Association of Race/Ethnicity-Specific Changes in Antihypertensive Medication Classes Initiated Among Medicare Beneficiaries With the Eighth Joint National Committee Panel Member Report

Key Points Question Have the initial antihypertensive medication regimens filled by older US adults with hypertension changed following publication of the Eighth Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure panel member report? Findings In this serial cross-sectional study of 41 340 Medicare beneficiaries, there was no statistically significant change in the proportion of Black beneficiaries initiating antihypertensive monotherapy with an angiotensin-converting enzyme inhibitor or angiotensin receptor blocker following publication of the panel members’ report. The proportion initiating β-blocker monotherapy remained high among all race/ethnicity groups. Meaning Many older US adults who initiate antihypertensive medication do so with non–guideline-recommended classes of medication.


Introduction
In December 2013, the panel members appointed to the Eighth Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC8) published a report with evidence-based treatment recommendations for patients initiating antihypertensive medication. 1 Both the JNC8 panel member report and the previous seventh (JNC7) guideline recommended thiazide-type diuretics as first-line therapy, regardless of race/ethnicity, among patients without compelling indications for specific classes of antihypertensive medication. 1,2 However, the JNC8 panel member report differed from the JNC7 guideline by including race/ ethnicity-specific recommendations for additional initial medications, either alone or as components of combination therapy. For non-Black adults, the JNC8 panel member report recommended thiazide-type diuretics, calcium channel blockers (CCBs), angiotensin-converting-enzyme inhibitors (ACEIs), or angiotensin receptor blockers (ARBs). Black adults were recommended thiazide-type diuretics or CCBs. 1 The panel did not recommend β-blockers as first-line therapy in the absence of either coronary heart disease (CHD) or heart failure with reduced ejection fraction.
The race/ethnicity-specific treatment recommendations in the JNC8 panel member report were based on randomized trial data showing that Black adults have smaller blood pressure (BP) and cardiovascular disease (CVD) risk decreases than White adults when treated with an ACEI. 3,4 Prespecified subgroup analyses of the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT) found that the risks of stroke, heart failure, and CVD were higher in Black participants randomized to an ACEI vs a thiazide-type diuretic, 3

and the risk of stroke was higher in
Black participants randomized to an ACEI vs a CCB. 4 No differences were present for White participants. β-Blockers were not recommended for first-line therapy because there is evidence of increased CVD risk with β-blockers compared with ARBs. 5 The purpose of the present study was to evaluate changes in the initial antihypertensive medication regimens filled between 2011 and 2018 among older US adults who were Black, White, or of other race/ethnicity. In addition, we evaluated changes in the initial classes of antihypertensive medication filled before vs after December 18, 2013, the date the JNC8 panel member report was published.

Methods
We conducted a serial cross-sectional analysis of a 5% sample of Medicare beneficiaries initiating antihypertensive medication between 2011 and 2018. Medicare is a national program funded by the US federal government that provides health insurance to US adults aged 65 years or older and individuals younger than 65 years who are disabled or have end-stage kidney disease. Deidentified data on Medicare beneficiaries' inpatient, outpatient, and prescription drug claims were obtained from the Centers for Medicare and Medicaid services. We prepared the current manuscript according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cross-sectional studies. The University of Alabama at Birmingham institutional review board approved this study and waived the requirement for obtaining informed consent because the research involved no more than minimal risk to the participants and we are unable to contact beneficiaries because the claims data are deidentified.
Study eligibility was determined for each calendar year, separately (eFigure 1 in the Supplement). We required beneficiaries to have at least 1 pharmacy fill for an antihypertensive medication between 2011 and 2018 (eTable 1 in the Supplement). The date of each beneficiary's first antihypertensive medication fill in each calendar year defined the index date for that person. Fills for antihypertensive medication were defined using prescription claims in the Medicare Part D file, and we considered all antihypertensive medication classes that beneficiaries filled within 7 days following their index date as part of their initial regimen. Analyses were restricted to beneficiaries 66 years of age or older and younger than 110 years of age on their index date. A minimum age of 66 years was required to allow 365 days prior to the index date (ie, the look-back period) to assess covariates and study eligibility criteria while including beneficiaries with Medicare coverage due to older age vs being disabled or having end-stage kidney disease. To ensure complete data capture, we required beneficiaries to have (1)  We excluded beneficiaries with more than 1 birth date or death date or, for those who died, any claims after their date of death. We restricted the analysis to beneficiaries initiating treatment by excluding beneficiaries with claims for antihypertensive medication during the look-back period. To evaluate calendar years before and after publication of the JNC8 panel member report, 1 we excluded beneficiaries whose index date was between December 18 and 31, 2013. Using their index date, we categorized beneficiaries into the year that they initiated antihypertensive medication. Because we were interested in studying treatment initiation among beneficiaries without compelling indications for specific antihypertensive medication classes, we restricted the analysis to those without a history of CHD, diabetes, heart failure, chronic kidney disease, or stroke (eTable 2 in the Supplement) 2 by using previously published algorithms (eTable 3 in the Supplement). We also excluded beneficiaries with a history of atrial fibrillation or flutter because β-blockers are indicated for heart-rate control in this population.

Class of Antihypertensive Medication Initiated
We categorized classes of antihypertensive medications as defined in the 2017 American College of Cardiology (ACC)/American Heart Association (AHA) hypertension guideline (eTable 1 in the Supplement). 6 Beneficiaries were categorized as initiating antihypertensive monotherapy or combination therapy.

Beneficiary Race/Ethnicity
We used the Medicare beneficiary summary files to define beneficiaries' race/ethnicity. Medicare obtains data on the race/ethnicity of beneficiaries from the Social Security Administration, which collects voluntarily reported race/ethnicity at the time of application for a Social Security number.
Because there were few Asian (n = 951), Hispanic (n = 909), and "other" (n = 1234) race/ethnicity beneficiaries who met the inclusion criteria, we combined these race/ethnicity groups into a single "other race/ethnicity" category for the analysis.
whether they received a Medicare low-income subsidy during the look-back period and their residence area-level median income (eTable 3 in the Supplement).

Statistical Analysis
We calculated characteristics of Medicare beneficiaries who initiated antihypertensive medication.
Among beneficiaries initiating monotherapy and, separately, combination therapy, we calculated the proportion initiating each antihypertensive class. For beneficiaries initiating antihypertensive combination therapy, we calculated the number and proportion who initiated the 4 most common combinations of dual and triple therapies. All calculations were completed by calendar year and race/ ethnicity. We assessed linear trends across calendar years using linear regression for continuous variables and Poisson regression for categorical variables.
We used Poisson regression with robust standard errors to calculate prevalence ratios (PRs) and 95% CIs for initiating antihypertensive medication with an ACEI or ARB vs other antihypertensive classes, for those initiating monotherapy or combination therapy, seperately. 8 We included ACEIs and ARBs in a single group because the JNC8 panel member report provided differential recommendations for these agents to Black and to non-Black adults. The primary exposure in the model was whether initiation occurred before or after publication of the JNC8 panel member report Black, White, and other race/ethnicity beneficiaries was tested by adding interaction terms for race/ ethnicity by calendar period to the regression model. The above analyses were repeated to calculate the PRs for initiating antihypertensive medication with a β-blocker vs other antihypertensive classes.
A 2-sided P < .05 was considered statistically significant. Statistical analyses were conducted using SAS, version 9.3 (SAS Institute Inc).

Initiation of Antihypertensive Medication Monotherapy
The proportion of Black beneficiaries initiating antihypertensive monotherapy who did so with an

JAMA Network Open | Cardiology
Race/Ethnicity-Specific Changes in Antihypertensive Medication in Medicare Beneficiaries

Initiation of Antihypertensive Medication Combination Therapy
There was no evidence for trends in the proportion of Black beneficiaries or beneficiaries of other race/ethnicity initiating antihypertensive combination therapy who did so with an ACEI or ARB, β-blocker, thiazide-type diuretic, or CCB from 2011 to 2018 (Figure 2; eTable 5 in the Supplement).
The proportion of White beneficiaries initiating combination therapy who did so with an ARB or CCB increased, whereas the proportion initiating treatment with a thiazide-type diuretic decreased. The most common dual antihypertensive combination therapy filled by beneficiaries in each race/ ethnicity group in 2018 was a thiazide-type diuretic with either an ACEI or an ARB (eTable 6 in the Supplement).

Factors Associated With Initiation of an ACEI or an ARB
There was no evidence of a change in the proportion of beneficiaries initiating monotherapy who did so with an ACEI or ARB after vs before publication of the JNC8 panel member report (PR, 1.00; 95% CI, 0.97-1.03) (

Factors Associated With Initiation of a β-Blocker
The proportion of beneficiaries initiating antihypertensive monotherapy who did so with a β-blocker decreased from before to after publication of the JNC8 panel member report (PR, 0.89; 95% CI, 0.84-0.93) ( Table 3). Black beneficiaries were less likely than White beneficiaries to initiate

Discussion
The current study suggests that there was no change in the proportion of Black Medicare beneficiaries initiating an ACEI or ARB as monotherapy in the period after compared with before publication of the JNC8 panel member report. Initiation of CCB monotherapy increased from 2011 to 2018 among both Black beneficiaries and White beneficiaries. However, use of β-blockers as initial monotherapy was high and decreased only modestly. Many older US adults who initiated antihypertensive medication did so with non-guideline-recommended classes of medication.
Although Black Medicare beneficiaries initiating antihypertensive monotherapy were less likely than White beneficiaries to do so with an ACEI or ARB, a high proportion initiated treatment with an ACEI or ARB, and this did not change from before to after publication of the JNC8 panel member report. There are decades of evidence suggesting racial differences in BP-lowering and CVD risk associated with renin-angiotensin-aldosterone system inhibitors. 3,9,10 However, studies examining trends in antihypertensive medication use among US adults since the 2000s have not reported decreases in the use of these agents among Black adults. 11,12 In the current study, more than 20% of Black Medicare beneficiaries without compelling indications for ACEI or ARB monotherapy initiated these classes of antihypertensive medication in 2018.  Although there were no changes from 2011 to 2018 in initiation of monotherapy with either an ACEI or ARB among any race/ethnicity group, there were changes in initiation of these agents when considered separately. Among Black beneficiaries, and separately, White beneficiaries initiating monotherapy, the proportion initiating treatment with an ACEI decreased, whereas the proportion initiating with ARB increased. The decrease in ACEI use and increase in ARB use may be due to evidence of their similar efficacy 13,14 and higher risk of adverse effects associated with ACEIs, 13 especially among Black individuals. 15,16 The proportion of beneficiaries initiating monotherapy who did so with a β-blocker decreased following publication of the JNC8 panel member report. Despite this decrease, in 2018 a high proportion of beneficiaries initiating monotherapy did so with a β-blocker. The JNC8 panel member report and the 2017 ACC/AHA hypertension guideline recommended that β-blockers not be used as initial therapy for hypertension in adults without CHD or without heart failure with reduced ejection fraction. 1,6 This recommendation was based on evidence of higher risks of CVD 5,17,18 and CVD mortality 17 in patients with hypertension treated with a β-blocker compared with other antihypertensive medication classes. In addition, a meta-analysis of randomized clinical trials found that among adults 60 years of age or older, the risk for stroke was higher for β-blockers compared with other classes of antihypertensive medication. 19 The continued prescribing for β-blockers for initial antihypertensive monotherapy represents a missed opportunity to decrease CVD risk among adults with hypertension.
Although initiation of monotherapy with non-guideline-recommended medications remained high at the end of the study period, there was evidence that some prescribing practices may have

JAMA Network Open | Cardiology
Race/Ethnicity-Specific Changes in Antihypertensive Medication in Medicare Beneficiaries changed from before to after publication of the JNC8 panel member report. In the present analysis, fills for CCBs among Black beneficiaries initiating monotherapy increased during the years studied.
This pattern has been present elsewhere. 20 A retrospective study examining changes in initial antihypertensive monotherapy across ethnic groups after publication of the 2006 National Institute for Health and Care Excellence hypertension guideline found that prescribing of CCBs increased among Black adults in the UK. 20 These findings suggest the need for research focused on effective strategies for implementing guideline-recommended prescribing practices.
Findings from the present study suggest that the race/ethnicity-specific recommendation for initial antihypertensive medication included in the JNC8 panel member report was not widely adopted. Lack of guideline adoption is a well-known, 21,22 multifaceted issue involving factors associated with patients, clinicians, health care institutions, and the guidelines themselves. 23,24 Adoption of the race/ethnicity-specific recommendation from the JNC8 panel member report may have been hindered by the controversies that surrounded its publication, 25 including a dissenting opinion by some of the panel members. 26 In addition, the race/ethnicity-specific recommendation was only of moderate strength and certainty (Grade B, moderate recommendation). 1 The 2017 ACC/AHA hypertension guideline also includes the race/ethnicity-specific recommendation for initial antihypertensive medication classes. 6 However, the recommendation is stronger (Class I; level of evidence, B-R). Sufficient data were not available in the current study to assess whether the race/ ethnicity-specific recommendation included in the 2017 ACC/AHA hypertension guideline affected prescribing patterns, but this should be investigated in future studies.

Strengths and Limitations
The present study has several strengths. We used data from a 5% random sample of Medicare beneficiaries. The majority of US adults 65 years of age or older have health insurance through the Medicare program, allowing for high generalizability to older US adults. The use of Medicare claims enabled the identification of the initial antihypertensive medication regimens filled by adults without compelling indications. This study also has limitations. The modest sample sizes of Black beneficiaries and beneficiaries of other race/ethnicity may have limited the ability to detect statistically significant changes in classes of antihypertensive medications initiated over time. We did not have data on BP levels; thus, we could not stratify results by BP levels at the time of medication initiation or assess whether beneficiaries achieved BP control with regimen that they initiated. We assumed the medication claims filled reflected prescribing decisions made by the beneficiaries' clinicians, but it is possible that patients themselves requested non-guideline-recommended therapies. We did not exclude beneficiaries with cognitive impairment from the analysis, an outcome domain that ACEIs and ARBs may affect beneficially. 27 We used Medicare claims to identify pharmacy fills and exclude beneficiaries with a history of compelling indications. Incomplete exclusion of beneficiaries with compelling indications could have biased the results because the general recommendation for initial antihypertensive medication therapy included in the JNC8 panel member report would not apply to those with compelling indications.

Conclusions
From 2011 to 2018, the proportion of Black Medicare beneficiaries initiating monotherapy who did so with an ACEI decreased, and the proportion initiating with an ARB increased. However, there was no evidence that the proportion of Black Medicare beneficiaries initiating monotherapy with either an ACEI or ARB changed from before to after publication of the JNC8 panel member report.
Furthermore, despite a modest decrease in the period following publication of the JNC8 panel member report, a substantial proportion of Medicare beneficiaries who initiated antihypertensive monotherapy did so with a β-blocker. Achieving population-wide BP goals and maximizing the CVD risk decrease associated with antihypertensive medication will require efforts to increase guidelinerecommended choices among US adults initiating antihypertensive medication.