Emergency department visit for atrial fibrillation: sex differences in treatment and outcomes in the Global RE-LY AF Registry

Studies of sex-based differences in atrial fibrillation (AF) suggest an influence of sex on cardiovascular death and stroke, however, results are conflicting.1,2 Discrepant findings could reflect sex-based differences in access to care, but no studies have explored sex-based differences in treatment and outcomes among countries with differing income levels and gender parity. Such data are needed to understand if sex-based care gaps exist and are associated with differences in outcomes. This knowledge could lead to country and sex-specific treatment recommendations. This study explores how sex differences in AF treatment and outcomes vary between countries based on their economic status and degree of gender parity in the Global RE-LY AF Registry. From the prospective RE-LY registry [n = 15 400 patients presenting to an emergency department (ED) with AF in 47 countries between 2007 and 2011],3,4 we excluded patients without AF as primary listed reason for ED presentation (n = 8561), or missing outcomes or CHADS2 score (n = 213), resulting in a sample of 6626 patients. We defined rhythm control as treatment with cardioversion, AF ablation, or the use of any anti-arrhythmic drug. Outcomes were obtained at one-year follow-up and included repeated hospitalization for AF, heart failure (HF) hospitalization, stroke or transient ischaemic attack (TIA), and death.3 Selected baseline variables are presented as means ± standard deviation, median [interquartile range (IQR)] or proportion. Sex-based differences in treatments and outcomes are presented as crude proportions, with odds ratios (OR) for female sex compared to male sex and P-values derived from multi-level logistic regression with a random effect on country, adjusted for CHADS2 score, which was the recommended risk stratification tool at the time of study initiation.4 To explore the influence on outcome risks by gender-based disparities or economic factors, we stratified on the World Economic Forum (WEF) Global Gender Gap score for 2011, which estimates country-level overall gender parity with a 0–100 score annually. We used World Bank classifications of income for 2011 to group countries as ‘low and lower-middle’, ‘upper-middle’, and ‘high’ income countries. Interaction parameters were assessed in CHADS2-adjusted logistic regression models. All statistical analyses were performed using Stata v 17.0 for Mac (StataCorp, College Station, TX, USA). The study conforms to the Declaration of Helsinki and received ethical approval at all sites. All subjects gave written informed consent. Overall, females were older (65.5 ± 14.4 vs. 61.5 ± 14.2 years, P < .0001), had a higher median CHADS2 score [1 (IQR 1) vs. 1 (IQR 2), P < .0001], and more permanent AF (21.5% vs. 18.9%, P = .008). The ED visit resulted in hospitalization in 56.1% of females and 53.6% of males (P = .09). Females were less likely to be treated with a rhythm control strategy, receive cardioversion, or AF ablation during follow-up, but were slightly more likely to be treated with anticoagulation (Table 1). The effect of female sex on rhythm control therapy utilization differed by country-level gender parity (P for interaction .006), and World Bank income groups (P for interaction .001) (Table 1). Contraindications to anticoagulation were reported in 7.0% of females and 6.0% of males (P = .002). Population proportion treated with different strategies and proportion of outcomes in males and females, with P-values and odds ratios for female sex compared to male sex, derived from CHADS2-adjusted multi-level logistic regression models with a random effect for country Bold values denote statistical significance. AF, atrial fibrillation; ED, emergency department; HF, heart failure; TIA, transient ischaemic attack; Q, quartile. aConsists of Ukraine, Nigeria, Tanzania, Cameroon, Kenya, Mozambique, Uganda, Senegal, Sudan, and India, and includes 1727 individuals and 155 repeat visits to hospital for AF, 123 hospitalizations for HF, 34 stroke events, and 110 deaths. bConsists of Argentina, Brazil, Ecuador, Chile, Colombia, Venezuela, Russia, Latvia, Turkey, Iran, South Africa, Thailand, and China, and includes 1925 individuals and 384 repeat visits to hospital for AF, 148 hospitalizations for HF, 76 stroke events, and 117 deaths. cIncludes Canada, USA, Denmark, Sweden, Ireland, the UK, Austria, Germany, the Netherlands, Italy, Spain, Australia, Bulgaria, the Czech Republic, Hungary, Poland, Slovakia, the United Arab Emirates, Saudi Arabia, Singapore, Japan, and Korea, and includes 2974 individuals and 785 repeat visits to hospital for AF, 182 hospitalizations for HF, 61 stroke events, and 111 deaths. dQ1 consists of Turkey, Saudi Arabia, Iran, Nigeria, Cameroon, India, and Korea. Includes 1550 individuals and 72 AF hospitalizations, 45 HF hospitalizations, 11 stroke events, and 86 deaths. eQ2 consists of Brazil, Colombia, Venezuela, Italy, the Czech Republic, Hungary, Slovakia, Ukraine, Kenya, Senegal, Japan, and the United Arab Emirates. Includes 1589 individuals and 252 AF hospitalizations, 114 HF hospitalizations, 40 stroke events, and 84 deaths. fQ3 consists of China, Thailand, Singapore, Uganda, Mozambique, Tanzania, Russia, Poland, Bulgaria, Australia, Austria, Chile, Ecuador, and Argentina. Includes 1544 individuals and 392 AF hospitalizations, 169 HF hospitalizations, 72 stroke events, and 109 deaths. gQ4 consists of Canada, USA, Denmark, Ireland, Sweden, the UK, Germany, Holland, Spain, Latvia, and South Africa. Includes 1924 individuals and 605 AF hospitalizations, 123 HF hospitalizations, 46 stroke events, and 59 deaths. *P < .01; **P < .001; ***P < .05. After adjustment for the CHADS2 score, female sex was associated with increased risk of stroke [OR 1.59, 95% confidence interval (CI) 1.15–2.19, P = .005], which remained significant after adjustment for anticoagulation use (OR 1.60, 95% CI 1.16–2.21, P = .004), and among patients with non-valvular AF (n = 5209) (OR 1.62, 95% CI 1.15–2.30, P = .006), including when using Bonferroni correction in the overall population. This risk was numerically more prominent in upper-middle and high-income countries (P for interaction .51), and in the top two quartiles of Global Gender Gap score (P for interaction .17) (Table 1). We also tested an extensively adjusted model including age, tobacco use, AF type, alcohol use, anticoagulation use, history of sleep apnoea, HF, diabetes, hypertension, stroke, chronic obstructive pulmonary disease, rheumatic heart disease, myocardial infarction, and severe valvular disease. Results were attenuated, but remained significant (OR 1.42, 95% CI 1.01–1.99, P = .045). Female sex was not associated with rehospitalization for AF, hospitalization for HF, or death. In this global, ED-based registry of AF, females were less likely to be managed with rhythm control, particularly in lower income countries and those with less gender equity. There was no overall sex-related difference in the one-year rate of AF or HF hospitalization or death, but females did have a greater risk of stroke, even after extensive multivariable adjustment. Studies of sex-based differences in stroke risk in AF need to be interpreted in the context of age, comorbidities, and oral anticoagulation use, and also consider competing risks of mortality affecting males. Increased stroke risk in AF among females has been reported, notably in the ATRIA study,2 and an Ontario population-based study.5 Other studies, including the PREFER-AF,1 which had high anticoagulation rates in both sexes, and a Quebec registry, that used time-updated covariates,6 did not detect such a difference. Our analysis adds to the understanding of the increased risk of stroke among females by showing that this is present primarily in countries with high gender parity and in high-income countries. In contrast to the effect of female sex on stroke, the lower use of rhythm control among females in this study was independent of comorbidities, and most pronounced in countries with lower gender parity. Prior to enrolment into this registry, data from the RACE and AFFIRM studies did not show an advantage for rhythm control.7,8 However, recent data from the EARLY-AF and EAST-AFNET4 studies suggest a benefit of rhythm control for prevention of AF recurrence and outcomes.9,10 In light of this, it is important that females are offered rhythm control therapies to the same extent as males. In the year after an ED visit for AF, females were less likely to receive rhythm control therapy, most notably in lower income countries or those with less gender parity. Despite this, there was no effect of female sex on AF or HF hospitalization or mortality. Female sex was associated with an increased risk of stroke/TIA, even after adjustment for comorbidities and anticoagulant use. L.S.J. receives consulting fees from MEDICALgorithmics. J.S.H. has research grants and speaking fees from BMS/Pfizer, Boehringer-Ingelheim, Boston Scientific, Novartis, Medtronic, and Servier. D.C. has received consultation fees from Roche Diagnostics and Trimedics, as well as speaker fees from Servier and BMS/Pfizer. Data will be made available by reasonable request to the corresponding author. The RE-LY AF registry was funded by Boehringer-Ingelheim. L.S.J. is supported by the Swedish Heart and Lung Foundation, the Swedish Society for Medical Research, the Swedish Research Council, and the Swedish Society for Medicine. The study was approved by an institutional review board and all patients gave written informed consent. None supplied.