Rate vs Rhythm Control for Atrial Fibrillation Management
Rate vs Rhythm Control for Atrial Fibrillation Management
The entire baseline ORBIT-AF population included 10,098 patients enrolled between June 29, 2010 and August 09, 2011. The current analysis excluded 37 patients: 1 due to missing AF type/diagnosis and 36 for missing AF management strategy. This yielded a final study population of 10,061 patients from 174 sites. Baseline characteristics of the study cohort are shown in Table I according to rate or rhythm control. Over two-thirds were managed with a primary strategy of rate only control (n = 6,859, 68%) and nearly one-third with rhythm control (n = 3,202, 32%). Those managed with rhythm control were younger and had less medical co-morbidity. They also had a higher body mass index, calculated creatinine clearance, and left-ventricular ejection fraction, but lower resting heart rate.
Patients managed with a rhythm control strategy were significantly more likely to be in sinus rhythm on their most recent electrocardiogram (Table II). They were more likely to have paroxysmal AF, higher symptom scores, and had lower CHADS2 scores. Almost one-third of patients in the rate control group had current or prior antiarrhythmic drug use, whereas 82% of patients managed with rhythm control were previously (or currently) taking antiarrhythmic therapy. Nearly one-third of the population had a prior cardioversion, and 11% had a prior electrophysiology intervention. These were both more common in the rhythm control group (P < .0001 for each).
Unadjusted rates of medical therapies for AF are shown in Table III. A significant proportion of rhythm-control patients were also taking pure rate-controlling medications, yet to a lesser extent compared with rate-control patients. Strategies for the prevention of thromboembolism in each group, stratified by CHADS2 score, are shown in Figure 1. Overall, rhythm-controlled patients were more likely to be taking aspirin alone (21% vs. 12%, P < .0001), and less likely to be treated with oral anticoagulation (69% vs. 79%, P < .0001), despite a lower rate of contraindication to systemic anticoagulation (12% vs 15%, P = .002). As the CHADS2 score increased, use of lone aspirin therapy decreased in favor of systemic anticoagulation. However, at all levels of risk, patients managed with rate control were significantly less likely to be prescribed aspirin alone and more likely to be on systemic anticoagulation. Therapies used to control heart rate, in unadjusted and adjusted analyses, are displayed in Figure 2. The use of nodal-blocking agents significantly favored the rate control group.
(Enlarge Image)
Figure 1.
Unadjusted use of antithrombotic therapies. Aspirin only included aspirin/dipyridamole, (n = 14, 0.14%). Any ADP inhibitor included clopidogrel or prasugrel, with or without aspirin, but no oral anticoagulant. Systemic anticoagulation included warfarin or dabigatran (with any antiplatelet). *P < .05 for the comparison between rate control and rhythm control groups. ADP: adenosine diphosphate.
(Enlarge Image)
Figure 2.
Unadjusted and adjusted comparisons of medical therapies between strategies. Multivariable rates adjusted for age, left atrial diameter, posterior wall thickness, level of education, site region, medical history of frailty, AF type, and provider specialty. BB, β-blocker; CCB, calcium-channel blocker.
Clinical features associated with selection of rhythm control (versus rate control) after multivariable adjustment are shown in Figure 3 (c-index 0.74). Referral to an electrophysiologist (adjusted OR 1.68, 95% CI 1.49–1.90, P < .0001), paroxysmal AF (adjusted OR 1.49, 95% CI 1.22–1.83, P < .0001), and more advanced educational background (post-graduate vs. some school, adjusted OR 1.48, 95% CI 1.20–1.83, P = .0002) all were associated rhythm control management. In contrast, older patients, those with longstanding AF, and those managed by primary care physicians were more likely to be treated with a rate control strategy. When the European Heart Rhythm Association (EHRA) score was added to the model, the presence of symptoms was also associated with selection of rhythm control strategy (compared to no symptoms [EHRA score I]): mild symptoms (EHRA score II) OR 1.25, 95% CI 1.12–1.38, severe symptoms (EHRA score III) OR 1.62, 95% CI 1.41–1.87, and disabling symptoms (ERHA score IV) OR 1.42, 95% CI 1.02–1.96. After adjustment for symptom burden, electrophysiology provider specialty, type of AF (paroxysmal), and younger age remained associated with a rhythm control strategy.
(Enlarge Image)
Figure 3.
Multivariable analysis of factors associated with AF management strategy. Boxes denote adjusted OR with lines to 95% CIs.
Results
The entire baseline ORBIT-AF population included 10,098 patients enrolled between June 29, 2010 and August 09, 2011. The current analysis excluded 37 patients: 1 due to missing AF type/diagnosis and 36 for missing AF management strategy. This yielded a final study population of 10,061 patients from 174 sites. Baseline characteristics of the study cohort are shown in Table I according to rate or rhythm control. Over two-thirds were managed with a primary strategy of rate only control (n = 6,859, 68%) and nearly one-third with rhythm control (n = 3,202, 32%). Those managed with rhythm control were younger and had less medical co-morbidity. They also had a higher body mass index, calculated creatinine clearance, and left-ventricular ejection fraction, but lower resting heart rate.
Patients managed with a rhythm control strategy were significantly more likely to be in sinus rhythm on their most recent electrocardiogram (Table II). They were more likely to have paroxysmal AF, higher symptom scores, and had lower CHADS2 scores. Almost one-third of patients in the rate control group had current or prior antiarrhythmic drug use, whereas 82% of patients managed with rhythm control were previously (or currently) taking antiarrhythmic therapy. Nearly one-third of the population had a prior cardioversion, and 11% had a prior electrophysiology intervention. These were both more common in the rhythm control group (P < .0001 for each).
Unadjusted rates of medical therapies for AF are shown in Table III. A significant proportion of rhythm-control patients were also taking pure rate-controlling medications, yet to a lesser extent compared with rate-control patients. Strategies for the prevention of thromboembolism in each group, stratified by CHADS2 score, are shown in Figure 1. Overall, rhythm-controlled patients were more likely to be taking aspirin alone (21% vs. 12%, P < .0001), and less likely to be treated with oral anticoagulation (69% vs. 79%, P < .0001), despite a lower rate of contraindication to systemic anticoagulation (12% vs 15%, P = .002). As the CHADS2 score increased, use of lone aspirin therapy decreased in favor of systemic anticoagulation. However, at all levels of risk, patients managed with rate control were significantly less likely to be prescribed aspirin alone and more likely to be on systemic anticoagulation. Therapies used to control heart rate, in unadjusted and adjusted analyses, are displayed in Figure 2. The use of nodal-blocking agents significantly favored the rate control group.
(Enlarge Image)
Figure 1.
Unadjusted use of antithrombotic therapies. Aspirin only included aspirin/dipyridamole, (n = 14, 0.14%). Any ADP inhibitor included clopidogrel or prasugrel, with or without aspirin, but no oral anticoagulant. Systemic anticoagulation included warfarin or dabigatran (with any antiplatelet). *P < .05 for the comparison between rate control and rhythm control groups. ADP: adenosine diphosphate.
(Enlarge Image)
Figure 2.
Unadjusted and adjusted comparisons of medical therapies between strategies. Multivariable rates adjusted for age, left atrial diameter, posterior wall thickness, level of education, site region, medical history of frailty, AF type, and provider specialty. BB, β-blocker; CCB, calcium-channel blocker.
Clinical features associated with selection of rhythm control (versus rate control) after multivariable adjustment are shown in Figure 3 (c-index 0.74). Referral to an electrophysiologist (adjusted OR 1.68, 95% CI 1.49–1.90, P < .0001), paroxysmal AF (adjusted OR 1.49, 95% CI 1.22–1.83, P < .0001), and more advanced educational background (post-graduate vs. some school, adjusted OR 1.48, 95% CI 1.20–1.83, P = .0002) all were associated rhythm control management. In contrast, older patients, those with longstanding AF, and those managed by primary care physicians were more likely to be treated with a rate control strategy. When the European Heart Rhythm Association (EHRA) score was added to the model, the presence of symptoms was also associated with selection of rhythm control strategy (compared to no symptoms [EHRA score I]): mild symptoms (EHRA score II) OR 1.25, 95% CI 1.12–1.38, severe symptoms (EHRA score III) OR 1.62, 95% CI 1.41–1.87, and disabling symptoms (ERHA score IV) OR 1.42, 95% CI 1.02–1.96. After adjustment for symptom burden, electrophysiology provider specialty, type of AF (paroxysmal), and younger age remained associated with a rhythm control strategy.
(Enlarge Image)
Figure 3.
Multivariable analysis of factors associated with AF management strategy. Boxes denote adjusted OR with lines to 95% CIs.
Source...