Trends in Pharmacologic Management of Atrial Fibrillation
Trends in Pharmacologic Management of Atrial Fibrillation
In an analysis of >275,000 patients with newly diagnosed AF treated in a large US health care system, we observed several important trends in AF management over the past decade. We found that the use of oral rate control alone remains the dominant strategy for treating new AF, with about 4 times more patients taking rate control alone compared with patients taking rhythm control medications with or without rate control.
We also found a significant reduction in the overall use of any rate control or rhythm control medication from 2002 through 2011, even after accounting for trends in the use of other pharmacologic and nonpharmacologic treatments for AF. The largest percent decreases were observed for digoxin and amiodarone. The dramatic decrease we found in digoxin, one of the oldest available cardiac medications, may reflect a growing awareness of the potential toxicity of the drug due to its narrow therapeutic window and limited efficacy. Amiodarone was by far the most frequently prescribed oral antiarrhythmic, although its use decreased by nearly 20% over our study period. Although amiodarone appears to be the most effective antiarrhythmic medication for maintenance of sinus rhythm in patients with AF, its long-term use is associated with significant toxicity. The newest antiarrhythmic drug, dronedarone, was used by only a small number of VA patients after Food and Drug Administration approval in 2009. The use of 2 β-blockers (metoprolol and carvedilol) increased in the VA since 2002.
Overall, the use of rate and rhythm control medications in our VA data was slightly lower compared with rates found in another recent study based on the Medicare 5% sample administrative data, which demonstrated that roughly three-fourths of patients with new AF received oral rate control and slightly >20% received a rhythm control agent during 2006 through 2007. However, in contrast to that study, our study excluded patients with previous cardioversion or antiarrhythmic medication experience. Without those exclusions, we found similar proportions of patients receiving oral rate control and rhythm control medications (73% and 17%, respectively).
Prospective randomized clinical trials comparing rate and rhythm control in patients with AF found no differences in major morbidity and mortality outcomes between the 2 strategies. Some data suggest that the use of antiarrhythmic medications has decreased significantly after results of the AFFIRM trial and other trials were published. For example, although the use of antiarrhythmic medications did not change from 1991 through 2003, another study suggests that a modest decrease in their use occurred from 2003 through 2006. Our data suggest that this trend continued through 2011. Similarly, other studies have shown significant decreases in digoxin use along with significant increases in the use of β-blockers through 2006—trends that also continued through 2011 according to our data. It is possible that use of these medications decreased, in part, due to a greater awareness of limitations of antiarrhythmic medications and digoxin. Other possible reasons for the decrease include unmeasured changes in patient characteristics, such as symptom severity or AF persistence, improved information distinguishing patients who are likely to have successful experience with antiarrhythmics, increased use of the ablation approach, and changes over time in the use of services outside the VA including ablation procedures.
This study has limitations that should be acknowledged. First, administrative data do not capture important prognostic variables (eg, electrocardiographic findings) that can influence choice of treatment for AF, and the accuracy of ICD-9-CM diagnosis is questionable for some conditions. Therefore, it is not possible using administrative data to capture all factors that contribute to treatment decisions. In particular, a single ICD-9-CM diagnosis code identifies AF and does not distinguish AF that is persistent or paroxysmal. We therefore did not attempt to account for these factors in our analysis. Second, findings based on the VA patient population may not generalize to the entire population of AF patients. Most VA patients are men, and rates of some comorbid conditions such as hypertension, diabetes, or chronic heart failure may be higher among VA patients compared with other populations. Moreover, some medications have limited availability in VA pharmacies. For example, flecainide was not widely available in the VA before 2009 and then was only available with a prescription from a VA cardiology specialty clinic. Third, some VA patients may also receive health services, including pharmaceuticals, in the private sector. With the passage of Medicare Part D starting in 2006, the use of pharmacies outside the VA may have increased. However, analysis of previously acquired Medicare enrollment data for VA patients revealed that only 5% of patients in this study were enrolled in Medicare Part D during the year of AF episode. Nevertheless, some patients may have other sources of insurance such as private or Medicaid. To address this limitation, we evaluated changes in the use of rate and rhythm control within patient subgroups defined by veteran enrollment priority, under the assumption that patients with lower priority are more likely to receive medications outside the VA. The use of rhythm control medications decreased similarly in all priority categories, whereas the use of rate control medications decreased more for patients with lower priority compared with patients with service-connected disabilities or low income (14% decrease vs 7% and 4%, respectively). This suggests that an increasing number of patients with minimally symptomatic AF that can be managed using rate control may receive their medication outside the VA. Fourth, the increasing use of ablative and other surgical techniques in recent years means that long-term pharmacologic therapy may not be necessary for all patients. Although our sensitivity analysis controlled for the use of alternative and complementary surgical and nonsurgical procedures for treating AF, some veterans who receive these procedures likely receive them outside the VA, and our data do not reflect procedures outside the VA.
In summary, this analysis of patients in the VA health care system provides contemporary data on current use of rate and rhythm control for newly diagnosed AF and how treatment has changed over time. Management of AF remains challenging and is continually evolving. To date, there remains ambiguity regarding best practices for managing AF, including inadequate information regarding optimal strategies for maintaining normal sinus rhythm while minimizing risk associated with long-term medication therapy, the relative benefits of rate control or rhythm control, and the long-term efficacy of the chosen strategy. The gaps in knowledge related to AF are reflected in a recent Institute of Medicine determination that ranked AF among the top priority areas for comparative effectiveness research.
Discussion
In an analysis of >275,000 patients with newly diagnosed AF treated in a large US health care system, we observed several important trends in AF management over the past decade. We found that the use of oral rate control alone remains the dominant strategy for treating new AF, with about 4 times more patients taking rate control alone compared with patients taking rhythm control medications with or without rate control.
We also found a significant reduction in the overall use of any rate control or rhythm control medication from 2002 through 2011, even after accounting for trends in the use of other pharmacologic and nonpharmacologic treatments for AF. The largest percent decreases were observed for digoxin and amiodarone. The dramatic decrease we found in digoxin, one of the oldest available cardiac medications, may reflect a growing awareness of the potential toxicity of the drug due to its narrow therapeutic window and limited efficacy. Amiodarone was by far the most frequently prescribed oral antiarrhythmic, although its use decreased by nearly 20% over our study period. Although amiodarone appears to be the most effective antiarrhythmic medication for maintenance of sinus rhythm in patients with AF, its long-term use is associated with significant toxicity. The newest antiarrhythmic drug, dronedarone, was used by only a small number of VA patients after Food and Drug Administration approval in 2009. The use of 2 β-blockers (metoprolol and carvedilol) increased in the VA since 2002.
Overall, the use of rate and rhythm control medications in our VA data was slightly lower compared with rates found in another recent study based on the Medicare 5% sample administrative data, which demonstrated that roughly three-fourths of patients with new AF received oral rate control and slightly >20% received a rhythm control agent during 2006 through 2007. However, in contrast to that study, our study excluded patients with previous cardioversion or antiarrhythmic medication experience. Without those exclusions, we found similar proportions of patients receiving oral rate control and rhythm control medications (73% and 17%, respectively).
Prospective randomized clinical trials comparing rate and rhythm control in patients with AF found no differences in major morbidity and mortality outcomes between the 2 strategies. Some data suggest that the use of antiarrhythmic medications has decreased significantly after results of the AFFIRM trial and other trials were published. For example, although the use of antiarrhythmic medications did not change from 1991 through 2003, another study suggests that a modest decrease in their use occurred from 2003 through 2006. Our data suggest that this trend continued through 2011. Similarly, other studies have shown significant decreases in digoxin use along with significant increases in the use of β-blockers through 2006—trends that also continued through 2011 according to our data. It is possible that use of these medications decreased, in part, due to a greater awareness of limitations of antiarrhythmic medications and digoxin. Other possible reasons for the decrease include unmeasured changes in patient characteristics, such as symptom severity or AF persistence, improved information distinguishing patients who are likely to have successful experience with antiarrhythmics, increased use of the ablation approach, and changes over time in the use of services outside the VA including ablation procedures.
This study has limitations that should be acknowledged. First, administrative data do not capture important prognostic variables (eg, electrocardiographic findings) that can influence choice of treatment for AF, and the accuracy of ICD-9-CM diagnosis is questionable for some conditions. Therefore, it is not possible using administrative data to capture all factors that contribute to treatment decisions. In particular, a single ICD-9-CM diagnosis code identifies AF and does not distinguish AF that is persistent or paroxysmal. We therefore did not attempt to account for these factors in our analysis. Second, findings based on the VA patient population may not generalize to the entire population of AF patients. Most VA patients are men, and rates of some comorbid conditions such as hypertension, diabetes, or chronic heart failure may be higher among VA patients compared with other populations. Moreover, some medications have limited availability in VA pharmacies. For example, flecainide was not widely available in the VA before 2009 and then was only available with a prescription from a VA cardiology specialty clinic. Third, some VA patients may also receive health services, including pharmaceuticals, in the private sector. With the passage of Medicare Part D starting in 2006, the use of pharmacies outside the VA may have increased. However, analysis of previously acquired Medicare enrollment data for VA patients revealed that only 5% of patients in this study were enrolled in Medicare Part D during the year of AF episode. Nevertheless, some patients may have other sources of insurance such as private or Medicaid. To address this limitation, we evaluated changes in the use of rate and rhythm control within patient subgroups defined by veteran enrollment priority, under the assumption that patients with lower priority are more likely to receive medications outside the VA. The use of rhythm control medications decreased similarly in all priority categories, whereas the use of rate control medications decreased more for patients with lower priority compared with patients with service-connected disabilities or low income (14% decrease vs 7% and 4%, respectively). This suggests that an increasing number of patients with minimally symptomatic AF that can be managed using rate control may receive their medication outside the VA. Fourth, the increasing use of ablative and other surgical techniques in recent years means that long-term pharmacologic therapy may not be necessary for all patients. Although our sensitivity analysis controlled for the use of alternative and complementary surgical and nonsurgical procedures for treating AF, some veterans who receive these procedures likely receive them outside the VA, and our data do not reflect procedures outside the VA.
In summary, this analysis of patients in the VA health care system provides contemporary data on current use of rate and rhythm control for newly diagnosed AF and how treatment has changed over time. Management of AF remains challenging and is continually evolving. To date, there remains ambiguity regarding best practices for managing AF, including inadequate information regarding optimal strategies for maintaining normal sinus rhythm while minimizing risk associated with long-term medication therapy, the relative benefits of rate control or rhythm control, and the long-term efficacy of the chosen strategy. The gaps in knowledge related to AF are reflected in a recent Institute of Medicine determination that ranked AF among the top priority areas for comparative effectiveness research.
Source...