Racial/Ethnic and Socioeconomic Disparities in Management of Incident Paroxysmal Atrial Fibrillation

This cohort study examines inpatient, outpatient, and pharmacy claims to identify use of antiarrhythmic drugs and catheter ablation among commercially insured adults with paroxysmal atrial fibrillation.


Introduction
Atrial fibrillation (AF), the most common sustained cardiac arrhythmia, is associated with substantial morbidity and mortality, including frequent hospitalizations, heart failure, and thromboembolic events. 1 Although a rate control strategy is a reasonable approach for managing AF in many patients, certain factors clearly favor a rhythm control strategy. 2 For patients with symptomatic paroxysmal AF, a rhythm control strategy has been shown to improve quality of life. 3 Paroxysmal AF often progresses to persistent AF, and a rhythm control strategy may halt the accompanying structural and electrical adverse remodeling. 2,4 Furthermore, recent evidence suggests that early pursuit of a rhythm control strategy is associated with improved cardiovascular outcomes. 5 Maintenance of sinus rhythm may be achieved with the use of antiarrhythmic drugs (AADs) or catheter ablation. 2,4,6 Current guidelines recommend catheter ablation as a first-line treatment for symptomatic paroxysmal AF. 2 Among patients with heart failure with reduced ejection fraction (HFrEF) and AF, catheter ablation has been shown to decrease mortality and heart failure hospitalizations as well as increase left ventricular ejection fraction. [7][8][9][10] Inequitable care delivery based on race/ethnicity is pervasive, and decreased access to cardiovascular therapies among patients from racial/ethnic minority groups and those with lower income has been demonstrated. 11 There is decreased adoption of advanced cardiovascular therapeutics among Black, Latinx, and low-income patients. [12][13][14][15] Demographic studies have shown that patients with AF who underwent catheter ablation were mostly White and male. 16 Racial and sex inequities in receipt of certain AF-related therapies were demonstrated among Medicare patients. 17 However, broader, larger-scale studies are lacking that evaluate the association of race and socioeconomic status with use of a rhythm control strategy, particularly catheter ablation, among commercially insured patients with paroxysmal AF. The objectives of this cohort study were to ascertain the cumulative rates of AAD and catheter ablation use for the management of paroxysmal AF and to investigate the presence of inequities in AF management by evaluating the association of race/ethnicity and socioeconomic status with their use in the United States.

Methods
The University of Pennsylvania Institutional Review Board determined that this research was exempt from the regulatory requirements of the federal Common Rule because no protected health information was used. We followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline. race/ethnicity, are available for each member. Socioeconomic data, including median household income, are available through zip code-linked enrollment data from the US Census Bureau.
We identified adult patients (aged Ն18 years) in the database with a diagnosis of paroxysmal AF documented with International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) code I48.0 between October 1, 2015, and June 30, 2019. We required the diagnosis to be coded on at least 2 occasions on separate dates either in the inpatient or outpatient setting. Patients entered the study cohort on the date of the second coded diagnosis for paroxysmal AF. Patients were excluded if they did not have continuous insurance enrollment for at least 1 year before and at least 6 months after study entry to ensure that comorbidities, clinical data, and prescription claims could be accurately identified. In addition, patients were excluded if they had no pharmacy claims for medication for 1 year before the study period to ensure that medication use was accurately captured in the data set.
To ensure that only incident paroxysmal AF was analyzed, we excluded patients with preexisting AF diagnoses based on International Classification of Diseases, Ninth Revision (ICD-9) code 427.31 before the ICD-10 diagnosis as well as patients with previous prescription claims for amiodarone hydrochloride, disopyramide phosphate, dofetilide, dronedarone hydrochloride, flecainide acetate, mexiletine hydrochloride, propafenone hydrochloride, quinidine, or sotalol hydrochloride before the ICD-10 diagnosis.
Rhythm control was defined by a prescription claim for 1 or more AAD after cohort entry (ie, amiodarone, disopyramide, dofetilide, dronedarone, flecainide, mexiletine, propafenone, quinidine, or sotalol) or by a Current Procedural Terminology (CPT) code for catheter ablation (93656). Rate control was defined as the absence of rhythm control. All patients who had a CPT code for catheter ablation, regardless of AAD use, were included in the catheter ablation group. The mean number of outpatient cardiology visits per 12 months after cohort entry was ascertained on the basis of having a visit with a cardiology practitioner with CPT codes 99201-99205 or 99211-99215.

Statistical Analysis
Summary statistics for patient characteristics are presented as medians with interquartile ranges (IQRs) or means with SDs for continuous data and as a total number and percentages for categorical data. Continuous variables were compared using an unpaired, 2-tailed t test, and categorical variables were compared using the χ 2 test. Cumulative yearly rates of AAD and catheter ablation use To assess the association of race/ethnicity, sex, and socioeconomic status with the use of rhythm control strategy (both AADs and catheter ablation), we estimated multivariable logistic regression models with the use of AADs or catheter ablation as the dependent variable and with demographic factors (age, sex, race/ethnicity, region of residence, zip code-linked median household income, and health insurance subset), clinical factors (anticoagulation status, number of outpatient cardiology visits per 12 months), and comorbidities (number of Elixhauser comorbidities, 18 coronary artery disease, chronic kidney disease, stroke or transient ischemic attack, obesity, hypertension, diabetes, peripheral vascular disease, dyslipidemia, HFrEF, and heart failure with preserved ejection fraction) as the independent variables.
To assess the association of race/ethnicity, sex, and socioeconomic status with the use of catheter ablation among those in whom a rhythm control strategy was selected, we included patients who had either a prescription for AADs or a CPT code for catheter ablation during the study period.
We estimated multivariable logistic regression models with the use of catheter ablation as the dependent variable and the aforementioned independent variables. Candidate variables included in the model were selected a priori on the basis of pathophysiological plausibility or because of significance on univariate analysis. Estimated adjusted odds ratios (aORs) are reported with 95% CIs.
Among those who underwent catheter ablation, the number of AADs prescribed after paroxysmal AF diagnosis and before catheter ablation was compared between racial/ethnic, sex, and zip code-linked median household income groups. Statistical analyses were performed using SAS, version 9.4 (SAS Institute, Inc). All statistical testing was 2-tailed, with P < .05 designated as statistically significant. to 2.8% among Black patients, from 1.1% to 2.5% among Latinx patients, and from 1.5% to 4.2% among White patients.

Rhythm Control vs Rate Control
In multivariable analyses (

Catheter Ablation vs Antiarrhythmic Drugs in Rhythm Control
Independent factors associated with catheter ablation use in multivariable analysis are summarized in

Discussion
To our knowledge, this study is the first to investigate for the presence of disparities in paroxysmal AF management and catheter ablation by race/ethnicity and socioeconomic status among patients with commercial insurance in the US. Between 2016 and 2019, we found that the rate of AAD and catheter ablation use increased among patients with paroxysmal AF, yet the rate of catheter ablation

JAMA Network Open | Cardiology
Racial/Ethnic and Socioeconomic Disparities in Management of Incident Paroxysmal Atrial Fibrillation remained low (only 3.2% of patients with incident paroxysmal AF), particularly among patients with HFrEF. Among a commercially insured population, significant disparities in treatment strategy based on race/ethnicity and socioeconomic status were observed. Black race and lower zip code-linked median household income were independently associated with lower use of a rhythm control strategy. Latinx ethnicity and lower zip code-linked median household income were independently associated with lower use of catheter ablation, even among patients in whom a rhythm control strategy was pursued.
In this study, Black race was independently associated with lower use of a rhythm control strategy. However, among those in whom a rhythm control strategy was pursued, Black race was not associated with lower rates of catheter ablation.  14,15,22 We were unable to account for language in the database, but appreciable inequities in health care use, particularly for specialized procedures and care, among non-English-speaking Latinx patients have been reported previously 15 and may have played a role in the findings of the present study.
The differences we observed in the care of Black and Latinx patients with paroxysmal AF were in the setting of a population with 100% commercial insurance and were adjusted for outpatient cardiology visits, which raise concerns about bias in care delivery. 23 Patient advocacy may be a factor in treatment decisions given that White patients may more frequently advocate for rhythm control strategies and potentially curative ablation procedures. Racial and ethnic differences in self-advocacy have been observed and are likely reflective of a long history of discrimination against racial/ethnic minority groups in the US health care system. 24 Clinician interactions with patients from racial/ethnic minority groups have been characterized by less patient input regarding treatment decisions. 25 Lower rates of patient activation (ie, confidence to engage in health care and treatment decisions among marginalized patients) because of frequent and recurrent inequitable treatment by the health system may play a role in the demonstrated differences in paroxysmal AF management. 26,27 We also found significant differences in the use of catheter ablation by socioeconomic status.
Overall, patients with a zip code-linked median household income of less than $50 000 were 17% less likely to receive a rhythm control strategy and were 39% less likely to receive catheter ablation compared with those with a median household income of $100 000 or more. In addition to barriers to accessing more advanced cardiac electrophysiology care, affordability of AADs and catheter ablation likely contributed to our findings. Similar to findings in patients from racial/ethnic minority groups, it has been shown that novel therapies are less likely to be used among patients with lower income. 28 This finding may also reflect clinician bias; clinicians may preferentially offer catheter ablation to patients with higher income because of perceptions that patients with lower income are unable to afford such treatment. 29

JAMA Network Open | Cardiology
We found that the rate of catheter ablation was low (3.2%) for the overall population and for patients with HFrEF (2.8%), which was surprising. Given the increasing evidence of benefits in this population, including improved mortality, decreased rate of hospitalization for heart failure, and increased left ventricular ejection fraction, 5-8 the American College of Cardiology/American Heart Association updated AF guidelines in 2019 to include a recommendation that catheter ablation may be reasonable to use for this population. 4 The study period preceded some of the key trials that led to this change in guidelines. [6][7][8] However, given more recent evidence of improved cardiovascular outcomes associated with pursuit of an early rhythm control strategy among all patients with AF regardless of symptoms, 5 the implementation of strategies that address inequitable access to rhythm control strategies is important moving forward.

Limitations
This study has several limitations. Because we obtained data from an administrative database, we were unable to fully characterize the complexity of patient care decisions. We were unable to describe how patient preference, clinician expertise, nuanced clinical decision-making for individual patients, and practice patterns impacted treatment decisions. The differences in patient preferences for invasive procedures may have been a factor in the differences in catheter ablation use. Decreased preference for invasive cardiac procedures among non-White patients has been shown to be mediated by lack of familiarity with these procedures, which in large part is related to differences and biases in clinician communication. 30 Improved communication regarding catheter ablation and its benefits for marginalized patients may improve equitable uptake.
The presence of symptoms or AF burden has been a key factor for many clinicians to pursue rhythm control and catheter ablation in paroxysmal AF. We were unable to discern AF burden or if patients had symptomatic AF with this data set. However, we suspect that these findings reflect inequities in care because we do not expect Black or Latinx patients or patients with lower income to experience less symptomatic disease. Some evidence suggests that non-White patients may have more symptoms with a greater impact on quality of life. 31 In addition, certain patients, such as those from a lower socioeconomic status, likely have a higher threshold to present for care unless substantial symptoms are present. Furthermore, even among those for whom a rhythm control strategy was selected (and therefore were presumably symptomatic or had other important indications for rhythm control), catheter ablation was used less frequently for patients from racial/ ethnic minority groups and patients with a lower income. Alternatively, the differences that we observed may reflect the overuse of therapies, such as catheter ablation, when not clinically indicated for White patients with higher income, which we could not discern from the database.
Patients with paroxysmal AF were selected using ICD-10 codes only given that ICD-10 requires specifying AF type (eg, paroxysmal, persistent), whereas ICD-9 does not. Nevertheless, it is still possible that some patients were misclassified. The findings may not apply to populations with persistent and long-standing persistent AF.
The population studied had commercial insurance, with a large proportion covered by Medicare Advantage, and was slightly older than the general population with AF. These characteristics may limit generalizability for other payer groups or younger populations. However, we anticipate that the observed differences may be even greater among those with Medicaid or without health care insurance.
Despite these limitations, the results of this study highlight important treatment differences based on race/ethnicity and socioeconomic status. These results emphasize the need for continued efforts to ensure that access to all AF treatments is equitable.

Conclusions
In this cohort study of a diverse, commercially insured population, rates of AAD and catheter ablation use for the management of patients with paroxysmal AF increased between 2016 and 2019.
However, catheter ablation use remained low, particularly among patients with HFrEF. Black race and lower zip code-linked median household income were independently associated with less use of a rhythm control strategy, whereas Latinx ethnicity and lower income were independently associated with lower use of catheter ablation. These findings suggest that both racial/ethnic and socioeconomic inequities may be present in accessing advanced AF treatments. Additional studies to better understand the barriers to these AF therapies to ensure equitable access to all therapies are warranted.