of adherence is one of the main causes of embolic events. On the other hand, there is no information on DOAC adherence, although it is expected to be greater than with VKA because of their fixed dosage, ease of administration, unnecessary monitoring, greater efficacy, and safety. To obtain a good follow-up in medication it is necessary to reduce the asymmetry of information, but it is also important to follow the recommendations of the international clinical practice guidelines.

Conclusion
The CARMEN-AF study will allow obtaining information about the current status of thromboprophylaxis in patients with non-valvular AF in Mexico. Anticoagulant treatment strategies in a total of 1,200 patients will be analyzed and their effectiveness will be evaluated.
In addition, the degree of compliance with the clinical practice guidelines may be assessed, which will allow the development of an educational strategy to improve adherence to this type of treatment.

Ethical disclosures
Protection of human and animal subjects. The authors state that for this investigation no experiments have been performed on humans or animals.
Confidentiality of data. The authors state that they have followed the protocols of their work center on the publication of patient data.
Right to privacy and informed consent. The authors state that in this article there are no patient data.

Funding
Bayer, Boehringer-Ingelheim, and Pfizer unrestrictedly support the CARMEN-AF Registry, only for academic purposes.

Conflicts of interest
The Registry has unrestrictedly support from the Pharmaceutical Industry. They did not intervene in the development or the design of the CARMEN-AF Registry. Data collection will be performed by a specialized agency, independent of the sponsoring industry. Data analysis will be the responsibility of the Scientific Committee members who will be also in charge of coordinating the publications that may be generated from these analyses.
None of the authors involved in this research have a conflict of interest of any kind with Bayer, Boehringer-Ingelheim, Pfizer, Medicaweb or with any of the participating medical societies. -Patient with diagnosis of chronic AF (permanent) documented by at least 2 electrocardiogram tracings taken on different days in the last 6 months. Tracings can be by 12-lead electrocardiogram, rhythm strip, or 24-hour Holter monitoring.
-Patient with diagnosis of paroxysmal/persistent AF documented for at least 30 seconds by 12-lead electrocardiogram, rhythm strip, 24-hour Holter monitoring, or atrial electrograms from a pacemaker. It is required that at least one episode has occurred and documented in the last 6 months.
-Patient with at least one risk factor for thromboembolism evaluated by CHA 2 DS 2 -VASC score.
-AF onset in immediate postoperative or 3 months in cardiac surgery.
-Mental inability to take anticoagulants.
-Inability to fulfill the follow-up visits.
-Patient already programmed for ablation of pulmonary veins.
-Pregnant or lactating women.