Epidemiological characteristics, management, and outcomes of atrial fibrillation in TUNISIA: Results from the National Tunisian Registry of Atrial Fibrillation (NATURE‐AF)

Abstract Background Contemporary registries on atrial fibrillation (AF) are scare in North African countries. Hypothesis In the context of the epidemiological transition, prevalence of valvular AF in Tunisia has decreased and the quality of management is still suboptimal. Methods NATURE‐AF is a prospective Tunisian registry, involving consecutive patients with AF from March 1, 2017 to May 31, 2017, with a one‐year follow‐up period. All the patients with an Electrocardiogram‐documented AF, confirmed in the year prior to enrolment were eligible. The epidemiological characteristics and outcomes were described. Results A total of 915 patients were included in this study, with a mean age of 64.3 ± 22 years and a male/female sex ratio of 0.93. Valvular AF was identified in 22.4% of the patients. The mean CHA2DS2VASC score in nonvalvular AF was 2.4 ± 1.6. Monotherapy with antiplatelet agents was prescribed for 13.8% of the patients. However, 21.7% of the subjects did not receive any antithrombotic agent. Oral anticoagulants were prescribed for half of the patients with a low embolic risk score. In 341 patients, the mean time in therapeutic range was 48.87 ± 28.69%. Amiodarone was the most common antiarrhythmic agent used (52.6%). During a 12‐month follow‐up period, 15 patients (1.64%) had thromboembolism, 53 patients (5.8%) had major hemorrhage, and 52 patients (5.7%) died. Conclusions NATURE‐AF has provided systematic collection of contemporary data regarding the epidemiological and clinical characteristics as well as the management of AF by cardiologists in Tunisia. Valvular AF is still prevalent and the quality of anticoagulation was suboptimal.


| INTRODUCTION
In the last decades, a significant change in the epidemiologic and etiologic patterns of cardiovascular diseases has been noted in North African countries, with a decrease in rheumatic heart disease and an increase in hypertensive and ischemic heart diseases. 1,2 A decrease in the incidence of acute rheumatic fever and rheumatic heart disease has also been observed in the last four decades in African countries. 3,4 It is also estimated that by 2050, prevalence of atrial fibrillation (AF) in Africa will be greater than in any other region in the world. 1 As for all heart diseases, there are insufficient contemporary population-based data, describing the epidemiological and management pattern of AF patients receiving routine medical care in North Africa, especially in Tunisia where rheumatic valvular disease was the most underlying etiology of AF in 2003. 5 Demographic and prognostic AF data from other ethnic groups, such as European, Asian, and American countries would not be extrapolated to our population. It is unknown whether occidental studies could be easily applied to low-to-middle income regions, such as North African populations, where reported data are still scarce.
Rheumatic heart disease is present in more than fifth of African AF patients 6-8 compared with 2% in North American AF patients. 9 Anticoagulation prescription rates are low in African AF patients and they have decreased progressively over time. Only 33% of the patients with valvular AF and 12% of those with nonvalvular AF are on anticoagulants at the six-month follow-up. 8 Thus, a register or a survey dealing with the demographic and prognostic characteristics of AF in Tunisia is essential to be able to identify its specific characteristics inherent partly to the ethnic particularities, but especially to the particularities of the local health system, in the context of the epidemiological and guidelines transitions. Hence, the National Tunisian Registry of Atrial Fibrillation (NATURE-AF) was performed as previously described. 10 The aim of the present registry was to describe the epidemiological characteristics, the quality of management and the outcomes over a 12-month follow up period.

| METHODS
NATURE-AF is a prospective, observational registry with a 1-year follow-up period. It included consecutive in-and outpatients with AF presenting to cardiologists between March 1, 2017 A multivariate analysis was performed with anticoagulant treatment (over or undertreated) as dependent factor. The independent variables were age, gender, body mass index, type of AF, and combined therapy. Univariate logistic regression was carried out with a 10% output threshold. The final model was performed using the parameters selected by the backward stepwise method of Wald. The selected variables in the final model were tested at the 5% threshold.
Interaction between the selected parameters was tested at the 10% threshold.
The TTR was calculated as described by Rosendaal et al, 11     None of the cardiovascular risk factors (hypertension, diabetes mellitus, dyslipidemia, and smoking) was significantly associated with TTR ( Table 2). For patients with nonvalvular AF, none of CHA 2 DS 2 VASc score, HAS-BLED score, or SAMe-TT2R2 score was significantly associated with the quality of anticoagulation (TTR≥65%).
After multivariate adjustment (

| Rate and rhythm control
Rate control was attempted in 48.4% of AF patients with the use of beta-blockers, digoxin, and calcium blockers in respectively 59.9%, 29.9%, and 20.3% of the patients.
Thus, a decrease in the prevalence of valvular atrial fibrillation and an increase in the prevalence of nonvalvular atrial fibrillation with a high prevalence of hypertension, congestive heart failure, and diabetes were noted in NATURE-AF. This is in accordance with the epidemiological transition seen in North Africa.

| Rate versus rhythm management
Of the two main strategies for the treatment of AF, the 'rate control' and 'rhythm control' were similarly chosen in our registry. In REALIZE-AF, 25 the rhythm control was the most commonly chosen strategy (57.5% vs. 37.2%).
Regarding the rate control drugs, beta-blockers and nondihydropyridine calcium channel blockers were more often used than digoxin in NATURE-AF. Amiodarone was the most common antiarrhythmic drug used (36%). The low rate of catheter AF ablation is mainly due to economic reasons. In fact, in North African countries, costs of catheter AF ablation are borne by the patient rather than a health insurance.
In the ReLY-AF registry, 8

| Outcomes
AF remains a major cause of morbidity and mortality. At 1 year, cardiovascular mortality was estimated to be 5.7% and thromboembolic complications to be 1.6% in this registry. These results were similar to EORP-AF pilot registry 26 and EHS, 29 where mortality was estimated to be 5.7% and 5.3%, respectively. The thromboembolic complications rate was similar to that revealed in EHS (1.8%) but higher than that shown in EOARP-AF stroke complications. 21,26,29

| Limitations
Our data should be interpreted in the context of their limitations. In fact, this study was conducted by only 92 private and public cardiologists who accepted to participate in this registry. Only patients who consented to the study were enrolled. Therefore, not all new documented AF patients were involved, particularly those who presented to first care medical centers (noncardiologists). The follow-up period extended over only 1 year; thus, a long term follow-up is required.

| SUMMARY
These data have important clinical and public health implications for North African populations, who are in an epidemiological transition.
Valvular AF was present in one-fifth of Tunisian patients. Almost half of the registered nonvalvular AF patients were at low risk of stroke and had a high rate of acenoucoumarol use. However, the quality of anticoagulation was poor and compliance with the treatment guidelines remained suboptimal. Despite the low economic resources, health policy should enhance educational strategies, screening, management, and prevention strategies, in addition to new medication and technologies, such as catheter AF ablation and the use of direct oral anticoagulants to improve AF management and to offer better risk benefit ratios. These results highlight the need for a strategy that might have particular advantages for middle-income North African countries, having limited resources.

ACKNOWLEDGMENT
We thank the trial investigators and coordinators at all the centers, the trial monitors and staff from the DACIMA, Rabie Razgallah MD for assistance with biostatistics, and the participating patients for their contribution to the trial.

CONFLICT OF INTEREST
The authors declare no potential conflict of interests.

AUTHOR CONTRIBUTIONS
The manuscript has been read and approved by all the authors.