Antihypertensive Therapy in African Americans: Findings From an Inner‐City Ambulatory Clinic

Abstract

J Clin Hypertens (Greenwich). 2010;12:187–192. ^©2010 Wiley Periodicals, Inc.

African Americans bear a greater burden of hypertension. Understanding prevailing epidemiologic patterns can facilitate the implementation and successful outcome of community programs. The authors assessed practice patterns of antihypertensive drug utilization and blood pressure (BP) control in a predominantly African American population in Brooklyn, NY, from January 1 to January 31, 2008. A total of 416 (53.1%) had hypertension, with a mean age of 61 years, and 267 (64%) were women. In general, 212 (50.9%) were at goal BP and 59.9% of those at goal were taking at least 2 drugs. Patient age correlated with the number of drugs used (r=0.14; P=.004). Patients taking β‐blockers and calcium channel blockers were older: 63.6 vs 60.1 years (P=.01) and 62.7 vs 60.3 years (P=.07), respectively. The pattern of antihypertensive use was as follows: angiotensin‐converting enzyme inhibitors, 194 (46.6%); calcium channel blockers, 162 (38.9%); diuretics, 162 (38.9%); β‐blockers, 133(32%); and angiotensin receptor blockers, 93 (22.4%). The findings of age associated with the class of medications used and a predominance of angiotensin‐converting enzyme inhibitors usage highlight possible gaps in appropriateness of antihypertensive therapy. The application of age‐appropriate race‐based antihypertensive therapy might improve BP control rates. These results strengthen arguments for investing in community‐based programs to overcome possible provider‐related and local health system barriers to achieving BP control goals.

Hypertension is the most common primary diagnosis in America and is responsible for 35.7 million office visits per year. ¹ Overall, 1 in 3 American adults have high blood pressure (BP), while 2 in 5 African American adults have high BP. ² African Americans bear a greater burden of disease, ³ , ⁴ have a variable response to conventional antihypertensive medications, ⁵ , ⁶ and develop more severe end‐organ damage. ⁷ , ⁸

In addition to pharmacologic therapy and lifestyle modifications, community‐based interventions focusing on African Americans may improve outcomes. In this light, the Centers for Disease Control and Prevention (CDC) has initiated the Racial and Ethnic Approaches to Community Health Across the US (REACH US) program. ⁹ , ¹⁰ , ¹¹ , ¹² Similarly, the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC 7) ¹³ recommends a 3‐pronged approach of pharmacologic therapy, lifestyle changes, and public health initiatives involving community participation and mobilization. Pharmacologic therapy remains a primary focus of care, however, as it prevents and reverses end‐organ damage and improves cardiovascular outcomes. ¹⁴ , ¹⁵ Correspondingly, to ensure a cost‐effective, sustainable, and successful community‐based intervention, the intensity and design of public health programs involving grassroots initiatives ought to be guided by a clear understanding of prevailing epidemiologic patterns of hypertension.

As such, the CDC calls for continuous surveillance of health status in minority communities so that culturally sensitive prevention strategies can be tailored to these communities and program interventions evaluated. ¹⁶ Knowledge of existing prescribing patterns, antihypertensive drug utilization, and BP control rates in the index community can provide useful information for establishing community programs to combat hypertension and gauge their effectiveness. We assessed practice patterns and BP control in a predominantly African American population serviced by a community health clinic affiliated with an internal medicine residency program.

Methods

Design and Setting

We performed a cross‐sectional retrospective study to examine the pattern of antihypertensive medication use and rates of BP control in a sample of predominantly African American patients at Interfaith Medical Center ambulatory care clinic, a 287 bed multisite community teaching health care system in Brooklyn, NY, which serves more than 250,000 predominantly African American patients, many of whom are of Caribbean origin. Patients were considered African American if they were self‐identified as black (defined as of African descent). Patients are serviced by internal medicine residents, as well as board certified internists. Established patients were defined as patients who had been with the clinic for at least 6 months with a diagnosis of hypertension. Hypertension was defined as either a BP ≥140/90 mm Hg on at least two occasions or the use of antihypertensive medications or a documented history of hypertension.

Participants

Consecutive patients visiting the medical clinic from January 1, 2008, to January 31, 2008, were screened. Patients with a diagnosis of hypertension following chart review were eligible. All medical records were electronically kept.

Variables

Variables included age, sex, race, antihypertensive medication, BP, and comorbid illnesses. The classes of antihypertensive medications were recorded according to JNC 7 classification, ¹³ which included angiotensin‐converting enzyme (ACE) inhibitors, angiotensin receptor blockers (ARBs), β‐blockers (BBs), calcium channel blockers (CCBs), thiazide‐type diuretics, direct vasodilators, α‐blockers, and centrally acting drugs.

Outcome Variables

BP control was assessed according to JNC 7 guidelines: both systolic BP <140 mm Hg and diastolic BP <90 mm Hg for patients without diabetes or renal disease (abnormal creatinine or presence of proteinuria). For patients with diabetes mellitus or renal disease, a BP goal of systolic BP <130 mm Hg and a diastolic BP <80 mm Hg was used. ¹³ The BP recorded on the most recent clinic visit during the study period was used to determine hypertension control. If a patient was seen more than once during the study period, only the most recent visit was counted.

Data Collection

Data was reviewed on all consecutive patients who attended the medical clinic during the period from January 1, 2008, to January 31, 2008. A 1‐month period was chosen to avoid duplication of patient data.

Study Size

A total of 784 patients attended the medical clinic during the month of January in 2008. Of these, 428 patients (54.6%) were found to be hypertensive and thus eligible for the study. Eleven non–African American patients and 1 patient who was attending the clinic for the first time were excluded, leaving 416 patients for analysis. Patient demographic information, antihypertensive medications used, and BP recorded on the most recent clinic visit during the study period were extracted from patients’ medication reconciliation records and clinic progress notes by accessing electronic medical records. The institutional review board of the Interfaith Medical Center approved the study prior to commencement.

Statistical Methods

Statistical software package STATA version 9 (StataCorp LP, College Station, TX) was used for analysis. Data are reported as mean ± standard deviation for continuous variables and other data are reported as percentages. Differences in BP control between sex and age subgroups were detected using chi‐square tests. Predictors of BP control were modeled using binomial logistic regression. The model included known risk factors for BP control (age, sex, diabetes, number of medications, and number of comorbidities). P values ≤.05 were accepted as statistically significant. Composite variables were created for classes of antihypertensive medications using all patient medications.

Results

A total of 416 patients (53.1%) had hypertension (Table I). The mean age was 61±13 years, (range 31–97 years), 267 (64%) were women, 154 (37%) had diabetes mellitus, and 34 (8.2%) had end‐stage renal disease on dialysis.

Table I.

 Baseline Characteristics of Patients

Characteristic	Patients (N=416)
Mean age, (SD), y	61 (13)
Women, No. (%)	267 (64)
Diabetes mellitus, No. (%)	154 (37)
Dyslipidemia, No. (%)	171 (41)
Coronary artery disease, No. (%)	33 (8)
End‐stage renal disease, No. (%)	34 (8)
Asthma, No. (%)	69 (17)
Chronic obstructive pulmonary disease, No. (%)	25 (6)
Human immunodeficiency virus, No. (%)	2 (0.5)
Atrial fibrillation, No. (%)	11 (3)
Congestive heart failure, No. (%)	25 (6)
Thyroid disease, No. (%)	26 (6)
Gout, No. (%)	11 (3)
Dementia, No. (%)	5 (1)
Hepatitis C, No. (%)	22 (5)
Glaucoma, No. (%)	7 (2)
Depression, No. (%)	13 (3)
Peripheral vascular disease, No. (%)	10 (2)
Other, No. (%)	85 (20)

The average number of antihypertensive medications used was 1.93±1.01: 140 (33.7%) of patients were taking 1 drug, 162 (38.9%) were taking 2 drugs, 100 (24%) were taking ≥3 drugs, and 14 patients (3.4%) were diet‐controlled. (Figure 1). The difference between number of antihypertensive medications used in men and women was not significant (1.92 vs 1.93; P=.93). A positive correlation between age and number of antihypertensive medications (r=0.14, P=.004) and a negative correlation between number of antihypertensive medications used and BP control (r=−0.13, P=.01) were observed. These findings persisted after controlling for possible sex effects.

Figure 1.

 Percentage distribution of antihypertensive medications used.

Overall, 212 (51%) patients were at target BP. Nearly 70% (63.5%) of patients without diabetes or end‐stage renal disease were at goal (<140/90 mm Hg), while only 57 (33.1%) of diabetic patients and/or patients with end‐stage renal disease were at goal BP (<130/80 mm Hg) (P<.001). There was no difference in BP control between women and men (51.3% vs 50.3%, respectively; P=.85). On average, the mean on‐treatment systolic BP was 132.4±18.5 mm Hg, with a diastolic BP of 73.6±9.1 mm Hg. Patients at goal were significantly younger (60 vs 63 years; P=.01), and there was a significant negative correlation between age and BP control (r=−0.12, P=.01). Table II shows the majority of patients at goal: 127 (59.9%) were taking ≥2 drugs, with a trend toward less BP control on monotherapy vs multiple agents (37.8% vs 62.3%, P=.21). On average, compared with patients who were controlled, uncontrolled patients were taking significantly more antihypertensive agents (2.05 vs 1.80; P=.01). Patients with diabetes mellitus were less likely to attain goal BP (P<.001), even after controlling for age, sex, and number of comorbidities.

Table II.

 Attainment of Goal Blood Pressure by Number of Antihypertensive Medications

No. of Drugs	No. of Patients (N=212)	Controlled, %
0 (diet)	8	57.1
1	77	55
2	86	53.1
3	32	45.1
4	8	36.4
5	1	16.7

The pattern of antihypertensive drug use was as follows: ACE inhibitors, 194 (46.6%); followed by CCBs, 162 (38.9%); and diuretics, 162 (38.9%). BBs were utilized by 133 (31.9%) patients, ARBs by 93 (22.4%), centrally acting agents by 18 (4.3%), vasodilators by 7 (1.7%), and α‐agonists by 13 (3.1%). Patients taking BBs were significantly older compared with patients taking other agents (63.6 vs 60.1 years; P=.01), and a similar trend was observed in patients taking CCBs (62.7 vs 60.3 years; P=.07). There was a trend for higher diuretic use in women as compared with men (42.3% vs 32.9%; P=.06) (Figure 2). Compared with nondiabetics, diabetic patients were more likely to receive an ACE inhibitor (55.8% vs 41.2%; P=.004) but less likely to receive a diuretic (31.2% vs 43.5%; P=.01) or CCB (33.1% vs 42.4%; P=.06). The rate of BB use was similar in both diabetic and nondiabetic patients (31.2% vs 32.4%; P=.79).

Figure 2.

 Antihypertensive drug class utilization by sex. ACEI indicates angiotensin‐converting enzyme inhibitor; CCB, calcium channel blocker; BB, β‐blocker; ARB, angiotensin receptor blocker.

Discussion

Prevalence of Hypertension

Hypertension has been the top‐ranked medical diagnosis by physicians at office visits 8 times since 1996. ² In our community, the prevalence of hypertension was 53.1%, nearly 20% higher than the national average of 39%. ¹⁷ A positive correlation has been found between age and number of medications used, suggesting that older patients have more difficult‐to‐control BP and inadvertently require more medications for BP control. The finding of age and sex associations with the class of medications used is similar to other studies on hypertension. ¹⁸ , ¹⁹ Nevertheless, the adequacy of sex‐ and age‐appropriate therapy in African Americans in this community should be further explored to detect possible local health system and provider barriers to BP control.

Antihypertensive Drug Utilization Patterns

JNC 7 acknowledges that when public health intervention strategies address the diversity of racial, ethnic, cultural, linguistic, religious, and social factors in the delivery of their services, the likelihood of their acceptance by the community increases. ¹³ In the delivery of services, however, an intimate understanding of the population dynamics with respect to hypertension is key to the appropriate distribution of available resources by the government, health care organizations, and pharmaceutical industry.

The National Institute for Health and Clinical Excellence (NICE), in a joint initiative with the British Health Society (BHS) in 2006, recommended CCBs and diuretics as first‐line agents in the absence of other compelling indications in African Americans older than 55 years. ²⁰ Interestingly, an overwhelming majority of patients were taking ACE inhibitors, which, in general, are costlier than diuretics and are associated with discontinuation of therapy due to a higher incidence of cough, especially in African American women. ²¹ Furthermore, the Antihypertensive and Lipid‐Lowering Treatment to Prevent Heart Attack (ALLHAT) trial showed that African Americans treated with ACE inhibitors had significantly higher BPs compared with the diuretic group and had a greater incidence of combined cardiovascular disease and strokes. ²² This pattern of antihypertensive drug use raises concerns about possible long‐term detrimental effects and highlights possible gaps in appropriateness of antihypertensive class use in this community. Patients taking BBs and CCBs tended to be older and, thus, were less likely to have their BP controlled. Similar patterns have been observed in other studies, with only 34% of adults older than 65 years achieving BP control across the United States. ²³ Polypharmacy in this age group leading to selective noncompliance, patient drug discontinuation due to the increased potential for systemic side effects, and potential outcome benefits of these same medications in patients with coexisting comorbidities such as ischemic heart disease are possible explanations for this pattern of drug use. Further research exploring this pattern is warranted. In addition, our findings in diabetics raise concerns of underutilization of CCBs and diuretics in African Americans with diabetes.

BP Control Rates

Heart disease death rates are 30% higher for African Americans, and stroke death rates are 41% higher, ⁹ further sounding a call to action for the reversing of the morbidity trend being charted by the ongoing hypertension epidemic. Although overall BP control rates of 51% in this community were above the national average, which generally ranges from 31% to 38%, ¹⁶ , ¹⁹ , ²⁴ subgroup analyses in the diabetic and end‐stage renal disease population were in the lower range. Using data from the National Health and Nutrition Examination Survey (NHANES), Gu and colleagues ¹⁸ found that among patients with hypertension, women were less likely than men to meet BP control targets and more often used diuretics, which is similar to our finding of women using diuretics. The possible benefit of thiazide diuretics in maintaining calcium homeostasis and thus regulating the progression of osteoporosis in women might explain this finding. The United Kingdom Prospective Diabetes Study (UKPDS) showed that the risk of death due to diabetes, stroke, and microvascular disease is significantly lower in patients with tighter BP control. ²⁵ The control rate for diabetics in this community is trending somewhat below the national average, and, as thus, leaves a gap for achieving the Healthy People goals of 2010, as well as for reducing morbidity in this community.

Limitations

Our study has several limitations. First, as with most observational studies, we are unable to determine the influence of important confounders such as lifestyle factors (eg, smoking, alcohol use, or dietary patterns), which might have affected the observed BP control rates. Second, the measurement of BP was not standardized across patients and only a single value was used to base our conclusion on BP control. It is unlikely, however, that this lack of standardization would significantly impact BP measurements in the index visit. By enrolling patients who were established in the clinic for at least 6 months, the likelihood that the index BP was not consistent with the general average BP of the patient was reduced. Third, we were unable to determine adherence to prescribed medications, thus control rates may be overemphasized or underemphasized with regards to antihypertensive drug classes. Fourth, since patient comorbidities were retrieved from documentations in patient charts, the potential for lack of documentation of disease exists, thus creating an information bias.

Conclusions

Overall, the BP control rate of close to 51% in this central Brooklyn community is commendable and encouraging. In patients without diabetes and end‐stage kidney disease, higher control rates are associated with ACE inhibitor use. With advancing age, use of BBs and CCBs are associated with poorer BP control. In diabetic patients, underutilization of CCBs and diuretics might explain the lower rates of BP control. Health care providers in this community should be provided with explicit education on appropriate antihypertensive therapy in African Americans with complicated and uncomplicated hypertension. The application of age‐appropriate race‐based antihypertensive therapy might improve BP control rates. Our study, in addition to providing a baseline cohort of patients to be followed for outcomes research, also strengthens arguments for community‐based interventions targeted at overcoming possible health care system– and provider‐related barriers to achieving BP control goals.