Sustainability of Blood Pressure Reduction in Black Barbershops

Abstract

Purpose of Review

The prevalence of hypertension (HTN) among non-Hispanic (NH) blacks increased from 41% to 55% with the release of the new 2017 ACC/AHA guidelines – the highest among any racial group. NH black men have less physician interaction and lower blood pressure (BP) treatment and control rates when compared to their female counterparts, necessitating community outreach. Here we review the LA Barbershop Blood Pressure Study (LABBPS) which demonstrated a community-based approach involving pharmacists, physicians, and barbers could improve BP control rates among black men.

Recent Findings

LABBPS was a cluster-randomized trial that evaluated both the efficacy and sustainability of a pharmacist-led HTN management program in which barbers promoted follow-up with pharmacists who prescribed antihypertensive therapy under collaborative practice agreements with intervention participant’s primary care providers. After 6 months researchers observed a 21 mmHg greater fall in systolic BP among intervention group participants when compared to the control group participants who received ‘usual care.’ The 6-month extension phase of the study showed that the impressive BP reduction achieved was sustained with less pharmacist contact.

Summary

Multi-disciplinary, community-based approaches to HTN management can be effective and are necessary to tackle the current disparity seen in BP control rates. The model developed in LABBPS represents one such approach.

INTRODUCTION

The 2017 ACC/AHA hypertension (HTN) guidelines lowered blood pressure (BP) thresholds for the diagnosis of HTN resulting in an additional 31 million Americans being classified as hypertensive.¹ The numbers were particularly bleak for non-Hispanic (NH) blacks who saw the prevalence of HTN increase from 41% to 55% - the highest prevalence of any racial group.¹ For black men, who have less physician interaction than their female counterparts and thus lower rates of treatment and control, the estimated prevalence is as high as 59%.^1–4 These discouraging statistics explain, at least in part, the higher cardiovascular risk and much lower life expectancy of black men when compared to other racial/ethnic groups.⁵

Given that traditional healthcare systems have largely failed to achieve HTN control among the majority of NH black men the development of novel, culturally appropriate diagnosis and treatment models will be “critical to stem the shamefully disparate burden of HTN-related death and disability among blacks.”⁶

The legacy of mistrust in healthcare systems and research that exists in the black community has been fueled by racial discrimination, most notably in the Tuskegee experiments,^7,8 and by discordant medical treatment.⁹ One proven way to build trust is by offering community-based and culturally appropriate health interventions in a non-medical setting. Traditionally, churches have been a venue where blacks could be successfully engaged. However, attendees are more often women and not men and health interventions have typically been limited to educational, diagnosis and referral programs, but not medical treatment.^10,11

Prior attempts to reduce cardiovascular risk of black adults in barbershops

Barbershops are a unique venue for health interventions due to several unique key attributes: 1) they are a place for social and cultural exchange among black men, 2) barbershop customers tend to be loyal customers with patronage that not infrequently lasts for decades, 3) they are frequented often enough—oftentimes weekly or bi-weekly—to allow effective follow up, 4) barbers are trusted community members whose health messages are not only heard but may also have a deeper impact than those given by medical providers.

The use of barbershops as place for health promotion and healthcare delivery dates back several hundred years, when barber-surgeons provided services such as lice removal, tooth-extraction and blood-letting. Almost 40 years ago, the late Dr. Elijah Saunders first introduced the concept of screening barbershop patrons for HTN in Baltimore, Maryland.¹² In recent years, a number of successful programs in the areas of HTN, prostate cancer and HIV have followed.^10,13–18 The Healthy Heart Community Prevention Project¹⁰ in New Orleans, Louisiana was an educational program, in which barbers and beauticians were taught about the importance of HTN and hyperlipidemia as cardiovascular risk factors and how to measure BP in their shops. Other components of this program included educational sessions in churches and clinical presentations for the medical and lay community. The program screened over 1300 people for HTN, educated several hundred, and referred many to their doctors to receive care. However, actual data on the BP-lowering effect of these interventions are missing.

The importance of quantifying the effects of community-based HTN programs to determine their success was underscored by Dr. Ronald G. Victor’s first barbershop study in Dallas, Texas.¹⁸ In this cluster-randomized trial, barbers from 9 active-intervention barbershops offered BP checks with haircuts and motivated black male patrons with elevated BP to pursue physician follow-up while patrons in 8 comparison shops received HTN pamphlets but not barber-BP checks. The overall intervention effect was small (−2.5 mm Hg [95% CI, −5.3 to 0.3 mm Hg]). Patrons from intervention barbershops who were referred back to their primary care physicians saw on average a BP reduction of 4.4 mmHg, which did not reach statistical significance when compared to the BP reduction observed in control barbershops. A posthoc analysis of the trial showed that physician inertia was a major obstacle for the success of the program.⁴ Posthoc analysis also revealed the importance of focused specialty care when large BP reductions (21 mmHg) were observed in patrons who were referred to a HTN-specialist clinic.⁴ With the lack of hard outcome data from prior barbershop interventions and the clinically non-significant results of the Dallas Barbershop Study, rigorous evidence for the benefit of barbershop health interventions has been lacking prior to LABBPS.

The Los Angeles Barbershop Blood Pressure Study¹⁹

LABBPS was, like the Dallas Barbershop Study, a cluster-randomized trial, where barbershops not individual participants were randomized to avoid group contamination. A cohort of 319 black male patrons from 52 Los Angeles County barbershops were enrolled and shops were randomly assigned to either a pharmacist-led intervention or an active control group. In the intervention group, barbers promoted follow-up with pharmacists who prescribed BP medication under a collaborative practice agreement (CPA) with patrons’ primary care providers (PCP). In the control group, barbers promoted follow-up with primary care providers and lifestyle modification.

All participants were self-identified NH blacks, between the ages of 35–79. They were also regular patrons of participating barbershops (≥1 haircut every 6 weeks for ≥6 months) and had systolic BP ≥140 mm Hg on 2 screening days at baseline. Women, men who planned to relocate, and those on dialysis or chemotherapy were main exclusion criteria.

The trial’s primary outcome was the change in systolic BP at 6 months and 12 months in both groups. Secondary outcomes included the change in diastolic pressure, BP control rates (<130/80 mmHg), number of antihypertensive drugs prescribed, and adverse drug reactions.

At baseline the groups were well-matched with respect to age, marital status, education level, household income, duration of barbershop patronage, healthcare access and cardiovascular risk. Of note, the participants in this trial were largely well-educated (65% with some college or higher degree), yet of relatively low-income (53% at or below 300% federal poverty level), and the majority had existing healthcare access (~77% had a primary care provider and ~86% had health insurance).

At baseline systolic BP was similar in the intervention and control groups (152.8 vs. 154.6 mmHg). At 6 months the mean intervention effect on systolic BP was −27±14 mmHg, and at 12 months it was −29±13 mmHg (both p<0.0001; Table 1). By study end a systolic BP of 124 mmHg was achieved in the intervention group compared to a systolic BP of 147 mmHg observed in the control group. Change in diastolic BP was also greater in the intervention group than in the control group (−15 mmHg, 95% CI, 9.5−19.5 mm Hg, p<0.0001; Table 1). Blood pressure reduction was consistent across barbershop clusters, both for the 6- and 12-month results (Figure 1).

Table 1.

Primary and Secondary Blood Pressure Outcomes at 12 months^*

	Intervention, N = 125	Control, N =163	Intervention Effect
Blood Pressure			Difference in Mean Change of BP (95% CI)	p-value†
Systolic Blood Pressure - mm Hg ‡
Baseline	152.4 ± 10.1	154.6 ± 12.0
12-months	123.8 ± 8.8	147.4 ± 15.7
Change	−28.6 ± 12.7	−7.2 ± 17.7	−20.8 (−27.7, −13.9)	<0.0001
Diastolic Blood Pressure - mm Hg
Baseline	91.9 ± 11.3	89.8 ± 11.3
12-month	74.1 ± 8.2	86.5 ± 12.6
Change	−17.8 ± 11.9	−3.3 ± 11.2	−14.5 (−19.5, −9.5)	<0.0001
Hypertension Control Rate after 12 months - no. (%)			Odds Ratio (95% CI)	p-value§
Blood Pressure <140/90 mm Hg	118 (94.4%)	47 (28.8%)	3.3 (1.8, 6.1)	0.0001
Blood Pressure < 135/85 mm Hg	110 (88.0%)	24 (14.7%)	6.7 (2.3, 18.9)	0.0004
Blood Pressure <130/80 mm Hg	85 (68.0%)	18 (11.0%)	9.1 (1.5, 56.6)	0.0177

BP, blood pressure; CI, confidence interval; SD, standard deviation; SBP, systolic blood pressure

^*Plus-minus values are means ± SD

^†p-values calculated from linear mixed effects models with random intercepts for clusters. The estimated intervention effect was controlled for baseline systolic blood pressure (or diastolic), primary care provider, and high cholesterol.

^‡Pre-specified primary outcome. Intraclass Correlation Coefficient from the linear mixed effects model for change in SBP is 0.01.

^§p-values calculated from generalized estimating equations with a compound symmetry working correlation to account for cluster effects. The estimated intervention effect was controlled for baseline SBP, primary care provider, and high cholesterol.

Figure 1. Systolic and Diastolic Blood Pressure at Baseline and 12 months According to Barbershop Cluster.

Shown are box plots for systolic and diastolic blood pressure according to barbershop cluster. The horizontal line inside each box indicates the median, the diamond indicates the mean, and the bottom and top of each box indicate the 25th percentile and 75th percentile, respectively. I bars indicate the upper adjacent value (75th percentile plus 1.5 times the interquartile range) and the lower adjacent value (25th percentile minus 1.5 times the interquartile range), and the circles outliers.

Nearly 70% of participants in the intervention group met the prespecified BP goal of <130/80 (the recommended BP goal of the 2017 ACC/AHA guidelines), while only 11% of participants met that goal in the control group (Table 1).

Trial Takeaways

The trial is most notable for the large intervention effect observed in this difficult to reach, high-risk population of middle to older aged black men. The 21 mmHg systolic BP difference observed between groups is close to 3 times greater than the 7 mmHg difference seen in most other pharmacist-led intervention trials^20–22 and an order of magnitude larger than the results of the Dallas Barbershop Study.¹⁸ Importantly, the results were sustained and even further solidified with less pharmacist interactions between 6-and 12 months.

HTN-specialty trained, and American Society of Hypertension certified pharmacists were able to titrate antihypertensive medications safely and effectively in a convenient setting that facilitated frequent follow-up. The social nature of the barbershop environment also helped educate participants and non-participants alike on the importance of BP control and medication adherence.

Overcoming physician inertia, which blunted the success of the Dallas Barbershop Study, was key and underscores the importance of specialty-care with a clear focus on meeting guideline-recommended BP goals.

Another important difference between study groups was not only the quantity of antihypertensive medications that were prescribed in the trial but notably the use of first-line medications, that are user friendly with once or twice daily dosing and an advantageous side effect profile. Long-acting calcium channel blockers (i.e., amlodipine) and angiotensin receptor blockers were initiated first, followed by a long-acting thiazide diuretic (i.e., indapamide) or an aldosterone blocker. Central sympatholytics and non-vasodilating beta blockers, which have unfavorite side effect profiles and high discontinuation rates^23,24 were largely avoided to ensure adherence to the prescribed regimen.

Generalizability of Results and Portability of Model

The benefits of BP-lowering in preventing a wide spectrum of CVD in hypertensive persons are well-established. However, the potential public health implications of the impressive BP reductions seen in LABBPS are dependent on the generalizability of the results to the rest of the U.S black male population and the portability of the model.

While the cohort studied in the trial was relatively diverse, there are a few noteworthy characteristics that might limit applicability of the results to the population at-large. First, the location of the trial in a large urban city like Los Angeles calls into question whether a similar initiative would work in a small rural area where presumably access to healthcare, social services, employment, and education may be limited. Cultural and lifestyle differences are also apparent in dense urban areas versus rural communities. Recently the effects of the Southern Diet on HTN rates have received significant attention and may limit generalizability to blacks living in that portion of the country.²⁵ Finally, while the LABBPS cohort was largely low-income, the vast majority had some form of health insurance and a regular PCP access thanks to California’s MediCal program. This may be contrasted with states in which there has been no Medicaid expansion and access to care remains limited to those of lower economic means.

If we assume these factors do not negate the applicability of the results to the greater NH African American population, we must ensure the model developed herein is portable/reproduceable. Perhaps the most critical first step toward widespread dissemination of our model is the expansion of collaborative practice between pharmacists and physicians or the elimination of the requirement altogether (as in Canada and the United Kingdom).²⁶ While 48 states allow some form of collaborative practice most do not allow the broad authorities granted to pharmacists in California.²⁷ While limits on collaborative practice might prevent pharmacists from providing autonomous chronic disease management in the community setting, the greater inclusion of pharmacists in multi-disciplinary team-based care models could be achieved. Similarly, integration of physician extenders like nurse practitioners and physician assistants may prove effective in future iterations of the model.

The second barrier to widespread adoption of the model would be identifying who could and would shoulder the cost of such an intervention. Public and private payors would be ideal targets, but CMS has yet to grant pharmacists full provider status and private payors may be gun shy since they often experience rapid turn-over in membership and would incur increased upfront costs while failing to reap the long-term cost-savings from improved BP control rates. Large health systems like Kaiser Permanente (KP) would also be well-positioned to adopt this model of care. In recent years they have made impressive strides in improving BP control rates among blacks, but these successes were limited to those members who routinely go to clinic. Although an attractive target, union rules such as those that govern pharmacists at KP could prohibit work in off-site, community-based settings.

Irrespective of the factors outlined above one aspect of the model that could easily be widely disseminated and/or adopted nationally is the evidence-based drug treatment algorithm used in the intervention group. The combination of a long-acting ARB and CCB controlled BP in 50% of those who reached <130/80 mmHg. The addition of a potent, long-acting thiazide type diuretic resulted in another 10% of participants reaching goal. Adoption of this simple protocol in both traditional clinic settings would improve BP control rates among those engaged with the healthcare system leaving access and follow-up as the main deterrents for HTN control among blacks.

Future Research and Advocacy

Ongoing efforts to disseminate the barbershop model are underway. A pilot, which aims to replicate the results of LABBPS (albeit on much smaller scale) in a different city with a different pharmacist-led team is planned. The cost-effectiveness of LABBPS is currently being assessed to determine fiscally viable business models and to assess potential savings to public and private payers.

Additional research is needed to identify which elements of the program are most important and will inform changes to the model that improve operational efficiencies while maintaining intervention potency.

In California where pharmacists have achieved provider status, important policy changes have been put forth to allow specialty-trained pharmacists to be reimbursed by Medi-Cal (state Medicaid program) for comprehensive medication management in patients with uncontrolled chronic diseases such as HTN. If successful, this could serve as a model for other states where pharmacist-led programs go un-reimbursed despite a large body of evidence that supports their effectiveness.^20,21,28

Finally, there are health disparities among other racial, ethnic and gender groups that may benefit from similar community-based approaches. These may include diabetes among Hispanics, maternal death among black women and mental health disorders among Native Americans.²⁹ Future research into the appropriate venue for delivery of such programs is warranted.

CONCLUSION

The LABBPS provides strong evidence to support community-based healthcare interventions for high-risk and hard-to-reach populations. This study taught us many lessons on how to overcome difficulties with health care access and engagement, medication adherence and physician inertia. Given that education alone appears to be insufficient to fight HTN-related disparities, the inclusion of specialty care will improve the effectiveness of future community outreach programs. LABBPS demonstrated that pharmacist-led specialty care can be extremely effective in chronic disease management and thus deserves the legal backbone to make such programs widely available, and more importantly reimbursable. Further research is warranted and underway to further explore whether dissemination of this particular model on a national level is feasible and cost-effective.

KEY POINTS.

• Non-Hispanic blacks and black men in particular, are disproportionately burdened by hypertension-related morbidity and mortality

• Traditional models of care have failed to adequately curtail this disparity requiring the development of new/novel models of care

• The LA Barbershop Blood Pressure Study provides persuasive evidence that multidisciplinary, community-based approaches to hypertension management are effective

• Cost-effectiveness and implementation research are warranted to determine whether this model of care could be scaled nationally