National Trends in Antihypertensive Treatment Among Older Adults by Race and Presence of Comorbidity, 2008 to 2017

Abstract

Background

In 2014, hypertension guidelines for older adults endorsed increased use of fixed-dose combinations, prioritized thiazide diuretics and calcium channel blockers (CCBs) for Black patients, and no longer recommend beta-blockers as first-line therapy.

Objective

To evaluate older adults’ antihypertensive use following guideline changes.

Design

Time series analysis.

Patients

Twenty percent national sample of Medicare Part D beneficiaries aged 66 years and older with hypertension.

Intervention

Eighth Joint National Committee (JNC8) guidelines

Main Measures

Quarterly trends in prevalent and initial antihypertensive use were examined before (2008 to 2013) and after (2014 to 2017) JNC8. Analyses were conducted among all beneficiaries with hypertension, beneficiaries without chronic conditions that might influence antihypertensive selection (hypertension-only cohort), and among Black patients, given race-based guideline recommendations.

Key Results

The number of beneficiaries with hypertension increased from 1,978,494 in 2008 to 2,809,680 in 2017, the proportions using antihypertensives increased from 80.3 to 81.2%, and the proportion using multiple classes and fixed-dose combinations declined (60.8 to 58.1% and 20.7 to 15.1%, respectively, all P<.01). Prior to JNC8, the use of angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, and CCBs was increasing. Use of CCBs as initial therapy increased more rapidly following JNC8 (relative change in quarterly trend 0.15% [95% CI, 0.13–0.18%), especially among Black beneficiaries (relative change 0.44% [95% CI, 0.21–0.68%]). Contrary to guidelines, the use of thiazides and combinations as initial therapy consistently decreased in the hypertension-only cohort (13.8 to 8.3% and 25.1 to 15.7% respectively). By 2017, 65.9% of Black patients in the hypertension-only cohort were initiated on recommended first-line or combination therapy compared to 80.3% of non-Black patients.

Conclusions

Many older adults, particularly Black patients, continue to be initiated on antihypertensive classes not recommended as first-line, indicating opportunities to improve the effectiveness and equity of hypertension care and potentially reduce antihypertensive regimen complexity.

Supplementary Information

The online version contains supplementary material available at 10.1007/s11606-022-07612-3.

INTRODUCTION

Hypertension is the most common chronic disease in the US population, impacting over 108 million people, including two-thirds of older adults.¹ Hypertension remains the most common modifiable risk factor for cardiovascular diseases (CVD), cerebrovascular disease, and renal failure. Older adults with hypertension experience over $2500 higher annual healthcare expenditures compared with those without hypertension.² Pharmacologic treatment of hypertension is highly cost-effective,³ yet more than half of older adults have blood pressure (BP) measurements above recommended goals.⁴

Over the past decade, recommended approaches to treating hypertension have evolved. In 2003, the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation and Treatment of High Blood Pressure (JNC7) recommended universal BP treatment targets below 140/90 mmHg and first-line use of five antihypertensive classes: angiotensin-converting enzyme (ACE) inhibitors, angiotensin II receptor blockers (ARBs), beta blockers, calcium channel blockers (CCBs), and thiazide diuretics.⁵ In December 2013, the Eighth Joint National Committee report (JNC8) relaxed systolic BP treatment goals from 140 to 150 mmHg for most adults over age 60 and made three key changes related to the selection of antihypertensives: no longer recommending beta blockers as first-line treatment for most patients, narrowing the recommended first-line classes for Black patients to thiazide diuretics and CCBs, and recommending the preferred use of fixed-dose combinations for patients on multiple antihypertensives.⁶ These recommendations were designed to decrease the burden of polypharmacy among older adults by narrowing recommendations to the most effective classes and encouraging the use of combination therapy. The 2017 American College of Cardiology/American Heart Association (ACC/AHA) Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults subsequently reinforced the JNC8 recommendations on selection of antihypertensives while lowering target BP goals for most older adults from 150/90 to 130/80 mmHg.⁷

Little is known about recent trends in antihypertensive utilization in the Medicare population, as prior research on patterns of antihypertensive use largely predated the release of JNC8 guidelines^8–11 or were limited by small sample sizes.^12,13 We therefore used data from a national sample of Medicare beneficiaries with Part D pharmaceutical coverage to examine national trends in antihypertensive use from 2008 through 2017. We hypothesized that following the release of the JNC8 guidelines and in accordance with their recommendations, the use of beta blockers would decrease, the use of thiazide diuretics and calcium channel blocker classes would increase in Black patients, and the overall use of fixed-dose combinations would increase.

METHODS

Study Population

We analyzed Medicare claims and enrollment data provided by the Centers for Medicare and Medicaid Services (CMS) for the period 2008–2017. Demographic characteristics were available for all enrollees from the Master Beneficiary Summary File, and data on prescription medication use were available for the 20% random sample of beneficiaries enrolled in Part D prescription drug coverage. Our sample included all persons at least 66 years old continuously enrolled in Medicare Parts A (inpatient hospital services), B (outpatient medical services), and D (prescription drug coverage) with at least 365 days of continuous Part D enrollment prior to the study period. We excluded beneficiaries enrolled in Medicare Advantage plans, for whom diagnoses from claims were not available.

We identified beneficiaries with hypertension using CMS Chronic Condition Warehouse (CCW) algorithm, which indicates whether an enrollee has been diagnosed with hypertension, based upon either a single inpatient stay or at least two ambulatory visits with a hypertension diagnosis over the preceding 2-year period.¹⁴ We also identified beneficiaries with other chronic conditions that may influence antihypertensive class choice including acute myocardial infarction, ischemic heart disease, atrial fibrillation, chronic kidney disease (CKD), diabetes, and heart failure.

Cohort Definitions

As antihypertensive selection is often guided by comorbidities, we analyzed two prespecified cohorts, defining an “overall cohort” which included all patients with hypertension and a “hypertension-only cohort” which excluded patients with cardiovascular comorbidities, CKD, or diabetes, as these comorbidities would be expected to influence antihypertensive choice. Second, as the JNC8 guideline made different first-line recommendations for Black beneficiaries, we created two identically defined sub-cohorts of the hypertension-only cohort, one including Black beneficiaries and one including non-Black beneficiaries.

Antihypertensive Use

We examined prevalent antihypertensive use for each cohort and choice of initial antihypertensive for those who had not filled any antihypertensives in the preceding year.

We grouped antihypertensives into six classes: angiotensin-converting enzyme inhibitors and angiotensin receptor blockers (ACEI/ARBs), beta blockers, CCBs, thiazide diuretics, other antihypertensives, and combinations. ACEIs and ARBs were combined as they share a similar biologic pathway and are frequently used interchangeably, and co-prescription is contraindicated.^6,7 Other antihypertensives included alpha-1 blocking agents, alpha-2 agonists, direct renin inhibitors, loop diuretics, and potassium-sparing diuretics. For analyses of prevalent use, we split combination medications into component classes and in separate analyses examined use of fixed-dose combinations. For analyses of initial antihypertensive selection, we identified combinations as the prescription of a fixed-dose combination medication or claims for 2 or more separate antihypertensive classes filled within a 7-day period.

Statistical Analyses

We calculated characteristics of all beneficiaries with Part D and those in each analytic cohort over time, examining differences using t-tests or chi-square tests as appropriate. We determined the annual proportion of beneficiaries prescribed each antihypertensive class and fixed-dose combinations for each cohort. We repeated this approach to examine choice of initial antihypertensive agents.

We conducted longitudinal analyses using a time-series framework. We estimated ordinary least-squares regressions with Newey-West standard errors to account for autocorrelation and included an indicator variable for quarter and the release of JNC8 as well as an interaction between the two.¹⁵ The pre-intervention (2008 Quarter 1 to 2013 Quarter 4) and post-intervention (2014 Quarter 1 to 2017 Quarter 4) periods were defined relative to release of JNC8 in December 2013.⁵ Model covariates included age, sex, race, Medicaid dual-eligible status (as low-income Medicaid beneficiaries may receive additional prescription drug subsidies), census region, and comorbidities (coronary artery disease, heart failure, atrial fibrillation, stroke/TIA, diabetes, CKD, and dementia). Pre- and post-intervention trends (slopes), the immediate level change in 2014 Quarter 1, and the difference in slopes between the pre- and post-intervention period are reported. We used the same analytic approach to examine choice of initial antihypertensive for those who had not received any antihypertensive prescriptions in the preceding 365 days. All confidence intervals are presented at 95% confidence level.

All analyses were conducted using SAS, version 9.2. The study protocol was approved by the Harvard Medical School Human Studies Committee and the CMS Privacy Board.

RESULTS

The proportion of Medicare Part D beneficiaries diagnosed with hypertension decreased from 82.4% in 2008 to 80.7% in 2017, while the absolute number increased from 1,978,494 to 2,912,003 due to increasing Medicare Part D enrollment (Table 1 and Supplemental Table S1). Over the same period, decreasing proportions of the overall hypertension cohort were female (65.8 to 59.0%), age 85 years or older (21.0 to 18.4%), and Black (9.1 to 8.0%). The proportion of beneficiaries with hypertension and no relevant comorbidities was stable at 21.5%.

Table 1.

Characteristics of Medicare Part D Enrollees with Hypertension, 2008 to 2017

	Overall hypertension cohort		Hypertension-only cohort		Black hypertension-only cohort		Non-Black hypertension-only cohort
	2008	2017	2008	2017	2008	2017	2008	2017
No.	1,978,494	2,912,003	425,311	626,675	30,182	37,392	395,129	589,283
Demographics, %
Female	65.8	59.0	72.0	66.2	71.7	67.6	72.1	66.1
Age category
66–69	17.6	19.7	27.6	32.3	31.2	41.5	27.4	31.7
70–74	22.9	25.7	27.3	32.3	27.4	29.7	27.3	32.5
75–79	20.4	21.1	19.0	18.8	17.4	15.3	19.1	19.0
80–84	18.2	15.2	13.9	9.3	12.3	7.4	14.0	9.4
85+	21.0	18.4	12.2	7.4	11.7	6.1	12.2	7.5
Race/ethnicity
White	80.0	81.0	84.2	84.4	0	0	90.7	89.8
Black	9.1	8.0	7.1	6.0	100	100	0.0	0.0
Hispanic	6.6	5.5	4.6	3.9	0	0	4.9	4.1
Asian	3.2	3.1	3.1	2.7	0	0	3.4	2.9
Other	1.1	2.3	1.0	3.0	0	0	1.1	3.2
Dual eligible	32.6	20.8	20.6	12.0	50.5	32.4	18.3	10.7
Region
Northeast	19.1	19.6	16.1	18.2	11.6	13.7	16.5	18.5
Midwest	24.4	23.5	26.8	24.9	15.9	18.0	27.6	25.4
South	40.2	39.2	39.5	37.8	67.2	62.1	37.4	36.3
West	16.4	17.7	17.6	19.0	5.3	6.1	18.5	19.9
Comorbidities, %
Coronary artery disease a	58.7	54.4	0	0	0	0	0	0
Heart failure	37.1	31.0	0	0	0	0	0	0
Atrial fibrillation	17.9	19.2	0	0	0	0	0	0
Diabetes	41.2	44.8	0	0	0	0	0	0
Chronic kidney disease	23.2	40.6	0	0	0	0	0	0
Stroke/TIA	20.0	17.8	8.2	6.2	8.9	6.3	8.1	6.2
Dementia	20.6	18.8	9.6	6.7	12.2	7.7	9.4	6.7
No relevant comorbidityb	21.5	21.5	100	100	100	100	100	100

Abbreviations: TIA, transient ischemic attack

Note: Demographics presented for Quarter 1 of 2008 and Quarter 1 of 2017, respectively. Supplemental Table 1 contains demographic information for all years

^aComposite of CCW Chronic Conditions categories of ischemic heart disease and acute myocardial infarction

^bHypertension relevant comorbidities include ischemic heart disease, acute myocardial infarction, heart failure, atrial fibrillation, diabetes, and chronic kidney disease

Prevalent Antihypertensive Use

Between 2008 and 2017, the proportions of beneficiaries using any antihypertensives increased from 80.3 to 81.2% while the proportions using multiple classes and fixed-dose combinations declined (60.8 to 58.1%, and 20.7 to 15.1%, respectively, all P<0.01) (Table 2). Throughout the study period, use of antihypertensives was lower in the cohort of Black beneficiaries with hypertension-only compared to non-Black beneficiaries with hypertension only, though both cohorts demonstrated an increase in monotherapy over time (Black cohort from 75.1 to 77.3% and non-Black cohort from 78.2 to 78.9%). The use of fixed-dose combinations declined in all cohorts over the study period.

Table 2.

Proportion of Medicare Beneficiaries with Hypertension Receiving Antihypertensives by Number of Antihypertensive Classes Used, 2008 to 2017

	Calendar year, % of beneficiaries										% change
	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	Overall	Post-JNC8
Overall cohort
0 antihypertensives	19.7	19.2	19.4	19.0	19.0	18.6	19.4	19.2	18.8	18.8	−0.9	−0.7
1 antihypertensives	19.5	19.8	20.0	20.4	20.8	21.2	21.5	22.2	22.7	23.2	3.6	1.7
2 antihypertensives	26.0	26.2	26.3	26.6	26.7	27.1	27.1	27.3	27.5	27.6	1.6	0.5
3 antihypertensives	21.3	21.4	21.3	21.2	21.1	21.0	20.7	20.5	20.3	20.1	−1.2	−0.5
>4 antihypertensives	13.4	13.4	13.1	12.8	12.4	12.0	11.3	10.9	10.6	10.3	−3.1	−1.0
>1 combination	20.7	20.0	19.3	18.5	17.6	17.2	16.8	16.1	15.6	15.1	−5.5	−1.7
Hypertension-only cohort
0 antihypertensives	22.0	22.1	22.4	22.2	22.3	21.6	22.1	21.5	21.3	21.2	−0.8	−0.8
1 antihypertensives	25.1	25.4	25.8	26.6	27.2	27.9	28.5	29.5	30.3	30.9	5.8	2.4
2 antihypertensives	29.2	29.2	29.1	29.2	29.3	29.6	29.5	29.5	29.5	29.5	0.2	0.0
3 antihypertensives	16.8	16.5	16.2	15.9	15.5	15.4	14.9	14.7	14.3	14.0	−2.8	−0.9
> 4 antihypertensives	6.9	6.7	6.5	6.2	5.8	5.6	5.1	4.8	4.7	4.5	−2.5	−0.6
> 1 combination	25.4	24.5	23.5	22.5	21.6	21.2	20.8	20.1	19.5	18.9	−6.5	−1.9
Black hypertension-only cohort
0 antihypertensives	24.9	23.4	24.3	23.3	23.3	23.0	23.9	23.0	22.7	22.7	−2.2	−1.2
1 antihypertensives	17.3	17.7	17.4	18.0	18.8	18.9	19.2	20.1	21.0	21.6	4.3	2.4
2 antihypertensives	27.4	28.1	28.2	28.3	28.5	29.3	28.9	29.4	29.5	29.7	2.3	0.8
3 antihypertensives	19.9	20.2	20.1	20.5	20.0	19.4	19.3	19.3	18.7	18.4	−1.5	−0.9
>4 antihypertensives	10.5	10.5	9.9	10.0	9.4	9.4	8.7	8.3	8.1	7.6	−2.9	−1.1
>1 combination	30.1	29.8	29.2	28.5	27.0	27.4	27.1	26.7	25.9	24.8	−5.3	−2.2
Non-Black hypertension-only cohort
0 antihypertensives	21.8	22.0	22.3	22.1	22.2	21.5	21.9	21.4	21.2	21.1	−0.6	−0.8
1 antihypertensives	25.6	26.0	26.4	27.2	27.7	28.5	29.1	30.1	30.9	31.5	5.8	2.4
2 antihypertensives	29.4	29.3	29.2	29.3	29.3	29.6	29.5	29.5	29.5	29.5	0.1	−0.1
3 antihypertensives	16.6	16.2	15.9	15.6	15.2	15.1	14.6	14.4	14.0	13.7	−2.9	−0.9
>4 antihypertensives	6.7	6.5	6.2	5.9	5.5	5.3	4.9	4.6	4.4	4.3	−2.4	−0.6
>1 combination	25.0	24.1	23.1	22.1	21.2	20.8	20.4	19.7	19.1	18.5	−6.5	−1.9

Note: Data shown for quarter 1 of each year. Combination refers to use of fixed-dose combination medications. Post-JNC8 time period refers to change from 2014 to 2017

In analyses adjusting for changes in beneficiary demographic characteristics during the study period, the most common antihypertensive classes filled in 2008 were ACEI/ARBs (47.1%), beta blockers (40.4%), thiazide diuretics (28.9%), and CCBs (28.6%) (Fig. 1A). By 2017, the proportion of beneficiaries using ACEI/ARBs increased by 4.2%, while the proportion using all other classes decreased with the largest declines for thiazide diuretics (6.4%) and beta blockers (4.4%). In general, time-series analyses show that trends in prevalent medication use, including increasing use of ACEI/ARBs and decreasing use of beta blockers and thiazide diuretics changed minimally following the release of JNC8 (Supplemental Table S2).

Figure 1.

Trends in prevalent antihypertensive use, 2008 to 2017. Abbreviations: ACE/ARB, angiotensin-converting enzyme inhibitors and angiotensin receptor blockers; CCB, calcium channel blockers. Note: Hypertension-only cohorts include beneficiaries without a history of acute myocardial infarction, ischemic heart disease, atrial fibrillation, chronic kidney disease, diabetes, or heart failure. Proportions amount to greater than 100% due to concurrent use of multiple antihypertensives. A Change in trend comparing pre-JNC8 period to post-JNC8 period: P=.14 for thiazides, P<.001 for ACE/ARBs, beta blockers, CCBs, and other classes. B Change in trend comparing pre-JNC8 period to post-JNC8 period: P=.003 for ACE/ARBs, P=.001 for CCBs, P<.001 for beta blockers, thiazide diuretics, and other classes.

In the hypertension-only cohort, use of ACEI/ARBs increased by 5.9% while use of thiazide diuretics and beta blockers declined by 5.8% and 3.6%, respectively, during the study period (Fig. 1B). Time-series analyses indicate that in the post-JNC8 period there were very small guideline-concordant changes in prescribing trends. Specifically, the rate of increase of use of CCBs modestly increased (0.03% change in quarterly trend in post-JNC8 period relative to pre-JNC8 period, CI 0.01 to 0.05%), the trend of decreasing use of beta blockers was modestly greater (−0.04% change in quarterly trend, CI −0.05 to −0.02%), and the trend of decreasing thiazide use was modestly attenuated (0.05% change in quarterly trend, CI 0.03 to 0.07%) (Supplemental Table S2).

There was greater baseline use of CCBs and thiazides in the Black compared to the non-Black hypertension-only cohorts (Fig. 2). In the Black hypertension-only cohort, there was an increase in the use of CCBs (0.11% change in quarterly trend, CI 0.04 to 0.17%), no change in the pre-JNC8 trend of decreasing thiazide use (0.01% change in quarterly trend, CI −0.06 to 0.08%), and a small decrease in the rate of decline in the use of beta blockers (−0.07% change in quarterly trend, CI −0.13 to −0.02%) following the release of JNC8. Trends in the non-Black hypertension-only cohort largely matched the hypertension-only cohort.

Figure 2.

Trends in prevalent antihypertensive use by race, 2008 to 2017. Abbreviations: ACE/ARB, angiotensin-converting enzyme inhibitors and angiotensin receptor blockers; CCB, calcium channel blockers. Note: Hypertension-only cohorts include beneficiaries without a history of acute myocardial infarction, ischemic heart disease, atrial fibrillation, chronic kidney disease, diabetes, or heart failure. Proportions amount to greater than 100% due to concurrent use of multiple antihypertensives. A Change in trend comparing pre-JNC8 period to post-JNC8 period: P<.001 for ACE/ARBs, P=.01 for beta blockers, P=.003 for CCBs, P=.29 for thiazide diuretics, P=.70 for other classes. B Change in trend comparing pre-JNC8 period to post-JNC8 period: P=.004 for CCBs, P<.001 for ACE/ARBs, beta blockers, thiazide diuretics, and other classes.

Choice of Initial Antihypertensive

In the overall cohort, there was a continuous rise in the use of ACEI/ARBs and CCBs as first antihypertensives prescribed and decline in the use of beta blockers, thiazide diuretics, and combination medications (Fig. 3A). In the post-JNC8 period, there were both guideline-concordant and discordant changes in first antihypertensive prescribing trends. The slope of increasing use of CCBs increased (0.15% change in quarterly trend, CI 0.13 to 0.18%), and the trend of decreasing thiazide use was slightly attenuated (0.04% change in quarterly trend, CI 0.02 to 0.06%), while there was no change in the steady decline of beta blocker use, which predated JNC8 (Supplemental Table S3). In contrast to guideline recommendations, there was steeper decline in the use of combination therapy in the post-JNC8 period (−0.18% change in quarterly trend, CI −0.21 to −0.15%). Among beneficiaries who were initiated on combination therapy, the most common component classes of combinations were ACE/ARBs and thiazide diuretics (Supplemental Figure S1).

Figure 3.

Trends in choice of initial antihypertensive, 2008 to 2017. Abbreviations: ACE/ARB, angiotensin-converting enzyme inhibitors and angiotensin receptor blockers; CCB, calcium channel blockers. Note: Hypertension-only cohorts include beneficiaries without a history of acute myocardial infarction, ischemic heart disease, atrial fibrillation, chronic kidney disease, diabetes, or heart failure. Combinations include both fixed-dose combinations and separate antihypertensive classes initiated within the same seven day period. A Change in trend comparing pre-JNC8 period to post-JNC8 period: P=.28 for ACE/ARBs, P=.10 for beta blockers, P<.001 for CCBs, P<.001 for thiazide diuretics, P=.009 for other classes, P<.001 for combinations. B Change in trend comparing pre-JNC8 period to post-JNC8 period: P=.05 for ACE/ARBs, P=.008 for beta blockers, P<.001 for CCBs, P=.06 for thiazide diuretics, P=.17 for other classes, P<.001 for combinations.

Among beneficiaries in the hypertension-only cohort, ACEI/ARBs were the most commonly prescribed first antihypertensive and their use grew throughout the study period, from 28.9 to 39.2% (Fig. 3B). Combination medication use declined continuously, from 25.1 to 15.7%, with a steeper decline following JNC8 (−0.24% change in quarterly trend, CI −0.30 to −0.18%). Beta blocker use declined throughout the study period, from 16.7 to 11.5%, with a slightly steeper decline following JNC8 (−0.07% change in quarterly trend, CI −0.12 to −0.02%). Despite no longer being recommended as first-line, use of beta blockers and other classes as first antihypertensives remained common in 2017, with 10.1% and 6.3% of beneficiaries in the hypertension-only cohort started on these classes.

In the Black hypertension-only cohort (Fig. 4A), combination medications accounted for 41.9% of first prescribed antihypertensives in 2008, declining to 22.1% in 2017. Of the two classes identified as first line for Black patients by JNC8, there was a substantial increase in the use of CCBs (11.4 to 32.4%) but a decline in use of thiazide diuretics (16.5 to 11.4%). The trend of increasing CCBs use predated JNC8 but more than doubled in the post-JNC8 period (0.44% change in quarterly trend, CI 0.21 to 0.68%).

Figure 4.

Trends in choice of initial antihypertensive by race, 2008 to 2017. Abbreviations: ACE/ARB, angiotensin-converting enzyme inhibitors and angiotensin receptor blockers; CCB, calcium channel blockers. Note: Hypertension-only cohorts include beneficiaries without a history of acute myocardial infarction, ischemic heart disease, atrial fibrillation, chronic kidney disease, diabetes, or heart failure. Combinations include both fixed-dose combinations and separate antihypertensive classes initiated within the same 7-day period. A Change in trend comparing pre-JNC8 period to post-JNC8 period: P=.06 for ACE/ARBs, P=.99 for beta blockers, P<.001 for CCBs, P=.75 for thiazide diuretics, P=.048 for other classes, P=.004 for combinations. B Change in trend comparing pre-JNC8 period to post-JNC8 period: P=.02 for ACE/ARBs, P=.006 for beta blockers, P<.001 for CCBs, P=.06 for thiazide diuretics, P=.42 for other classes, P<.001 for combinations.

Despite increased use of CCBs among Black beneficiaries, only 65.9% of the Black hypertension-only cohort were initiated on recommended first-line or combination therapy in 2017. In contrast, 80.3% of the non-Black hypertension-only cohort received recommended first-line or combination therapy (Fig. 4B).

DISCUSSION

Our analysis of trends in antihypertensive prescribing for older Medicare beneficiaries shows that antihypertension medication selection evolved between 2008 and 2017, but that these changes largely predated the release of JNC8 recommendations. Prevalent and incident use of ACEI/ARBs and CCBs have increased, consistent with their classification as first-line options. Antihypertensive selection differed by the presence of cardiovascular comorbidities, but was largely consistent with guidelines on antihypertensive class selection. Contrary to guidelines, beta blockers continue to be prevalent with only a modest decrease during the study period and there has been a guideline-discordant decline in the use of fixed-dose combinations and in initiating patients on combination therapy. Antihypertensive selection also differed by race, with a lower proportion of Black patients initiated on guideline-reccomended first-line therapies. These findings suggest additional opportunities to improve care provided to older adults with hypertension.

By 2017, 80% of the hypertension-only cohort were initiated first-line antihypertensive therapy, a higher rate of guideline adherent treatment selection than has been observed for diabetes, heart failure, or statin use.^16–18 Although trends in selection among antihypertensive classes were consistent with JNC8 recommendations, these trends largely predated JNC8. These results are consistent with clinicians changing practice in response to the clinical trials underlying guideline recommendations. For example, beta blockers were found to have inferior stroke and mortality protection compared to other antihypertensives classes more than 10 years prior to JNC8.^19,20 The observed growth in CCB use and decline in thiazide use, despite the continued recommendation of both as a first-line therapy may similarly reflect a response to clinical trials.²¹

Our findings with respect to fixed-dose combinations were unexpected. One prior study documented a lower baseline use of fixed-dose combinations in the Medicare Part D population and a decline between 2009 and 2014, predating JNC8.¹² One-third of older adults with hypertension take three or more antihypertensives and fixed-dose combinations have been promoted as a strategy to both decrease polypharmacy and improve hypertension control rates, as endorsed by JNC8 and subsequent hypertension guidelines given evidence of improved adherence.^22,23 Reasons for low uptake may include a limited selection of classes, clinicians’ inability to titrate medications individually, and cost. Though branded fixed-dose combinations are expensive,²⁴ a recent analysis found that 90% are available in generic form;²³ hence, patient cost is unlikely to explain persistent declines in their use. Nonetheless, the need for cautious and individualized prescribing remains, particularly for older patients with multimorbidity for whom adverse drug events and prescribing cascades, including those resulting from antihypertensives, are common.^25,26

Use of beta blockers among patients without heart failure or cardiovascular disease continues to be common despite strong evidence of better outcomes with other first-line classes. Clinicians likely had patients who had been tolerating these medications well and chose not to take them off of them or change classes. Though such clinical inertia might make sense given the primary focus of hypertension guidelines on BP goals, the relative ineffectiveness of beta blockers for patients without cardiovascular comorbidities likely results in the need for more medications to achieve goal BPs and contributes to polypharmacy burden.

We found different prescribing patterns among Black beneficiaries as compared to non-Black beneficiaries. Black beneficiaries were more likely to be initiated on CCBs and thiazide diuretics than non-Black beneficiaries, though the use of thiazide diuretics declined in both groups. These results are consistent with JNC8 and ACC/AHA recommendations, which were based on two earlier randomized clinical trials comparing antihypertensive monotherapies.^27–29

Despite the increased use of CCBs among Black beneficiaries, only 65.9% of Black beneficiaries without compelling comorbidities for other classes were initiated on guideline-directed first-line or combination therapy in 2017 compared to 80.3% of the overall hypertension-only cohort. This lower proportion is largely driven by race-based differences in first-line class recommendations for Black patients. If the same recommendations were applied for all races, 83.7% of Black beneficiaries would have been initiated on first-line therapy. Numerous prior studies have characterized racial inequities in hypertension outcomes, with Black patients having lower rates of reaching antihypertensive treatment targets and higher rates of death due to cardiovascular disease.^4,30,31 Our findings indicate that lower rates of prescribing first-line antihypertensive classes to Black patients may be one mediator of these disparities, along with structural and economic barriers to accessing high-quality healthcare.³¹ Lower rates of first-line antihypertensive prescribing to Black patients may be a result of clinicians being unaware of, or choosing not to follow, race-based guideline recommendations, may reflect clinical inertia, or may reflect patient and clinician concerns around changing current treatments in response to race-based recommendations. Increasing attention to race as a dynamic socio-economic-cultural construct and not a biologic construct has led to reconsideration of the appropriateness of clinical algorithms and guidelines that include race. Our study and others³² indicate that though race-based hypertension guidelines are likely to influence clinical prescribing, there is no evidence that these recommendations have led to narrowing of disparities in blood pressure control.^4,31 Thus, future guidelines should reevaluate these recommendations and critically appraise whether maintaining them is consistent with what we understand about the biology of race-based differences.³³

Our study has several potential limitations. First, our analyses were limited to administrative claims, so we relied on diagnosis codes to identify older adults with hypertension and other comorbidities which may lead to under-estimates of certain chronic disease. Second, our study focused on a sample of Medicare enrollees who also had Part D drug coverage; older adults without Part D prescription drug benefit may have different patterns of antihypertensive use. Third, antihypertensives may be used for indications other than hypertension and pharmacy claims do not include information on medication indication. For this reason, we performed analyses stratifying beneficiaries by the presence and absence of comorbidities that commonly influence the choice of antihypertensive classes. Fourth, we were not able to identify whether some beneficiaries identified in incident use cohorts may have used antihypertensives more than 365 days in the past, and these patients’ class choices may have been driven in part by past medication experiences.

In conclusion, we found substantial changes in antihypertensive use among Medicare beneficiaries with hypertension between 2008 and 2017, particularly greater use of first-line ACEI/ARBs and CCBs. Opportunities remain to improve treatment regimens by reducing the use of second-line classes as initial therapy and increasing the use of fixed-dose combinations, which may reduce pill burden and regimen complexity for older adults. JNC8 guidelines did not appear to have had a substantial impact on antihypertensive prescribing, indicating that overcoming clinical inertia to further improve hypertension treatment will require population-level efforts to translate guidelines into practice.

Author Contribution

Dr. Landon and Mr. Souza had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis as the guarantor. Drs. Anderson and Landon attest that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted.

Study concept and design: Anderson, Landon. Acquisition, analysis, or interpretation of data: All the authors. Drafting of the manuscript: Anderson. Critical revision of the manuscript for important intellectual content: All the authors. Statistical analysis: Anderson, Souza, Zaslavsky. Obtained funding: Ayanian, Zaslavsky, Landon. Administrative, technical, or material support: Ayanian, Zaslavsky, Landon. Study supervision: Landon.

Funding

Research reported in this publication was supported by National Institute of Aging of the National Institutes of Health under award number P01AG032952. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Dr. Anderson was supported by grants from the National Institute on Aging (L30AG060493 and R03AG064373) and American College of Cardiology.

Declarations

Conflict of Interests

Dr. Anderson reports receiving honoraria from Alosa Health, a nonprofit educational organization with no relationship to any drug or device manufacturers, related to deprescribing education.