Blood Pressure Control and Adherence to Drug Treatment in Patients with Hypertension Treated at a Specialized Outpatient Clinic: A Cross-Sectional Study.

OBJECTIVE: To evaluate and identify variables associated with the control of hypertension and adherence to antihypertensive drug treatment in a group of patients with hypertension monitored in a specialized, highly complex outpatient service.
METHODS: A prospective, cross-sectional study was carried out in the hypertension unit of a tertiary teaching hospital. Patients diagnosed with hypertensive aged 18 years and over and accompanied for at least six months were included in the study. Patients with secondary hypertension and pregnant women were excluded. The sample consisted of 253 patients. Adherence/concordance to antihypertensive treatment was assessed using the Morisky Green Levine Scale. Blood pressure control was set for values less than 140/90 mmHg. Variables with p≤0.20 in univariate analysis were included in multiple logistic regression. The level of significance adopted was p ≤0.05.
RESULTS: Most of patients were white, married and women, with a mean age of 65 (13.3) years old, low income, and education levels. Blood pressure control and adherence were observed in 69.2% and 90.1% of the patients, respectively. Variables that were independently associated with blood pressure control were (OR, odds ratio; CI, 95% confidence interval): married marital status (OR 2.3; CI 1.34-4.28), use of calcium channel blockers (OR 0.4; CI 0.19-0.92) and number of prescribed antihypertensive drugs (OR 0.78; CI 0.66-0.92). Adherence was not associated with any of the variables studied.
CONCLUSION: There was a high frequency of patients with satisfactory adherence to antihypertensive drug treatment. Blood pressure control was less frequent and was associated with social and treatment-related factors.

Introduction

Hypertension is a chronic cardiovascular disease which affects as many as 1.39 billion people globally.1 Hypertension is the main risk factor for cardiovascular disease and mortality worldwide.2,3 The prevalence of hypertension has increased in several countries over the years, with a global increase from 25.9% in 2000 to 31.1% in 2010,4 being higher in low and middle-income countries (31.5%) than in high-income countries (28.5%). In Brazil, the scenario is similar and hypertension affects about 31%5 of adult individuals and more than 60% of the elderly.6

Despite the robust evidence on the positive impact on cardiovascular morbidity and mortality, and on the reduction of lesions in target organs when blood pressure is reduced,7 control rates of hypertension are still unsatisfactory, even in high-income countries. In low and middle-income countries, the scenario seems to be even worse.8 The Prospective Urban Rural Epidemiology (PURE) study sought to assess this variability and blood pressure control rates found in low, medium-low, medium-high, and high-income countries were only 12.7%, 9.9%, 15.6%, and 19.0%, respectively.9 An observational study10 evaluated control rates in several regions of Brazil, with the objective of identifying this estimate in a middle-income country and almost half (46.4%) of persons with hypertension were outside the control target.

Effective control of blood pressure is the main objective of the treatment of hypertension.11 However, several factors influence this control, such as therapeutic inertia, the complexity of the individual, the socioeconomic and education level, access to unstructured health services/health systems, and adherence to treatment, which is certainly one of the most important factors.12–15 The World Health Organization (WHO) has defined adherence as

the degree to which a person’s behavior - taking the medication, following a diet and/or making lifestyle changes - corresponds to the recommendations agreed on with a healthcare provider/practitioner.16

In another document on the subject, adherence was defined as a process characterized by three major components: Initiation, Implementation and Discontinuation. Initiation occurs when the patient takes the first dose of a prescribed medication.; implementation is the extent to which a patient’s actual dosing corresponds to the prescribed dosing regimen from initiation until the last dose; and discontinuation marks the interruption, occurs when the patient stops taking the prescribed medication, for whatever reason(s).17 However, adherence/concordance is still a major challenge in the treatment of hypertension, with inadequate prevalence in most countries,18 with rates of around 50% in developed countries and about only 30% in developing countries.16

Much of the data on blood pressure control in low- and middle-income countries, such as Brazil, originates from primary healthcare services.11,19 This knowledge gap can be a barrier to the development of public policies for persons with hypertension of greater complexity, that is, those who have comorbidities and greater complications of the disease, and therefore, need assistance in tertiary care. In this context, it is relevant to know the profiles of hypertensive patients, and to identify the prevalence and the variables associated with control and adherence to treatment in this population. Therefore, the present study aimed to assess and identify variables associated with the control of arterial hypertension and adherence to antihypertensive drug treatment in a group of persons with hypertension accompanied in a specialized, outpatient clinic.

Methods

Design, Population, and Sample

A prospective cross-sectional study was conducted between August 2018 and July 2019. The clinic cares for approximately 800 persons with hypertension in a specialized and outpatient service located in a tertiary teaching hospital in the city of São Paulo, Brazil. The sample size was calculated, considering a control rate of 60%18 and a significance level of 5%, leading to the selection of 253 patients. The medical records were analysed and included persons with hypertension aged 18 years or older, who had been accompanied at the outpatient clinic for at least six months, and who voluntarily agreed to participate in the study by signing the Free and Informed Consent Form, were included in the study. Patients with secondary arterial hypertension and pregnant women were excluded from the study (Figure 1). The invitation to participate in the study was made at the outpatient’s routine medical visit. The interview for data collection was scheduled for another time and held in individual offices, by trained nurses.

Figure 1

Flowchart of inclusion and exclusion of the hypertensive patients.

Blood Pressure Control and Drug Treatment Adherence

Blood pressure control was assessed by blood pressure measurement and defined by systolic blood pressure less than 140 mmHg and diastolic blood pressure less than 90 mmHg. Blood pressure was measured by the principal investigator, in a calm and private environment, with a validated automatic device (Omron HEM-705 CP).20 Before blood pressure measuring, the investigator made sure that the patient did not have a full bladder; did not exercise for at least 60 minutes; not having ingested alcoholic beverages, coffee or food and not having smoked in the last 30 minutes.

The patients were kept seated, with their legs uncrossed, feet flat on the floor, back against the chair, and relaxed; with their one arm at the heart level with the palm of their hand facing up. A cuff suitable for the size of the arm was used. Three consecutive blood pressure measurements were taken, with an interval of two minutes between each measurement. The first blood pressure value was excluded and the mean value of the last two measurements was used.

Adherence/concordance to antihypertensive treatment was assessed using the Morisky Green Levine Medication Adherence Scale,21 which was validated for the Portuguese language in Brazil.22 The scale determines adherence to treatment based on the patient’s response to four questions that present a choice of dichotomous responses (yes or no). Adherence to drug treatment was considered when all responses were negative (final score = 0) and a positive response to one of the four items classifies the patient as non‐adherent to medication.

Study Variables

Biosocial data, habits and lifestyle, anthropometric measurements, body composition and laboratory data were collected. All instruments used were validated for the Portuguese language in Brazil.

Age, sex, race, marital status and education were obtained by self-report. For socioeconomic classification, the “Brazil Economic Classification Criterion” was used, which estimates the purchasing power of urban people and families.23 This classification assesses the possession and quantity of goods such as household appliances, housing, and education. The higher the score, the higher the socioeconomic status stratified into the following classes: A1, B1, B2, C1, C2, D, and Class E.23

Habits and lifestyles were also assessed by self-report and included smoking, alcohol consumption and physical activity. Regarding smoking, patients were classified as smokers, non-smokers and ex-smokers. The intake of alcoholic beverages was assessed by the Alcohol Use Disorders Test (AUDIT),24 which classifies patients according to scores on low-risk (between 0 and 7 points) consumption, risk use (8 to 15 points), harmful use (16 to 19 points), and probable dependence (≥20 points). Physical activity was assessed by a short version of the International Physical Activity Questionnaire (IPAQ)25 which classifies patients into: very active, active, irregularly active, or sedentary.

The psycho-emotional characteristics were analyzed using the SRQ-20-Self Report Questionnaire26 which assesses common non-psychotic mental disorders through 20 dichotomous questions (yes or no) and for each affirmative answer, 1 point is added. A person with a common mental disorder is considered when the patient reaches a final score> 6 points for men or> 8 points for women.

Anthropometric measures such as weight, height, body mass index, waist circumference, waist circumference, hip circumference, and neck circumference were also evaluated. Body composition was assessed using a calibrated bioimpedance scale and included: body fat percentage, visceral fat level, skeletal muscle percentage, and body age. All measurements were performed by the investigator. We also included data on fasting blood glucose, glycated hemoglobin, lipid profile, glomerular filtration rate analyzed by the MDRD equation, urea, creatinine, and proteinuria, in addition to information on the histories of comorbidities and drug treatments, all obtained from the patient’s electronic medical record.

Statistical Analysis

The level of significance adopted was 5% and used for categorical variables, Pearson’s chi-square test or Fischer’s exact tests, and for quantitative variables, the Wilcoxon-Mann–Whitney, Brunner-Munzel tests, and t-tests were used. In the logistic regression model, the candidate variables for predictors were those with a p-value <0.20 in the bivariate analysis of comparison between patients with controlled and uncontrolled blood pressure. The final model was presented according to the variables selected via LASSO (least absolute shrinkage and selection operator), in order to minimize the cross-validation error and respecting the standard error criterion.

Ethics

The research Ethics Committee of School of Nursing, University of São Paulo and Medical School, University of São Paulo approved the study, with reference numbers 3.003.912 and 2.831.454. All participants received a clear, plain-language summary of the study design, benefits, and risks and then signed voluntary informed consent forms. This study complies with the Declaration of Helsinki.

Results

Of the 253 persons with hypertension who composed the sample, 175 (69.2%) had their blood pressure under control and 228 (90.1%) adhered to drug treatment. Most of the analysed persons with hypertension were females (61.7%), white (63.2%), married (52.8%), and most (47.0%) belonged to a low socioeconomic stratum (C2). Most patients presented completed secondary education (44.3%), a monthly income in the range of 392.00 dollars. The mean age of the patients was 65.0 (13.3) years. Controlled persons with hypertension were statistically different (p <0.05) from uncontrolled ones only in relation to marital status, with a predominance of married individuals in the first group (58.9% vs 39.0%, respectively) (Table 1).

Table 1

Biosocial Variables, According to Groups of Controlled and Uncontrolled Persons with Hypertension Seen at a Specialized Outpatient Clinic

Variables	Hypertensives					Total (n=253)		p-value
	Controlled (n=175)		Not Controlled (n=78)
	n	%	n	%		n	%
Sex
Female	104	66.7	52	33.3		156	61.7	0.275*
Male	71	73.0	26	27.0		97	38.3
Age, years - Mean (SD)	65.2 (13.3)		64.5 (13.5)			65.0 (13.3)		0.587**
Ethnicity
White	116	72.5	44	27.5		160	63.2	0.133*
Not white	59	63.4	34	36.6		93	36.8
Marital status
Married	103	77.4	30	22.6		133	52.8	0.014*
Unmarried	25	61.0	16	39.0		41	16.7
Divorced	14	50.0	14	50.0		28	11.1
Widower	33	64.7	18	35.3		51	19.9
Education, years - Mean (SD)	10.39 (3.8)		9.9 (4.1)			10.2 (3.9)		0.274**
Education level
Illiterate	1	50.0	1	50.0		2	0.8	0.755***
Incomplete primary education	10	55.6	8	44.6		18	7.1
Complete primary education	50	66.7	25	33.3		75	29.6
Incomplete secondary education	4	57.0	3	43.0		7	2.8
Complete secondary education	83	74.0	29	26.0		112	44.3
Incomplete tertiary education	6	67.0	3	33.0		9	3.6
Complete tertiary education	21	70,0	9	30,0		30	11,9
Monthly income (US$) - Mean (SD)	429.66 (313.25)		391.67 (345.47)			417.92 (323.34)		0.093****
Socioeconomic strata
A1	1	100.0	0	0.0		1	0.4	0.218***
B1 e B2	26	81.2	6	18.8		32	12.6
C1	45	70.3	19	29.3		64	25.3
C2	79	66.4	40	33.6		119	47.0
D-E	24	64.9	13	35.1		37	14.6

Notes: *p obtained by Pearson’s x2 test; **Mann–Whitney test; ***Fisher’s exact test; ****Brunner-Munzel tests.

Regarding habits and lifestyles, it was found that just over half of the patients (56.1%) reported not being smokers and only 21.7% practiced physical activity on a regular basis. A representative part of the sample (40.3%) used alcohol, however, almost all (98.8%) presented low-risk consumption (Table 2).

Table 2

Characteristics of Habits and Lifestyles According to Groups of Controlled and Uncontrolled Persons with Hypertension Seen at a Specialized Outpatient Clinic

Variables	Hypertensives				Total (n=253)		p-value
	Controlled (n=175)		Not Controlled (n=78)
	n	%	n	%	n	%
Smoking
Yes	12	63.2	7	36.8	19	7.5	0.721*
No	97	68.3	45	31.7	142	56.1
Ex-smoker	66	71.7	26	28.3	92	36.4
Physical activity
Active	43	78.2	12	21.8	55	21.7	0.094*
Irregularly active	72	71.3	29	28.7	101	39.9
Sedentary	60	61.8	37	38.2	97	38.3
Alcoholism
Yes	75	73.5	27	26.5	102	40.3	0.218*
No	100	66.2	51	33.8	151	59.7
Alcohol Use Disorders Test (AUDIT) final score
Low risk consumption	173	69.2	77	30.8	250	98.8	1.000**
Risk consumption	2	66.6	1	33.3	3	1.2

Notes: *p obtained by Pearson’s x2 test; **Fisher’s exact test.

The data in Table 3 show that the body mass index was at the upper limit of the overweight range [29.5 (5.3) Kg/m2] and the expressive majority (81.4%) showed overweight/obesity. The means of abdominal circumference, waist-hip ratio, and neck circumference were altered, in the increased/greatly increased ranges. The variables of body composition, body fat, and visceral fat were in the high/very high ranges and the majority (56.1%) had a percentage of skeletal muscle qualified as low. The body age estimated by bioimpedance, was above the real biological age. Controlled persons with hypertension were statistically different (p <0.05) from uncontrolled ones, as they presented lower body mass index averages [29.0 (5.4) vs 30.6 (4.9) Kg/m2], waist circumferences [101.6 (13.7) vs 101.7 (10.4) cm], body fat index [36.1 (10.5) vs 38.1 (9.3)] and body age [64.9 (12.0) vs 68.1 (11.1)].

Table 3

Anthropometric Characteristics and Body Composition, According to Groups of Controlled and Uncontrolled Persons with Hypertension Seen at a Specialized Outpatient Clinic

Variables	Hypertensives				Total (n=253)		p-value
	Controlled (n=175)		Not Controlled (n=78)
	n	%	n	%	n	%
Body mass index (Kg/m2) - Mean (SD)
	29.0 (5.4)		30.6 (4.9)		29.5 (5.3)		0.015*
Abdominal circumference (cm) - Mean (SD)
	101.6 (13.7)		101.7 (10.4)		101.7 (12.7)		0.027*
Waist/Hip Ratio - Mean (SD)
	0.99 (0.60)		0.94 (0.06)		0.98 (0.50)		0.942*
Neck circumference (cm) - Mean (SD)
	39.7 (3.1)		40.0 (2.9)		39.8 (3.0)		0.191*
Body composition variables
Body Fat - Mean (SD)
	36.1 (10.5)		38.1 (9.3)		36.9 (10.2)		0.037**
Visceral Fat - Mean (SD)
	11.9 (4.2)		12.8 (4.7)		12.2 (4.4)		0.141*
Skeletal muscle - Mean (SD)
	26.1 (5.2)		26.0 (4.8)		26.1 (5.1)		0.917**
Body Age (years) - Mean (SD)
	64.9 (12.0)		68.1 (11.1)		65.9 (11.8)		0.033*

Notes: *p obtained by the Mann–Whitney test; **Two Sample t-test.

The analysis of laboratory tests showed that 15.6% had fasting glucose values compatible with diabetes. However, considering the values of glycated hemoglobin, the percentage rose to 30.4%. It is noteworthy that 40.7% had a medical diagnosis of diabetes mellitus. Most had desirable triglyceride values (63.6%), total cholesterol (69.6%) and HDL-cholesterol (81.4%). In relation to LDL-cholesterol, just over half (56.5%) presented values considered as excellent, however, dyslipidemia reported in medical records was the most prevalent morbidity (71.5%). Most presented values of urea and creatinine within the normal range (82.2% and 61.7%, respectively) and the estimated glomerular filtration rate, despite having the average with normal value, 31.6% of the persons with hypertension showed values <60 mL/min and 28.1% had protein in their urine. At lower frequencies, the presence of other comorbidities such as chronic kidney disease (19.0%), neoplasia (15.4%), and heart failure (13.4%) were also observed. Regarding psycho-emotional characteristics, 23.7% of the studied persons with hypertension were classified as having common mental disorders (Table 4).

Table 4

Comorbidities and Laboratory Markers, According to Groups of Controlled and Uncontrolled Persons with Hypertension Seen at a Specialized Outpatient Clinic

Variables		Hypertensives							Total (n=253)		p-value
		Controlled (n=175)				Not Controlled (n=78)
		n		%		n	%		n	%
Fasting blood glucose (mg/dL) - Mean (SD)
		109.2 (30.1)				113.2 (30.6)			110 (30.1)		0.110*
Normal		81		46.3		28	35.9		109	43.1	0.206**
Increased risk for diabetes		61		34.9		36	46.2		97	38.3
Diabetes		33		18.9		14	17.9		47	15.6
Glycated hemoglobin (HbA1c) - Mean (SD)
		6.2 (1.2)				6.5 (1.5)			6.3 (1.3)		0.078*
Normal		67		38.3		22	28.2		89	35.2	0.274**
Pre Diabetes		56		32.0		31	39.7		87	34.4
Diabetes mellitus		52		29.7		25	32.1		77	30.4
Triglycerides (mg/dL) - Mean (SD)
		139.7 (85.7)				148.7 (71.8)			142.5 (81.6)		0.140*
Desirable level		118		67.4		43	55.1		161	63.6	0.061**
Total Cholesterol (mg/dL) - Mean (SD)
		174.4 (37.7)				182.9 (44.2)			177.1 (39.9)		0.109*
Desirable level		124		70.9		52	66.7		176	69.6	0.504**
Low Density Lipoproteins -LDL-c (mg/dL) - Mean (SD)
		94.9 (31.8)				101.8 (35.7)			97.1 (33.1)		0.100*
Great level		102		58.3		41	52.6		143	56.5	0.397**
High Density Lipoproteins HDL-c (mg/dL) - Mean (SD)
		55.3 (16.9)				55.7 (15.12)			55.4 (16.3)		0.564*
Desirable level		141		80.6		65	83.3		206	81.4	0.603**
Serum urea (mg/dL) - Mean (SD)
		40.1 (22.9)				41.3 (18.5)			40.5 (21.6)		0.258*
Normal		144		82.3		64	82.1		208	82.2	0.964**
Altered		31		17.7		14	17.9		45	17.8
Creatinine (mg/dL) - Mean (SD)
		1.1 (0.9)				1.2 (0.7)			1.1 (0.82)		0.437*
Normal		111		63.4		45	57.7		156	61.7	0.387**
Altered		64		36.6		33	42.3		97	38.3
Estimated glomerular filtration rate MDRD (mL/min) - Mean (SD)
		71.3 (23.2)				66.4 (22.3)			69.8 (22.1)		0.119***
MDRD†<60 mL/min		53		30.3		27	34.6		80	31.6	0.495**
Proteinuria
Absent		132		75.4		50	64.1		182	71.9	0.065**
Present		43		24.6		63	80.8		71	28.1
Dyslipidemia		122		69.7		59	75.6		181	71.5	0.336**
Diabetes mellitus		70		40.0		33	42.3		103	40.7	0.731**
Chronic renal failure		31		17.1		17	21.8		48	19.0	0.446**
Neoplasm		24		13.7		15	19.2		39	15.4	0.263**
Cardiac insufficiency		22		12.6		12	15.4		34	13.4	0.545**
Arrhythmia		17		9.7		6	7.7		23	9.1	0.606**
Brain stroke		17		9.7		5	6.4		22	8.7	0.390**
Acute myocardial infarction		14		8.0		7	9.0		21	8.3	0.796**
Depression		11		6.3		5	6.4		16	6.3	0.970**
Peripheral vascular disease		9		5.1		3	3.8		12	4.7	0.655**
Chronic airway disease		6		3.4		3	3.8		9	3.6	0.869**
Digestive diseases		6		3.4		2	2.6		8	3.2	0.717**
Self Report Questionnaire (SRQ–20) final score
Presence of Common Mental Disorders
		39		65.0		21	35.0		60	23.7	0.424**
Blood pressure (mmHg) - Mean (SD)
Systolic blood pressure	133.5 (7.1)				150.1 (13.0)				138.6 (12.0)		<0.001****
Diastolic blood pressure	79.1 (6.8)				84.8 (6.6)				80.9 (7.2)		<0.001*

Notes: *p obtained by the Mann–Whitney test; **Pearson’s x2 test; ***Two Sample t-test; ****Brunner-Munzel.

Abbreviation: †MDRD, Modification of Diet in Renal Disease.

The average of the total number of drugs prescribed per patient was 8.0 (4.0) and the average number of antihypertensive drugs was 3.7 (1.9). Only 9.5% of persons with hypertension used only one antihypertensive medication, 42.7% two or three, and slightly less than half (47.8%) used four or more antihypertensive drugs. The most prescribed class of antihypertensive drugs was diuretics (78.3%), followed by calcium channel blockers (68.8%), AT1 receptor blockers (61.7%), beta-blocker (50.2%), angiotensin-converting enzyme inhibitor (30.4%), and at lower frequencies, centrally acting agents (15.8%), vasodilators (10.0%) and alpha-blockers (7.1%). Controlled persons with hypertension compared to uncontrolled ones were statistically different (p <0.05) since they presented the lowest total average of drugs prescribed per patient [7.6 (3.8) vs 9.1 (4.1)] and antihypertensive drugs [3.4 (1.7) vs 4.5 (2.1)]; in addition to a lower frequency of antihypertensive classes: calcium channel blockers (61.7% vs 84.6%), beta-blocker (44.0% vs 64.1%), centrally acting agents (12, 6% vs 23.1%), and vasodilators (5.7% vs 19.2%) (Table 5).

Table 5

Characteristics of Drug Treatment, According to Groups of Controlled and Uncontrolled Persons with Hypertension Seen at a Specialized Outpatient Clinic

Variables	Hypertensives				Total (n=253)		p-value
	Controlled (n=175)		Not Controlled (n=78)
	n	%	n	%	n	%
Number of medications - Mean (SD)
	7.6 (3.8)		9.1 (4.1)		8.0 (4.0)		<0.001*
Number of antihypertensive drugs - Mean (SD)
	3.4 (1.7)		4.5 (2.1)		3.7 (1.9)		<0.001**
One	20	11.4	4	5.1	24	9.5	<0.001***
Two to three	87	49.7	21	26.9	108	42.7
Four or more	68	38.9	53	67.9	121	47.8
Therapeutic classes
Diuretic	132	75.4	66	84.6	198	78.3	0.103***
Calcium channel blockers
	108	61.7	66	84.6	174	68.8	<0.001***
AT1 receiver blockers (BRAs)
	109	62.3	47	60.3	156	61.7	0.760***
Adrenergic beta-blocker inhibitors
	77	44.0	50	64.1	127	50.2	0.003***
Angiotensin-converting enzyme inhibitor
	53	30.3	24	30.8	77	30.4	0.939***
Central action agents	22	12.6	18	23.1	40	15.8	0.35***
Vasodilators	10	5.7	15	19.2	25	10.0	<0.001***
Alpha blockers	14	8.0	4	5.1	18	7.1	0.413***

Notes: *p obtained by the Mann–Whitney test; **Brunner-Munzel test; ***Pearson’s x2 test.

The multiple logistic regression analysis showed that there was an association (p <0.05) with blood pressure control the variables: marital status, use of calcium channel blockers, and the number of prescribed antihypertensive drugs. Thus, married persons with hypertension were 2.3 times more likely to control blood pressure than unmarried ones; those using calcium channel blockers were 55.9% less likely to control blood pressure; and with each antihypertensive medication added to the treatment, the chance of blood pressure control decreased by 21.3% (Table 6).

Table 6

Logistic Regression Model and Blood Pressure Control in Persons with Hypertension Seen at a Specialized Outpatient Clinic

Variable	Odds Ratio	Confidence Interval (95%)	p-value
Marital status
Unmarried	1
Married	2.382	1.345–4.283	0.003
Calcium channel blockers
No	1
Yes	0.441	0.199–0.928	0.036
Number of antihypertensive drugs in use
	0.787	0.668–0.923	0.004

None of the variables studied was associated with adherence to antihypertensive drug treatment.

Discussion

The main results in this study showed that there was little agreement between the rates of blood pressure control and adherence/concordance to antihypertensive treatment (69.2% vs 90.1%). However, married marital status, the use of calcium channel blockers and the number of antihypertensive drugs were variables that were independently associated with blood pressure control.

The prevalence of hypertension control found here (~70%), although insufficient, was similar to that of developed countries and a reference in the treatment of hypertension, such as Canada, whose current estimate is 68%.27 Data from the National Health and Nutrition Examination Survey estimated that the proportion of blood pressure control increased from 31.8% in 1999–2000 to 48.5% in 2007–2008 (p <0.001), to 53.8% in 2013 −2014 (p = 0.14), and then dropped to 43.7% in 2017–2018 (p = 0.003).28 This decline is a worrying finding since the lack of hypertension control is directly related to the occurrence of cardiovascular diseases and death from all causes.6 Data from the 2017 Global Burden of Disease show that in Brazil the total number of deaths attributable to high systolic blood pressure increased by 53.4%, ranking first as the main risk factor for death in the period from 1990 to 2017.29 From this perspective, the global estimates for control are alarming, as seen in the “May Measurement Month, 2017” conducted in more than 80 countries, showing that among the persons with hypertension treated, only 46.3% had their blood pressure controlled.30 Results from Brazilian studies are similar, with rates around 50%.9,31,32 These control data are not very satisfactory, especially considering that in the majority, they result from persons with hypertension seen in primary healthcare, where patients with less severe disease are concentrated.

The measurement of blood pressure, despite being considered a standard procedure for the diagnosis and follow-up of people with hypertension, is subject to the influence of the observer and the environment.33,34 Thus, the optimized result of the prevalence of control found in the present study may also have been influenced by the fact that the patient was evaluated in an environment outside the place of treatment and blood pressure measurements were performed by a nurse, probably attenuating the white coat effect.

The prevalence of adherence to antihypertensive treatment, assessed by the Morisky Green Levine Scale, was higher than the data found in the literature. A systematic review with meta-analysis with the inclusion of studies that also used self-report measures showed that the prevalence of adherence/concordance ranged from 36.7% to 66.2%.35 In the present study it was found that adherence was not a predictor of blood pressure control and in persons with hypertension who reported being adherent to treatment, about a third did not have their blood pressure under control. The use of a self-report instrument to assess adherence may have influenced this result, since this method may overestimate the results. However, the instrument used is widely applied in studies with the same theme, and can be considered the most applicable tool to clinical practice, in view of its practicality and low cost. Measuring adherence/concordance to antihypertensive treatment is a complex task. There is no gold standard based on the combination of methods and various dimensions involving the treatment adherence process recommended by the World Health Organization.15 In this study, direct assessment was not possible due to financial issues.

Interestingly, marital status was the only biosocial variable that was associated with blood pressure control. Married patients with hypertension were more likely to control blood pressure in comparison to unmarried ones. Similarly, a meta-analysis study showed the influence of marital status on the incidence of cardiovascular disease, evidencing that single individuals were more likely to have cardiovascular disease compared to married couples.36 A study that evaluated adherence to drug treatment in patients after acute myocardial infarction also showed that married persons with hypertension were more adherent to treatment.37 The protective effect of a married marital status can be attributed to changes in behavior and social networks resulting from the union and the established family bond. The association between marital status and hypertension has been described in other studies.38–40 However, the influence of marital status in the control of hypertension is not frequent in the literature and expanding knowledge about this aspect is relevant. Family and social support is a fundamental part of the health-disease process and the fact of having a partner can be a facilitating agent in the treatment of hypertensive persons.

Uncontrolled persons with hypertension had a higher body mass index, waist circumference, body fat and body age in the bivariate analysis (p <0.05), when compared with the controlled subjects. Robust evidence41 already demonstrates this relationship and reinforces the need for measures that promote the adoption of healthier behaviors, especially among hypertensive persons, with the objective of achieving the blood pressure control goal.

The laboratory profile of hypertensive persons, consistent with the most prevalent comorbidities, pointed to an important percentage of persons with alterations in glycated hemoglobin, LDL and glomerular filtration rate, which are relevant markers when investigating the degree of severity of the hypertensive persons. Despite not being associated with blood pressure control, these results reinforce the severity profile of the studied persons, of whom 30.8% still remained uncontrolled. Another important finding was that control was associated with characteristics of drug treatment. With each antihypertensive added to the treatment, the chance of pressure control decreased by 21.3%. This result may be a consequence of the fact that the greater the number of medications, the greater the difficulty in correctly taking the medications and the lower the chance of adhering to the treatment, reflecting a worsening of control. However, only about 30% of hypertensive persons respond to treatment with monotherapy, therefore requiring more than one medication.42,43 However, there is evidence of low adherence to treatment in hypertensive persons with polypharmacy.44–46 The association of fixed doses of medications would be an alternative to reduce the number of medications taken, however, with the Brazilian public system for dispensing antihypertensive drugs it is not possible. It should be noted that the studied sample was composed of complex patients, with other diseases associated with hypertension, such as diabetes, dyslipidemia, in addition to lesions in target organs, justifying the need for more complex drug treatments.

The uncontrolled persons with hypertension presented prescriptions, in greater quantity (p<0.05), in the bivariate analysis of some pharmacological classes, however only the use of calcium channel blockers remained associated with the blood pressure control at the final multivariate model. Hypertensive persons using calcium channel blockers had a 55.9% less chance of controlling blood pressure. The negative association between the use of this medication and the control of blood pressure can be related to several factors, such as greater severity of persons with hypertension and more difficulty in controlling the disease. Calcium channel blockers are effective antihypertensives, reduce cardiovascular morbidity and mortality, have good tolerability and present safety in the treatment of arterial hypertension.47

Limitations

An important limitation of the present study was the use of a self-report instrument to assess adherence. It is a form of evaluation that can underestimate or overestimate the result. The ideal would be the combination of direct and indirect methods, however, the limitation of financial resources in the Brazilian reality was a limiting factor. It is worth mentioning that the Morisky Green Levine Scale has good psychometric characteristics, is an easy-to-use, widely used, low-cost method, in addition to identifying important attitudes towards barriers related to adherence. Another limitation is related to the cross-sectional design, which does not allow establishing relationships between cause and effect and the fact that it was carried out in a single center, resulting in difficulty in generalizing the data.

Conclusion

Despite the presence of important risk factors and target organ injuries in the studied persons with hypertension, the prevalence of control and adherence was satisfactory. The multiprofessional team has a privileged position in the health system to strengthen the social support network of persons with hypertension. Strategies that favour the integration between hypertensive persons, family, health services and other social facilities can be beneficial for the control of blood pressure and adherence to drug treatment. The goal of controlling persons with hypertension is not an easy task to be achieved, but population strategies must be implemented. Having a usual source of care, optimizing real adherence, and minimizing therapeutic inertia can lead to higher rates of blood pressure control. A collaborative partnership between patients, health professionals, and the health system, incorporates a multilevel approach to the control of hypertension and can bring satisfactory results. The findings of the present study made it possible to evaluate adherence to antihypertensive drug treatment in an indirect way, blood pressure control, and associated variables, in the largest public hospital complex in Latin America and the results can contribute to the improvement of public health policies.