138
Clinical methods and pathophysiology
Identifying masked uncontrolled hypertension in the community pharmacy setting Daniel Sabater-Hernándeza,b, Jose Sendra-Lillob, Jose J. Jiménez-Monleónc, Shalom I. Benrimoja, Desire Durksa, Fernando Martínez-Martínezb,d and Alejandro de la Sierrae,f; on behalf of the MEPAFAR study workgroup Masked uncontrolled hypertension (MUCH) is associated with an increased cardiovascular risk. This condition is frequent in the community pharmacy (i.e., CP-MUCH), but there is no evidence on the factors associated with its presence in that setting. The aim of this analysis was to explore these factors. A sample of 98 treated hypertensive patients from the MEPAFAR study, with normal community pharmacy blood pressure (CPBP < 135/85 mmHg), were analyzed. Blood pressure (BP) was also measured at home (4 days) and monitored for 24 h. CP-MUCH was identified when either ambulatory (daytime) or home BP averages were equal to or higher than 135/85 mmHg. A multivariate logistic regression analysis was carried out to identify the factors associated with CP-MUCH. The prevalence of CP-MUCH tends to be higher as systolic and diastolic CPBP increase, reaching 47% in patients with both systolic and diastolic CPBP equal to or higher than 123 mmHg and 79 mmHg, respectively. The multivariate regression analysis showed only systolic CPBP as an independent factor of CP-MUCH [ ≥ 123 mmHg: odds ratio = 16.46 (P = 0.012); from 115 to 122.9 mmHg: odds ratio = 10.74 (P = 0.036); systolic CPBP < 115 mmHg as the reference]. Further
Introduction Masked uncontrolled hypertension (MUCH) is defined in those treated hypertensive patients who have elevated ambulatory (ABP) or home blood pressure (HBP) despite normal physician office measurements [1]. This condition is found in 22% of treated patients and has been associated with an increased cardiovascular risk [2]. This higher risk may be explained by the fact that ABP and HBP have been shown to be stronger predictors of target organ damage and cardiovascular events than blood pressure (BP) measured in a physician’s office [3,4]. MUCH could remain undetected if only office BP readings are available for evaluation. As a consequence, if normal BP figures are measured in the physician’s office, the necessary adjustments in treatment may not be made and the underlying risk associated with elevated ABP or HBP would therefore not be well managed. As List of the MEPAFAR study workgroup: Hernández-Peña N., Sabater-Díaz J.M., Artiles-Campelo A., Jorge-Rodríguez M.E., García-Morales M.F., García-Rodríguez A.R., Merino-Barber L., Santana-Perez F., Díaz-Merino N., Artiles-Ruano M.E., Contardi-Lista A.M.
assessment, using ambulatory and/or home BP monitoring, is recommended in patients with normal CPBP, but systolic CPBP equal to or higher than 115 mmHg. A more feasible approach would be evaluating patients with systolic CPBP equal to or higher than 123 mmHg and diastolic CPBP equal to or higher than 79 mmHg. Blood Press Monit 20:138–143 Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved. Blood Pressure Monitoring 2015, 20:138–143 Keywords: Community Pharmacy Services, hypertension, masked hypertension, risk factors a Graduate School of Health, University of Technology Sydney, New South Wales, Australia, bAcademic Centre in Pharmaceutical Care, cFaculty of Medicine, d Faculty of Pharmacy, University of Granada, Granada, eDepartment of Internal Medicine, Hospital Mutua Terrassa and fFaculty of Medicine, University of Barcelona, Barcelona, Spain
Correspondence to Daniel Sabater-Hernández, PharmD, PhD, Graduate School of Health, University of Technology Sydney, Level 4, Building 7, 67 Thomas St, PO Box 123, Ultimo NSW 2007, Australia Tel: + 61 2 9514 7201; fax: + 61 2 9514 8300; e-mail: [email protected] Received 5 September 2014 Revised 4 December 2014 Accepted 8 December 2014
measurement of out-of-office BP in all normotensive individuals is not a feasible or an efficient strategy in daily clinical practice, several studies have been carried out to identify the factors associated with the presence of MUCH [5–9]. By considering these factors and identifying patients who are more likely to have MUCH, a further assessment, using ABP or HBP monitoring (ABPM or HPBM), could be recommended. MUCH has also been identified as a relatively frequent condition in the community pharmacy (i.e., CP-MUCH) [10,11]. Particularly in this case, if normal BP figures are measured in the community pharmacy, pharmacists would not refer patients to a physician, and again, the necessary changes in treatment would not be made. Given that CP-MUCH could influence both decisionmaking processes and health outcomes, the presence of this condition should be investigated carefully when pharmacists deliver pharmaceutical care services to treated hypertensive patients. To our knowledge, there are no reports in the literature on the factors associated with CP-MUCH. As a
1359-5237 Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
DOI: 10.1097/MBP.0000000000000105
Identifying masked hypertension Sabater-Hernández et al. 139
consequence, community pharmacists do not have guidelines that would allow them to better identify patients who are likely to present with CP-MUCH. In light of the prevalence of CP-MUCH and given its clinical implications, a subgroup analysis of patients with normal community pharmacy BP (CPBP) who were included in the MEPAFAR study [11] was carried out to explore the factors associated with CP-MUCH.
Methods The MEdida de la Presión Arterial en FARmacia (MEPAFAR) study was a cross-sectional study carried out from June 2008 to June 2009 in eight Spanish community pharmacies. The main aim of the study was to assess the agreement between CPBP, daytime ABP, and HBP in treated hypertensive patients [11], and thus the capacity of the CPBP measurement method to evaluate the effectiveness of antihypertensive treatment. The protocol of the MEPAFAR study was assessed and approved by the Research Ethics Committee of the University of Granada (Spain). BP measurement methods have been described comprehensively in previous publications [11,12]. Briefly, CPBP was measured at four visits, by the same pharmacist at each pharmacy, over a 4-week period. At each visit, three BP measurements were taken (2–3 min apart) on the control arm (arm on which CPBP was higher on the first visit). Patients’ visits to the pharmacy were scheduled at the same time for all four time points (± 1 h). The mean CPBP was calculated discarding the data from the first visit and the first measurement from each visit. Table 1
For the purpose of this subanalysis, normal CPBP was defined by systolic BP (SBP) less than 135 mmHg and diastolic BP (DBP) less than 85 mmHg. At home, patients monitored their BP over 4 consecutive working days, taking three measurements in the morning (2 min apart, 6 a.m. to 9 a.m.) and three measurements in the evening (6 p.m. to 9 p.m.) on the nondominant arm. HBP readings were stored in the device’s memory. The mean HBP was calculated discarding values obtained on the first day and using the first and second measure from each morning and each evening. Normal HBP was defined by SBP less than 135 mmHg and DBP less than 85 mmHg. The clinically validated OMRON M10-IT automatic electronic device (Omron Corp., Tokyo, Japan) [13] was used both at home and in the community pharmacy. ABPM was performed on a working day (24 h) using the nondominant arm. Measurements were taken every 20 min (7 a.m. to 10 p.m.) or 30 min (10 p.m. to 7 a.m.). The clinically validated Spacelabs Medical 90207-5Q monitor (Spacelabs Inc., Redmond, Washington, USA) was used [14]. The average daytime ABP was calculated according to a sleep diary kept by each patient. Normal daytime ABP was defined by SBP less than 135 mmHg and DBP less than 85 mmHg. On the basis of recently acknowledged definitions [15,16], CP-MUCH was defined when an average CPBP (SBP/DBP) of less than 135/85 mmHg was combined with one or both of the following: average daytime ABP
General characteristics of the sample and classification by blood pressure profile Alla (n = 98)
Age (years) [mean (SD)] Female [n (%)] BMI (kg/m2) [mean (SD)] Smokers [n (%)] Dyslipidemia [n (%)] Diabetes [n (%)] History of CVD [n (%)] Antihypertensive drugs [n (%)] One Two Three or more Nonadherence [n (%)] Systolic BP (mmHg) [mean (SD)] Community pharmacy Home Daytime ambulatory Diastolic BP (mmHg) [mean (SD)] Community pharmacy Home Daytime ambulatory
Controlledb (n = 74)
CP-MUCH (n = 24)
P-valuec
55.5 60 30.4 13 33 16 5
(11.4) (61.2) (4.5) (13.3) (33.7) (16.3) (5.1)
54.6 45 30.4 8 27 10 4
(11.5) (60.8) (4.2) (11.0) (36.5) (13.5) (5.4)
58.3 15 30.3 5 6 6 1
(11.0) (62.5) (5.5) (20.8) (25.0) (25.0) (4.2)
0.166 0.883 0.954 0.218 0.301 0.210 1.000
47 32 19 30
(48.0) (32.7) (19.4) (30.6)
35 23 16 23
(47.3) (31.1) (21.6) (31.1)
12 9 3 7
(50.0) (37.5) (12.5) (29.2)
0.596
0.860
118.4 (9.7)d 121.9 (10.8) 122.9 (10.3)
116.3 (9.7)e 118.6 (8.6) 119.7 (8.7)
124.9 (6.8)d 132.3 (10.3) 132.7 (8.5)
< 0.001 < 0.001 < 0.001
75.4 (6.1)f 76.7 (6.5)e 74.4 (7.9)f
74.5 (6.1) 75.1 (5.6)e 72.7 (7.3)f
78.0 (5.3)f 81.8 (6.1) 79.6 (7.4)
0.013 < 0.001 < 0.001
ABP, ambulatory blood pressure; BP, blood pressure; CP-MUCH, community pharmacy masked uncontrolled hypertension; CVD, cardiovascular disease; DBP, diastolic blood pressure; HBP, home blood pressure; SBP, systolic blood pressure. a All patients with controlled community pharmacy blood pressure. b Controlled blood pressure by both ABP and HBP. c For the differences between individuals with and without CP-MUCH. d Statistically significant differences with both HBP and ABP. e Statistically significant differences with HBP. f Statistically significant differences with ABP.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
140 Blood Pressure Monitoring 2015, Vol 20 No 3
equal to or higher than 135/85 mmHg and/or average HBP equal to or higher than 135/85 mmHg.
quantitative independent variables (i.e. SBP, DBP, age, and BMI) were entered into the model in their original form. However, none of them showed a linear relationship with the dependent variable when linearity was checked. Thus, these quantitative variables were converted into categorical variables on the basis of their tertiles (i.e. each variable was stratified into three groups with similar numbers of individuals). A second multivariate logistic model was constructed using CP-MUCH, defined by daytime ABP only, as the dependent variable. The goodness-of-fit of the models was assessed using the Hosmer–Lemeshow test (the model was considered acceptable if the test was not statistically significant). A P-value less than 0.05 was considered statistically significant.
To characterize the study sample, the following variables were collected: age, sex, smoking status, BMI, number of antihypertensive drugs used, adherence to antihypertensive drugs (Morisky–Green test [17]), history of previous cardiovascular disease (stroke, myocardial infarction, angina, or peripheral artery disease), and presence of diabetes or dyslipidemia (documented diagnosis or previously prescribed drug treatment). Statistical analysis
SPSS, version 15.0 (SPSS Inc., Chicago, Illinois, USA) was used for the statistical calculations. Mean and SD were used to summarize quantitative variables. Qualitative variables were described using frequencies and percentages. To compare the quantitative variables, Student’s t-test for independent samples was used. Differences between CPBP, HBP, and daytime ABP were assessed by repeated-measures analysis of variance applying the Bonferroni correction. χ2 and Fisher’s exact tests were used for comparisons of proportions.
Results The final sample of the MEPAFAR study included 169 treated hypertensive patients, 98 of whom had CPBP less than 135/85 mmHg. This subgroup of 98 patients represented the sample of the present analysis. The general characteristics of the sample are shown in Table 1. CP-MUCH was found in 24 (24.5%) individuals, 11 of whom had uncontrolled HBP only; eight had uncontrolled daytime BP only; and five had both home and daytime uncontrolled BP. Patients with CP-MUCH presented significantly higher CPBP, HBP, and ABP than those without the condition; however, none of the mean BP values were above the threshold limits (Table 1). Among patients with CP-MUCH, those with both home and daytime uncontrolled BP (n = 5)
To identify the independent factors associated with CPMUCH, a multivariate logistic regression analysis was used. As candidate independent variables for the multivariate models, CPBP (SBP and DBP) and all the variables collected to characterize the study sample were considered. Univariate logistic regression was used to select the independent variables that were finally entered in the multivariate models (P < 0.2). A-priori, Fig. 1
140.0 139.6
137.7
135.6
135.0 125.6
121.6 122.9
80.7 85.8
85.9
85.0 76.4 76.8
35.0
SBP DBP Uncontrolled daytime ABP only (n=8)
128.6
125.9
81.2 75.7
SBP DBP Uncontrolled HBP only (n=11) CPBP
HBP
74.8
77.9
SBP DBP Uncontrolled daytime ABP and HBP (n =5)
Daytime ABP
Average blood pressure figures in patients with CP-MUCH. ABP, ambulatory blood pressure; CPBP, community pharmacy blood pressure; CPMUCH, community pharmacy masked uncontrolled hypertension; DBP, diastolic blood pressure; HBP, home blood pressure; SBP, systolic blood pressure.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Identifying masked hypertension Sabater-Hernández et al. 141
Table 2 Multivariate logistic analysis and prevalence of community pharmacy masked uncontrolled hypertension across different groups (n = 98) Prevalence of CPMUCH [n (%)] Sex Female (n = 60) Male (n = 38) Age Less than or equal to 50 years (n = 34) 51–61 years (both included) (n = 31) 62 years or more (n = 33) Smoking habit Nonsmoker (n = 84) Smoker (n = 13) Diabetes No (n = 82) Yes (n = 16) Dyslipidemia No (n = 65) Yes (n = 33) History of CVD No (n = 93) Yes (n = 5) BMI ≤ 27.7 kg/m2 (n = 32) 27.8–31.9 kg/m2 (both included) (n = 34) ≥ 32.0 kg/m2 (n = 32) Number of antihypertensive drugs One (n = 47) Two (n = 32) Three or more (n = 19) Adherence No (n = 30) Yes (n = 68) Community pharmacy SBP < 115 mmHg (n = 34) 115–122.9 mmHg (both included) (n = 29) ≥ 123 mmHg or more (n = 35) Community pharmacy DBP < 75 mmHg (n = 33) 75–78 mmHg (both included) (n = 32) ≥ 79 mmHg (n = 33)
Adjusted OR
P-value
15 (25) 9 (23.7)
NI
6 (17.6)
Reference
10 (32.3)
1.23
0.764
8 (24.2)
1.59
0.511
19 (22.6) 5 (38.5) 18 (22.0) 6 (37.5)
NI
Reference 1.26 0.707
18 (27.7) 6 (18.2)
NI
23 (24.7) 1 (20.0)
NI
9 (28.1) 8 (23.5)
NI
7 (21.9) 12 (25.5) 9 (28.1) 3 (15.8)
NI
7 (23.3) 17 (25.0)
NI
1 (2.9) 9 (31.0) 14 (40.0) 5 (15.2) 6 (18.8) 13 (39.4)
Reference 10.74 0.036 16.46
0.012
Reference 0.88 0.862 2.08
0.314
Hosmer–Lemeshow test (χ = 6.625; P = 0.469); Nagelkerke R : 0.267. CVD, cardiovascular disease; CP-MUCH, community pharmacy masked uncontrolled hypertension; DBP, diastolic blood pressure; NI, not included in the multivariate model (P ≥0.2 in the univariate logistic model); OR, odds ratio; SBP, systolic blood pressure. 2
2
presented the highest SBP figures [128.6 (SD: 6.6) at the pharmacy, 140.0 (SD: 6.4) mmHg at home, and 139.6 (SD: 3.4) mmHg by ambulatory monitoring] (Fig. 1). As shown in Table 2 and Fig. 2, the proportion of individuals with CP-MUCH tends to be higher as systolic and diastolic CPBP increases; for example, the prevalence of CP-MUCH reached 47% (seven out of 15) in patients with both systolic and diastolic CPBP equal to or higher than 123 mmHg and 79 mmHg, respectively. On the basis of the univariate regression analysis, age, diabetes, and both systolic and diastolic CPBP were selected as factors to be entered into the multivariate model (Table 2). The multivariate regression analysis indicated
systolic CPBP equal to or higher than 115 mmHg as an independent factor of CP-MUCH [from 115 to 122.9 mmHg: odds ratio (OR) = 10.74 (P = 0.036); ≥ 123 mmHg: OR = 16.46 (P = 0.012); systolic CPBP < 115 mmHg as the reference]. When only daytime ABP was used to identify CP-MUCH (n = 13), again, only systolic CPBP was found to be associated [ ≥ 123 mmHg: OR = 10.85 (P = 0.036); systolic CPBP < 115 mmHg as the reference].
Discussion Measurement of BP in community pharmacies is a widespread practice that is recommended by several hypertension societies [18–20]. Nevertheless, some caution should be exercised when considering and using this method; specifically, when normal figures are obtained, CPBP readings might encourage erroneous conclusions about hypertension control in some cases [10,11]. In fact, in this study, CP-MUCH was identified in a quarter of the patients with normal CPBP. Ideally, ABPM and/or HBPM would be recommended in all patients with normal CPBP; however, this is not feasible or achievable in practice. In light of this, our results might help community pharmacists identify individuals who are more likely to present CP-MUCH and, thus, prioritize the use of the out-of-pharmacy BP methods among treated hypertensive patients with normal CPBP. On the basis of the results of the multivariate analysis, CP-MUCH was associated with systolic CPBP and should be investigated when systolic figures equal to or higher than 115 mmHg are obtained. However, because of the low BP cut-off point, this recommendation still encourages a wide use of ABPM and/or HBPM that might not be achievable in some cases (for example, a patient’s lack of willingness to monitor BP). Additional results of this study allow identification of certain circumstances in which either the prevalence of CP-MUCH or the relevance of the situation might be greater. When CP-MUCH was defined by means of ABPM only, the multivariate analysis showed that individuals with systolic CPBP equal to or higher than 123 mmHg are more likely to present CP-MUCH. This is particularly noteworthy when it is considered that ABPM is the reference method in the management of hypertension. In addition, although diastolic CPBP was not associated with CPMUCH, the prevalence of CP-MUCH increased (across any SBP strata) as diastolic CPBP was closer to the uppernormal limits (≥79 mmHg; Fig. 2). Overall, HBP and ABP figures were not remarkably increased in patients with CP-MUCH (Table 1 and Fig. 1). Only individuals with both home and daytime uncontrolled BP showed high SBP figures. Interestingly, four out of these five individuals had systolic CPBP equal to or higher than 123 mmHg. On the basis of these observations taken together, community pharmacists would be encouraged to further assess BP control (using ABPM and/or HBPM)
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
142 Blood Pressure Monitoring 2015, Vol 20 No 3
Fig. 2
47% 50%
36% 33%
43%
20%
20%
25%
≥123 mmHg 0% 0%
5% 115 − 122.9 mmHg
0% ≥79 mmHg 75 − 78 mmHg
Clinical methods and pathophysiology
Identifying masked uncontrolled hypertension in the community pharmacy setting Daniel Sabater-Hernándeza,b, Jose Sendra-Lillob, Jose J. Jiménez-Monleónc, Shalom I. Benrimoja, Desire Durksa, Fernando Martínez-Martínezb,d and Alejandro de la Sierrae,f; on behalf of the MEPAFAR study workgroup Masked uncontrolled hypertension (MUCH) is associated with an increased cardiovascular risk. This condition is frequent in the community pharmacy (i.e., CP-MUCH), but there is no evidence on the factors associated with its presence in that setting. The aim of this analysis was to explore these factors. A sample of 98 treated hypertensive patients from the MEPAFAR study, with normal community pharmacy blood pressure (CPBP < 135/85 mmHg), were analyzed. Blood pressure (BP) was also measured at home (4 days) and monitored for 24 h. CP-MUCH was identified when either ambulatory (daytime) or home BP averages were equal to or higher than 135/85 mmHg. A multivariate logistic regression analysis was carried out to identify the factors associated with CP-MUCH. The prevalence of CP-MUCH tends to be higher as systolic and diastolic CPBP increase, reaching 47% in patients with both systolic and diastolic CPBP equal to or higher than 123 mmHg and 79 mmHg, respectively. The multivariate regression analysis showed only systolic CPBP as an independent factor of CP-MUCH [ ≥ 123 mmHg: odds ratio = 16.46 (P = 0.012); from 115 to 122.9 mmHg: odds ratio = 10.74 (P = 0.036); systolic CPBP < 115 mmHg as the reference]. Further
Introduction Masked uncontrolled hypertension (MUCH) is defined in those treated hypertensive patients who have elevated ambulatory (ABP) or home blood pressure (HBP) despite normal physician office measurements [1]. This condition is found in 22% of treated patients and has been associated with an increased cardiovascular risk [2]. This higher risk may be explained by the fact that ABP and HBP have been shown to be stronger predictors of target organ damage and cardiovascular events than blood pressure (BP) measured in a physician’s office [3,4]. MUCH could remain undetected if only office BP readings are available for evaluation. As a consequence, if normal BP figures are measured in the physician’s office, the necessary adjustments in treatment may not be made and the underlying risk associated with elevated ABP or HBP would therefore not be well managed. As List of the MEPAFAR study workgroup: Hernández-Peña N., Sabater-Díaz J.M., Artiles-Campelo A., Jorge-Rodríguez M.E., García-Morales M.F., García-Rodríguez A.R., Merino-Barber L., Santana-Perez F., Díaz-Merino N., Artiles-Ruano M.E., Contardi-Lista A.M.
assessment, using ambulatory and/or home BP monitoring, is recommended in patients with normal CPBP, but systolic CPBP equal to or higher than 115 mmHg. A more feasible approach would be evaluating patients with systolic CPBP equal to or higher than 123 mmHg and diastolic CPBP equal to or higher than 79 mmHg. Blood Press Monit 20:138–143 Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved. Blood Pressure Monitoring 2015, 20:138–143 Keywords: Community Pharmacy Services, hypertension, masked hypertension, risk factors a Graduate School of Health, University of Technology Sydney, New South Wales, Australia, bAcademic Centre in Pharmaceutical Care, cFaculty of Medicine, d Faculty of Pharmacy, University of Granada, Granada, eDepartment of Internal Medicine, Hospital Mutua Terrassa and fFaculty of Medicine, University of Barcelona, Barcelona, Spain
Correspondence to Daniel Sabater-Hernández, PharmD, PhD, Graduate School of Health, University of Technology Sydney, Level 4, Building 7, 67 Thomas St, PO Box 123, Ultimo NSW 2007, Australia Tel: + 61 2 9514 7201; fax: + 61 2 9514 8300; e-mail: [email protected] Received 5 September 2014 Revised 4 December 2014 Accepted 8 December 2014
measurement of out-of-office BP in all normotensive individuals is not a feasible or an efficient strategy in daily clinical practice, several studies have been carried out to identify the factors associated with the presence of MUCH [5–9]. By considering these factors and identifying patients who are more likely to have MUCH, a further assessment, using ABP or HBP monitoring (ABPM or HPBM), could be recommended. MUCH has also been identified as a relatively frequent condition in the community pharmacy (i.e., CP-MUCH) [10,11]. Particularly in this case, if normal BP figures are measured in the community pharmacy, pharmacists would not refer patients to a physician, and again, the necessary changes in treatment would not be made. Given that CP-MUCH could influence both decisionmaking processes and health outcomes, the presence of this condition should be investigated carefully when pharmacists deliver pharmaceutical care services to treated hypertensive patients. To our knowledge, there are no reports in the literature on the factors associated with CP-MUCH. As a
1359-5237 Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
DOI: 10.1097/MBP.0000000000000105
Identifying masked hypertension Sabater-Hernández et al. 139
consequence, community pharmacists do not have guidelines that would allow them to better identify patients who are likely to present with CP-MUCH. In light of the prevalence of CP-MUCH and given its clinical implications, a subgroup analysis of patients with normal community pharmacy BP (CPBP) who were included in the MEPAFAR study [11] was carried out to explore the factors associated with CP-MUCH.
Methods The MEdida de la Presión Arterial en FARmacia (MEPAFAR) study was a cross-sectional study carried out from June 2008 to June 2009 in eight Spanish community pharmacies. The main aim of the study was to assess the agreement between CPBP, daytime ABP, and HBP in treated hypertensive patients [11], and thus the capacity of the CPBP measurement method to evaluate the effectiveness of antihypertensive treatment. The protocol of the MEPAFAR study was assessed and approved by the Research Ethics Committee of the University of Granada (Spain). BP measurement methods have been described comprehensively in previous publications [11,12]. Briefly, CPBP was measured at four visits, by the same pharmacist at each pharmacy, over a 4-week period. At each visit, three BP measurements were taken (2–3 min apart) on the control arm (arm on which CPBP was higher on the first visit). Patients’ visits to the pharmacy were scheduled at the same time for all four time points (± 1 h). The mean CPBP was calculated discarding the data from the first visit and the first measurement from each visit. Table 1
For the purpose of this subanalysis, normal CPBP was defined by systolic BP (SBP) less than 135 mmHg and diastolic BP (DBP) less than 85 mmHg. At home, patients monitored their BP over 4 consecutive working days, taking three measurements in the morning (2 min apart, 6 a.m. to 9 a.m.) and three measurements in the evening (6 p.m. to 9 p.m.) on the nondominant arm. HBP readings were stored in the device’s memory. The mean HBP was calculated discarding values obtained on the first day and using the first and second measure from each morning and each evening. Normal HBP was defined by SBP less than 135 mmHg and DBP less than 85 mmHg. The clinically validated OMRON M10-IT automatic electronic device (Omron Corp., Tokyo, Japan) [13] was used both at home and in the community pharmacy. ABPM was performed on a working day (24 h) using the nondominant arm. Measurements were taken every 20 min (7 a.m. to 10 p.m.) or 30 min (10 p.m. to 7 a.m.). The clinically validated Spacelabs Medical 90207-5Q monitor (Spacelabs Inc., Redmond, Washington, USA) was used [14]. The average daytime ABP was calculated according to a sleep diary kept by each patient. Normal daytime ABP was defined by SBP less than 135 mmHg and DBP less than 85 mmHg. On the basis of recently acknowledged definitions [15,16], CP-MUCH was defined when an average CPBP (SBP/DBP) of less than 135/85 mmHg was combined with one or both of the following: average daytime ABP
General characteristics of the sample and classification by blood pressure profile Alla (n = 98)
Age (years) [mean (SD)] Female [n (%)] BMI (kg/m2) [mean (SD)] Smokers [n (%)] Dyslipidemia [n (%)] Diabetes [n (%)] History of CVD [n (%)] Antihypertensive drugs [n (%)] One Two Three or more Nonadherence [n (%)] Systolic BP (mmHg) [mean (SD)] Community pharmacy Home Daytime ambulatory Diastolic BP (mmHg) [mean (SD)] Community pharmacy Home Daytime ambulatory
Controlledb (n = 74)
CP-MUCH (n = 24)
P-valuec
55.5 60 30.4 13 33 16 5
(11.4) (61.2) (4.5) (13.3) (33.7) (16.3) (5.1)
54.6 45 30.4 8 27 10 4
(11.5) (60.8) (4.2) (11.0) (36.5) (13.5) (5.4)
58.3 15 30.3 5 6 6 1
(11.0) (62.5) (5.5) (20.8) (25.0) (25.0) (4.2)
0.166 0.883 0.954 0.218 0.301 0.210 1.000
47 32 19 30
(48.0) (32.7) (19.4) (30.6)
35 23 16 23
(47.3) (31.1) (21.6) (31.1)
12 9 3 7
(50.0) (37.5) (12.5) (29.2)
0.596
0.860
118.4 (9.7)d 121.9 (10.8) 122.9 (10.3)
116.3 (9.7)e 118.6 (8.6) 119.7 (8.7)
124.9 (6.8)d 132.3 (10.3) 132.7 (8.5)
< 0.001 < 0.001 < 0.001
75.4 (6.1)f 76.7 (6.5)e 74.4 (7.9)f
74.5 (6.1) 75.1 (5.6)e 72.7 (7.3)f
78.0 (5.3)f 81.8 (6.1) 79.6 (7.4)
0.013 < 0.001 < 0.001
ABP, ambulatory blood pressure; BP, blood pressure; CP-MUCH, community pharmacy masked uncontrolled hypertension; CVD, cardiovascular disease; DBP, diastolic blood pressure; HBP, home blood pressure; SBP, systolic blood pressure. a All patients with controlled community pharmacy blood pressure. b Controlled blood pressure by both ABP and HBP. c For the differences between individuals with and without CP-MUCH. d Statistically significant differences with both HBP and ABP. e Statistically significant differences with HBP. f Statistically significant differences with ABP.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
140 Blood Pressure Monitoring 2015, Vol 20 No 3
equal to or higher than 135/85 mmHg and/or average HBP equal to or higher than 135/85 mmHg.
quantitative independent variables (i.e. SBP, DBP, age, and BMI) were entered into the model in their original form. However, none of them showed a linear relationship with the dependent variable when linearity was checked. Thus, these quantitative variables were converted into categorical variables on the basis of their tertiles (i.e. each variable was stratified into three groups with similar numbers of individuals). A second multivariate logistic model was constructed using CP-MUCH, defined by daytime ABP only, as the dependent variable. The goodness-of-fit of the models was assessed using the Hosmer–Lemeshow test (the model was considered acceptable if the test was not statistically significant). A P-value less than 0.05 was considered statistically significant.
To characterize the study sample, the following variables were collected: age, sex, smoking status, BMI, number of antihypertensive drugs used, adherence to antihypertensive drugs (Morisky–Green test [17]), history of previous cardiovascular disease (stroke, myocardial infarction, angina, or peripheral artery disease), and presence of diabetes or dyslipidemia (documented diagnosis or previously prescribed drug treatment). Statistical analysis
SPSS, version 15.0 (SPSS Inc., Chicago, Illinois, USA) was used for the statistical calculations. Mean and SD were used to summarize quantitative variables. Qualitative variables were described using frequencies and percentages. To compare the quantitative variables, Student’s t-test for independent samples was used. Differences between CPBP, HBP, and daytime ABP were assessed by repeated-measures analysis of variance applying the Bonferroni correction. χ2 and Fisher’s exact tests were used for comparisons of proportions.
Results The final sample of the MEPAFAR study included 169 treated hypertensive patients, 98 of whom had CPBP less than 135/85 mmHg. This subgroup of 98 patients represented the sample of the present analysis. The general characteristics of the sample are shown in Table 1. CP-MUCH was found in 24 (24.5%) individuals, 11 of whom had uncontrolled HBP only; eight had uncontrolled daytime BP only; and five had both home and daytime uncontrolled BP. Patients with CP-MUCH presented significantly higher CPBP, HBP, and ABP than those without the condition; however, none of the mean BP values were above the threshold limits (Table 1). Among patients with CP-MUCH, those with both home and daytime uncontrolled BP (n = 5)
To identify the independent factors associated with CPMUCH, a multivariate logistic regression analysis was used. As candidate independent variables for the multivariate models, CPBP (SBP and DBP) and all the variables collected to characterize the study sample were considered. Univariate logistic regression was used to select the independent variables that were finally entered in the multivariate models (P < 0.2). A-priori, Fig. 1
140.0 139.6
137.7
135.6
135.0 125.6
121.6 122.9
80.7 85.8
85.9
85.0 76.4 76.8
35.0
SBP DBP Uncontrolled daytime ABP only (n=8)
128.6
125.9
81.2 75.7
SBP DBP Uncontrolled HBP only (n=11) CPBP
HBP
74.8
77.9
SBP DBP Uncontrolled daytime ABP and HBP (n =5)
Daytime ABP
Average blood pressure figures in patients with CP-MUCH. ABP, ambulatory blood pressure; CPBP, community pharmacy blood pressure; CPMUCH, community pharmacy masked uncontrolled hypertension; DBP, diastolic blood pressure; HBP, home blood pressure; SBP, systolic blood pressure.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Identifying masked hypertension Sabater-Hernández et al. 141
Table 2 Multivariate logistic analysis and prevalence of community pharmacy masked uncontrolled hypertension across different groups (n = 98) Prevalence of CPMUCH [n (%)] Sex Female (n = 60) Male (n = 38) Age Less than or equal to 50 years (n = 34) 51–61 years (both included) (n = 31) 62 years or more (n = 33) Smoking habit Nonsmoker (n = 84) Smoker (n = 13) Diabetes No (n = 82) Yes (n = 16) Dyslipidemia No (n = 65) Yes (n = 33) History of CVD No (n = 93) Yes (n = 5) BMI ≤ 27.7 kg/m2 (n = 32) 27.8–31.9 kg/m2 (both included) (n = 34) ≥ 32.0 kg/m2 (n = 32) Number of antihypertensive drugs One (n = 47) Two (n = 32) Three or more (n = 19) Adherence No (n = 30) Yes (n = 68) Community pharmacy SBP < 115 mmHg (n = 34) 115–122.9 mmHg (both included) (n = 29) ≥ 123 mmHg or more (n = 35) Community pharmacy DBP < 75 mmHg (n = 33) 75–78 mmHg (both included) (n = 32) ≥ 79 mmHg (n = 33)
Adjusted OR
P-value
15 (25) 9 (23.7)
NI
6 (17.6)
Reference
10 (32.3)
1.23
0.764
8 (24.2)
1.59
0.511
19 (22.6) 5 (38.5) 18 (22.0) 6 (37.5)
NI
Reference 1.26 0.707
18 (27.7) 6 (18.2)
NI
23 (24.7) 1 (20.0)
NI
9 (28.1) 8 (23.5)
NI
7 (21.9) 12 (25.5) 9 (28.1) 3 (15.8)
NI
7 (23.3) 17 (25.0)
NI
1 (2.9) 9 (31.0) 14 (40.0) 5 (15.2) 6 (18.8) 13 (39.4)
Reference 10.74 0.036 16.46
0.012
Reference 0.88 0.862 2.08
0.314
Hosmer–Lemeshow test (χ = 6.625; P = 0.469); Nagelkerke R : 0.267. CVD, cardiovascular disease; CP-MUCH, community pharmacy masked uncontrolled hypertension; DBP, diastolic blood pressure; NI, not included in the multivariate model (P ≥0.2 in the univariate logistic model); OR, odds ratio; SBP, systolic blood pressure. 2
2
presented the highest SBP figures [128.6 (SD: 6.6) at the pharmacy, 140.0 (SD: 6.4) mmHg at home, and 139.6 (SD: 3.4) mmHg by ambulatory monitoring] (Fig. 1). As shown in Table 2 and Fig. 2, the proportion of individuals with CP-MUCH tends to be higher as systolic and diastolic CPBP increases; for example, the prevalence of CP-MUCH reached 47% (seven out of 15) in patients with both systolic and diastolic CPBP equal to or higher than 123 mmHg and 79 mmHg, respectively. On the basis of the univariate regression analysis, age, diabetes, and both systolic and diastolic CPBP were selected as factors to be entered into the multivariate model (Table 2). The multivariate regression analysis indicated
systolic CPBP equal to or higher than 115 mmHg as an independent factor of CP-MUCH [from 115 to 122.9 mmHg: odds ratio (OR) = 10.74 (P = 0.036); ≥ 123 mmHg: OR = 16.46 (P = 0.012); systolic CPBP < 115 mmHg as the reference]. When only daytime ABP was used to identify CP-MUCH (n = 13), again, only systolic CPBP was found to be associated [ ≥ 123 mmHg: OR = 10.85 (P = 0.036); systolic CPBP < 115 mmHg as the reference].
Discussion Measurement of BP in community pharmacies is a widespread practice that is recommended by several hypertension societies [18–20]. Nevertheless, some caution should be exercised when considering and using this method; specifically, when normal figures are obtained, CPBP readings might encourage erroneous conclusions about hypertension control in some cases [10,11]. In fact, in this study, CP-MUCH was identified in a quarter of the patients with normal CPBP. Ideally, ABPM and/or HBPM would be recommended in all patients with normal CPBP; however, this is not feasible or achievable in practice. In light of this, our results might help community pharmacists identify individuals who are more likely to present CP-MUCH and, thus, prioritize the use of the out-of-pharmacy BP methods among treated hypertensive patients with normal CPBP. On the basis of the results of the multivariate analysis, CP-MUCH was associated with systolic CPBP and should be investigated when systolic figures equal to or higher than 115 mmHg are obtained. However, because of the low BP cut-off point, this recommendation still encourages a wide use of ABPM and/or HBPM that might not be achievable in some cases (for example, a patient’s lack of willingness to monitor BP). Additional results of this study allow identification of certain circumstances in which either the prevalence of CP-MUCH or the relevance of the situation might be greater. When CP-MUCH was defined by means of ABPM only, the multivariate analysis showed that individuals with systolic CPBP equal to or higher than 123 mmHg are more likely to present CP-MUCH. This is particularly noteworthy when it is considered that ABPM is the reference method in the management of hypertension. In addition, although diastolic CPBP was not associated with CPMUCH, the prevalence of CP-MUCH increased (across any SBP strata) as diastolic CPBP was closer to the uppernormal limits (≥79 mmHg; Fig. 2). Overall, HBP and ABP figures were not remarkably increased in patients with CP-MUCH (Table 1 and Fig. 1). Only individuals with both home and daytime uncontrolled BP showed high SBP figures. Interestingly, four out of these five individuals had systolic CPBP equal to or higher than 123 mmHg. On the basis of these observations taken together, community pharmacists would be encouraged to further assess BP control (using ABPM and/or HBPM)
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
142 Blood Pressure Monitoring 2015, Vol 20 No 3
Fig. 2
47% 50%
36% 33%
43%
20%
20%
25%
≥123 mmHg 0% 0%
5% 115 − 122.9 mmHg
0% ≥79 mmHg 75 − 78 mmHg
