| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Hypertension. 2004;43:10.)
© 2004 American Heart Association, Inc.
Scientific Contributions |
From the Department of Preventive Medicine and Epidemiology, Stritch School of Medicine, Loyola University (K.W.-M., R.S.C., H.K.), Maywood, Ill; Departamento de Medicina Preventiva y Salud Pública, Facultad de Medicina, Universidad Autónoma de Madrid (J.R.B.), Madrid, Spain; Instituto Superiore di Sanità, Laboratorio di Epidemiologia e Biostatistica (S.G.), Rome, Italy; the Department of Community Health and Epidemiology, Faculty of Medicine, Dalhousie University (M.R.J.), Halifax, Nova Scotia, Canada; the Cardiovascular Studies Unit, Imperial College of Science, Technology, and Medicine (N.P.), London, UK; the Department of Epidemiology and Public Health, University College London Medical School (P.P.), London, UK; the Department of Medicine, University Hospital (B.S.), Umeå, Sweden; and the Robert-Koch Institut (M.T.), Berlin, Germany.
Correspondence to Dr Katharina Wolf-Maier, Department of Preventive Medicine and Epidemiology, Stritch School of Medicine, Loyola University, 2160 S First Ave, Maywood, IL 60153. E-mail k.wolf-maier{at}t-online.de
| Abstract |
|---|
|
|
|---|
10% in European countries had their blood pressure controlled. At the 140/90 mm Hg cutpoint, two thirds to three quarters of the hypertensives in Canada and Europe were untreated compared with slightly less than half in the United States. Although guidelines vary among countries, resulting in different case definitions, this does not account entirely for the varying success of different national control efforts. Low treatment and control rates in Europe, combined with a higher prevalence of hypertension, could contribute to a higher burden of cardiovascular disease risk attributable to elevated blood pressure compared with that in North America.
Key Words: hypertension, detection and control antihypertensive therapy blood pressure monitoring models, statistical cross-sectional studies
| Introduction |
|---|
|
|
|---|
Hypertension has the unusual attribute of being sufficiently common to represent a public health concern, yet its control depends primarily on the successful treatment of individual patients by physicians. A crucial step in this process is therefore the explicit recommendations given to medical professionals and the practical consequences of these treatment strategies in particular health systems. Vigorous debate continues over appropriate thresholds for initiating treatment of high BP.6 During the last decade, risk stratification, with the intent of focusing attention on persons with high levels of CVD risk, has been promoted by specialists, decision makers in health systems, and professional advisory bodies.711 This approach has led to conflicting international treatment thresholds, however, that might in turn lead to differing levels of hypertension treatment and control. The degree of intensity of screening for target-organ damage in persons at low to medium risk will also lead to variable case definition.12,13 The public health impact of these various standards for evaluation and treatment urgently needs to be evaluated.
Unlike most medical conditions, community surveillance has been the most common approach to evaluating the success of efforts to treat and control high BP. Although surveys are not a perfect evaluation tool, they are necessary to obtain information about persons who are unaware that they have hypertension or are not compliant with medical advice. During the past decade, many countries have conducted large-scale, national health surveys to determine the prevalence and treatment of hypertension in addition to other CVD risk factors.1417 When measurements are comparable, these national surveys can also be used to make international comparisons.18 This information might provide insights into ways to improve public health strategies to prevent target-organ damage. The aim of this study was to use original data from surveys to compare levels of hypertension treatment and control in the United States, Canada, England, Germany, Sweden, Italy, and Spain.
| Methods |
|---|
|
|
|---|
|
Data Selection
BP Measurements
The mercury sphygmomanometer was used for BP measurements in every country except for England, where the Dinamap 8100 was used. All studies had at least 2 measurements, and the second BP was used to create the mean for the age-gender groups. Further details are presented in the earlier report of the prevalence findings.19
Hypertension Treatment and Control
Hypertension was defined by 2 standard criteria, namely, systolic BP (SBP)
160 or diastolic BP (DBP)
95 mm Hg; SBP
140 or DBP
90 mm Hg; or current use of antihypertensive medication. Awareness was defined as answering "yes" to the question: "Have you ever been told that you had high blood pressure?" Treatment was defined as current use of antihypertensive medications. In England, medication use was documented, whereas in all other surveys it was based on self-report. Control was defined as a BP <160/90 or <140/90 mm Hg among medicated hypertensives. The control rate (or, more precisely, the proportion) was the number of treated hypertensives with a BP <160/90 or <140/90 mm Hg divided by the total number of hypertensives. Control in treated hypertensives was defined as the number of controlled hypertensives divided by the number of treated hypertensives. Nonpharmacologic therapies have begun to receive greater attention, and some individuals might have controlled their elevated BP without drugs; we had inconsistent information in these samples to incorporate those interventions into the analyses.
Treatment Guidelines
Treatment guidelines vary among the studied countries and over time (Table 2). In the United States, irrespective of risk status, persons with an SBP or DBP of 140 or 90 mm Hg or greater are presently candidates for treatment, and among patients with CVD, renal disease, or diabetes, treatment is recommended for a BP
130/85 mm Hg (Joint National Committee IV). At the time of National Health and Nutrition Survey (NHANES) III, however, the guidelines were less stringent, with less emphasis on SBP. The Canadian guidelines in force in the 1990s recommended treatment at 160/100 mm Hg in "low-risk" individuals, decreasing to 140/90 mm Hg in patients with diabetes or renal disease.9 A variety of European guidelines have been promulgated, although most are broadly consistent with the World Health Organization/International Society for Hypertension approach that sets 150/95 mm Hg as the threshold in low-risk individuals, decreasing to 130/85 mm Hg in those with diabetes or renal disease.7,8,10 Revisions of these guidelines were adopted at various points in time relative to the surveys as well. Ideally, calculation of treatment and control should be based on the guideline in force at the time of data collection. Although that approach would yield historically accurate information within a given country, it would defeat the purpose of a comparison that requires a common standard across countries. To make the analysis relevant to current practice, we have therefore chosen to use the 2 most widely applied threshold values in discussions of treatment and control. We recognize that in several instances, 140/90 mm Hg will not have been the stated goal in a given country. Contrariwise, many persons with BPs <160/95 mm Hg will have been started on treatment by physicians, thus falsely elevating the apparent control rate when 160/95 mm Hg is used as the threshold. As a sensitivity analysis of the impact of risk stratification on treatment thresholds, we applied the national guidelines in force in Canada and the risk stratification in England and Spain to the data from those countries. Finally, to estimate the burden of uncontrolled, high BP in a category of patients who were included under all guidelines, we calculated the prevalence of BP >160/95 mm Hg; these results were calculated by dividing the number of persons with BP >160 or 95 mm Hg, regardless of treatment status, by the total number of participants.
|
Data Analysis
Hypertension prevalence and treatment and control rates were age-adjusted by averaging the 10-year age-gender groups. To achieve maximum overlap of data from the available surveys, we restricted the analysis to 35 to 64 years for age-adjusted results and used the age range 35 to 74 for age-specific data, omitting Spain from the category >65 years. In the United States, NHANES III data on whites, blacks, and Hispanics were combined with the appropriate weighting for population size. The English sample was risk-stratified to apply the local treatment algorithm that defines subgroups of hypertensives by using a model derived from the Framingham Heart Study, and the participants were divided at a 10-year CHD risk threshold of 15%.24,25 The Canadian and Spanish samples were also risk-stratified by using a global risk algorithm.9
| Results |
|---|
|
|
|---|
Hypertension Treatment
Modest heterogeneity was observed in age- and gender-adjusted treatment rates at both the 160/95 and 140/90 mm Hg cutpoints among the studied European countries (Tables 3 and 4
). Based on the current standard of 140/90 mm Hg, England had the lowest level of treatment (25%), followed by Sweden and Germany (both 26%), Spain (27%), and Italy (32%). Treatment of hypertension was highest in the United States (53%), followed by Canada (36%).
|
|
Women in all countries were more likely to be treated than men; gender differences were especially strong in the United States and Canada (44% in men vs 63% in women in the United States and 28% in men vs 45% in women in Canada, based on 140/90 mm Hg; Table 4). Heterogeneity was apparent in the gender-age treatment patterns in the United States compared with all other countries (data not shown). Although low levels of treatment were noted in Europe in the younger age groups of men, they approximated those in the United States by age 65. Among women, the proportion of hypertensives receiving treatment was reasonably constant across the age range in the United States, whereas it rose rapidly in other countries, reaching 50% in the elderly.
Hypertension Control
In the United States, two thirds of the hypertensive population had their BP controlled at the 160/95 mm Hg threshold, and the corresponding figure was 49% in Canada (Table 3 and Figure 1). Within Europe, control ranged from 23% in Spain to 38% in England. Gender differences were modest overall, favoring women in all countries. The percentage of hypertensives reported to be on treatment and having a BP <140/90 mm Hg was much lower in all European countries, ie,
10%. In contrast, 17% and 29% of hypertensives in Canada and the United States, respectively, were at the treatment goal (Table 4 and Figure 2). Control was 2-fold higher in women compared with men in Spain, Italy, Canada, and the United States. Consistent with treatment patterns, the proportion of hypertensive men at this threshold whose BP was controlled increased markedly with age, especially in the United States, where it rose from 9% at age 35 to 44% to 30% at age >65 years (Figure 3).
|
|
|
Hypertension control at the 140/90 mm Hg threshold among women in the youngest age group ranged from 5% to 13% in the European countries, 21% in Canada, and 36% in the United States (Figure 4). Levels of control among older women (65 to 74 years) were highest in the United States (37%), whereas levels of control in Canada were similar to those in Europe (5% to 17%).
|
The primary obstacle to an unbiased comparison across countries is the large number of persons with a pretreatment BP between 140/90 mm Hg and 160/95 mm Hg. These persons will be designated as "controlled" if the threshold of 160/95 mm Hg is applied, when they did not meet the case criteria, thus falsely elevating the rate. To eliminate this source of bias, a complementary analysis was undertaken to determine the proportion of study participants whose BP was uncontrolled at the 160/95 mm Hg threshold (ie, had BP higher than this value, whether treated or not). In the United States, the prevalence of uncontrolled hypertension at this level was 5.3% compared with 7.4% in Canada and 12% to 25% in Europe.
Although differences in criteria for diagnosis and thresholds for treatment alter case definitions, less variation would be anticipated in the level of control among treated cases. In fact, this was generally the case (Tables 3 and 4
). Although control in treated patients was still highest in the United States (54%), rates in Canada and England were not much lower (47% and 40%, respectively), but Germany (30%), Italy (28%), and especially Spain (19%) and Sweden (21%) had lower results at the 140/90 mm Hg threshold.
Control in Hypertensives With Low and High CVD Risk
Based on most non-US guidelines, low-risk patientsgenerally defined as those with a BP <150/95 mm Hg without CVD risk factorswould not be candidates for treatment. Unfortunately, not all surveys had the CVD risk measurements required to apply risk stratification algorithms. The potential impact of excluding low-risk patients was examined in Spain, England, and Canada as illustrative examples by using the definitions in force in each of the specific countries.14,16,17 In Spain and England, control in low-risk hypertensives was 8% and 10% at the 140/90 mm Hg threshold, whereas these values were 1% and 4%, respectively, in those classified as high risk. In Canada, on the other hand, the comparable control rate was 15% for persons at low risk and 12% among those at high risk. Thus, high-risk patients (ie, those targeted by local guidelines) actually had lower levels of treatment and control in Spain and England compared with Canada.
| Discussion |
|---|
|
|
|---|
We were not able to determine whether treatment and control levels differed by presence of specific CVD risk factors or diabetes/kidney disease. However, we demonstrated that patients designated as high risk by the treatment guidelines in England, Spain, and Canada had lower levels of BP control compared with those at low risk. This seemingly paradoxical result demonstrates the complexity of using community surveillance as a method of assessing the success of CVD prevention strategies. At least 2 explanations can be offered. First, many of the lower-risk patients are women, who are always better treated. Second, among the high-risk are the elderly and those with higher pretreatment BPs, which are inherently more difficult to control. Previous studies have documented lower levels of control in older hypertensive men at high risk compared with those at low risk27 and similarly low levels of hypertension treatment and control among high-risk patients with CHD in several European countries.28 After risk stratification, patients at high risk would need to be treated at a BP threshold lower than the threshold for those at low risk and consequently, should be better controlled. Of course, restricting treatment to only high-risk individuals is more cost-effective when measured at the level of the individual patient encounter. It might well be, however, that from the perspective of a health system that is already seeing patients for other reasons, adding treatment to those at lower risk adds little marginal cost. Concentrating only on high-risk patients will certainly limit the public health impact of BP control, because low-risk persons account for most of the population-attributable risk. As noted here, the United States has a much lower proportion of uncontrolled hypertensives and stroke rates that are about half of those in Europe,29 which might in part be a result of a broader treatment strategy.
Community surveys do not determine hypertension status on the same basis as clinical guidelines. Most surveys measure BP on only one occasion; depending on the analysis used, this has been shown to yield many false-positive hypertensive cases and a smaller number of false-negatives. However, using the BP
140/90 mm Hg cutpoint, the net effect of misclassification is small; in NHANES III, for example, the prevalence of hypertension declined only 2% from the first to the second visit. Of course, net reduction in the number of hypertensive cases will translate into a correspondingly higher control rate. The validity of these comparisons between countries therefore depends critically on the comparability of the survey methods. Variation in mean BP levels in these surveys as a result of methodology has been discussed in detail in an earlier report.19 Moreover, the time frame over which the studies were performed varied slightly, as did the date of implementation of various guidelines. Given the timing of the surveys and the varied introduction of guidelines, however, it would not have been possible to conduct a comparative analysis without using a single standard. Future studies are needed to assess the impact of introducing new guidelines on control rates. Although these aspects limit the accuracy of our results, the overall differences between the United States, Canada, and Europe are large and unlikely to change over relatively short periods of time. Moreover, the data are consistent with other published epidemiologic literature and CVD mortality rates in the sampled countries.28,29
The attributes of persons unlikely to have their hypertension controlled has been extensively examined.27,3033 Lack of control was much more common in relation to SBP compared with DBP.30,31 The presence of other CVD risk factors was not predictive of treatment success in the Framingham Heart Study population, in conformity to our results.32 Reduction of BP to "optimal" levels could prevent up to 50% of CHD events,33 and it is frustrating to recognize that despite enormous attention to this problem, only modest gains might have been made in treatment and control during the last decade in the United States.34 The large between-country variation described here, however, provides a perspective on the range of what is feasible, suggesting that organized interventions within the health care system can yield large, positive results.
It is unlikely that any of the countries studied have reached the maximum level of treatment and control that can be attained. For example, BP control at <140/90 mm Hg was achieved in routine practice settings among two thirds of 33 000 participants in a recent clinical trial.35 Hypertensives who are at high risk of CVD but who have not yet suffered an event represent a crucial opportunity for clinical prevention.5 To further reduce CVD secondary to hypertension, more efforts are needed in the public at large and among health care providers to increase the awareness of the associated risk and benefits of treatment. In addition, system-wide approaches can help physicians achieve greater success in following current guidelines. Possible ways to increase adherence to guidelines include education by respected personnel, implementation of reminder systems, outlining of vital recommendations from the public health perspective, and improved presentation by user-friendly format and annual updates.36
Perspectives
Much of the variation in the success rate in hypertension control reported here appears to be attributable to the treatment strategy adopted by individual countries. The structure of financial incentives within these systems could additionally play an important role in this health care outcome. Although further intervention efforts, as outlined earlier, need to be pursued within individual health systems, the implications of broader versus more restrictive guidelines also need to be examined.
| Acknowledgments |
|---|
Received June 12, 2003; first decision July 7, 2003; accepted October 16, 2003.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
K. Fox Benefits of perindopril all along the cardiovascular continuum: the level of evidence Eur. Heart J. Suppl., September 1, 2008; 10(suppl_G): G4 - G12. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. Ingelsson, P. Gona, M. G. Larson, D. M. Lloyd-Jones, W. B. Kannel, R. S. Vasan, and D. Levy Altered Blood Pressure Progression in the Community and Its Relation to Clinical Events Arch Intern Med, July 14, 2008; 168(13): 1450 - 1457. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Scobie, S. Keyhani, P. L. Hebert, and M. A. McLaughlin Response to Gender and Blood Pressure Control Hypertension, June 1, 2008; 51(6): e49 - e49. [Full Text] [PDF] |
||||
![]() |
S. Antonopoulos, S. Kokkoris, S. Gerakari, S. Mikros, T. Nitsotolis, D. Vikeli, P. Korantzopoulos, and G. Giannoulis Comparison of Monotherapy Versus Combination Antihypertensive Therapy in Elderly Patients With Essential Hypertension Angiology, May 1, 2008; 59(2): 230 - 235. [Abstract] [PDF] |
||||
![]() |
J. Ma and R. S. Stafford Screening, Treatment, and Control of Hypertension in US Private Physician Offices, 2003-2004 Hypertension, May 1, 2008; 51(5): 1275 - 1281. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. P. Rossi, M. Bolognesi, D. Rizzoni, T. M. Seccia, A. Piva, E. Porteri, G. A.M. Tiberio, S. M. Giulini, E. Agabiti-Rosei, and A. C. Pessina Vascular Remodeling and Duration of Hypertension Predict Outcome of Adrenalectomy in Primary Aldosteronism Patients Hypertension, May 1, 2008; 51(5): 1366 - 1371. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Weikert, H. Boeing, T. Pischon, C. Weikert, A. Olsen, A. Tjonneland, K. Overvad, N. Becker, J. Linseisen, A. Trichopoulou, et al. Blood Pressure and Risk of Renal Cell Carcinoma in the European Prospective Investigation into Cancer and Nutrition Am. J. Epidemiol., February 15, 2008; 167(4): 438 - 446. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Turchin, M. Shubina, A. H. Chodos, J. S. Einbinder, and M. L. Pendergrass Effect of Board Certification on Antihypertensive Treatment Intensification in Patients With Diabetes Mellitus Circulation, February 5, 2008; 117(5): 623 - 628. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. A. Preston, P. Harvey, O. Herfert, G. Dykstra, J. W. Jukema, F. Sun, and D. Gillen A Randomized, Placebo-Controlled Trial to Evaluate the Efficacy, Safety, and Pharmacodynamic Interaction of Coadministered Amlodipine and Atorvastatin in 1660 Patients With Concomitant Hypertension and Dyslipidemia: The Respond Trial J. Clin. Pharmacol., December 1, 2007; 47(12): 1555 - 1569. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Addo, L. Smeeth, and D. A. Leon Hypertension In Sub-Saharan Africa: A Systematic Review Hypertension, December 1, 2007; 50(6): 1012 - 1018. [Abstract] [Full Text] [PDF] |
||||
![]() |
N. N. Qureshi, J. Hatcher, N. Chaturvedi, T. H Jafar, and Hypertension Research Group Effect of general practitioner education on adherence to antihypertensive drugs: cluster randomised controlled trial BMJ, November 17, 2007; 335(7628): 1030 - 1030. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Rieder Getting into a healthy `CV success zone': effective strategies to prevent CVD Eur. Heart J. Suppl., May 1, 2007; 9(suppl_B): B4 - B7. [Full Text] [PDF] |
||||
![]() |
F. J. He and G. A. MacGregor Blood pressure is the most important cause of death and disability in the world Eur. Heart J. Suppl., May 1, 2007; 9(suppl_B): B23 - B28. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Vaidyanathan, C. Reynolds, C.-M. Yeh, M.-N. Bizot, H. A. Dieterich, D. Howard, and W. P. Dole Pharmacokinetics, Safety, and Tolerability of the Novel Oral Direct Renin Inhibitor Aliskiren in Elderly Healthy Subjects J. Clin. Pharmacol., April 1, 2007; 47(4): 453 - 460. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Redon, L. Cea-Calvo, J. V. Lozano, J. C. Marti-Canales, J. L. Llisterri, J. Aznar, J. Gonzalez-Esteban, and on behalf of the investigators of the PREV-ICTUS S Differences in Blood Pressure Control and Stroke Mortality Across Spain: The Prevencion de Riesgo de Ictus (PREV-ICTUS) Study Hypertension, April 1, 2007; 49(4): 799 - 805. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. S. Cooper Using Public Health Indicators to Measure the Success of Hypertension Control Hypertension, April 1, 2007; 49(4): 773 - 774. [Full Text] [PDF] |
||||
![]() |
M. Rienstra, D. J. Van Veldhuisen, H. J.G.M. Crijns, I. C. Van Gelder, and for the RACE investigators Enhanced cardiovascular morbidity and mortality during rhythm control treatment in persistent atrial fibrillation in hypertensives: data of the RACE study Eur. Heart J., March 2, 2007; 28(6): 741 - 751. [Abstract] [Full Text] [PDF] |
||||
![]() |
Y. R. Wang, G. C. Alexander, and R. S. Stafford Outpatient Hypertension Treatment, Treatment Intensification, and Control in Western Europe and the United States Arch Intern Med, January 22, 2007; 167(2): 141 - 147. [Abstract] [Full Text] [PDF] |
||||
![]() |
Y. R. Wang Lack of Effect of Guideline Changes on Hypertension Control for Patients With Diabetes in the U.S., 1995-2005 Diabetes Care, January 1, 2007; 30(1): 49 - 52. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. R. Banegas, J. Segura, J. Sobrino, F. Rodriguez-Artalejo, A. de la Sierra, J. J. de la Cruz, M. Gorostidi, A. Sarria, L. M. Ruilope, and for the Spanish Society of Hypertension Ambulatory Effectiveness of Blood Pressure Control Outside the Medical Setting Hypertension, January 1, 2007; 49(1): 62 - 68. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. Zhang, L. Thijs, and J. A. Staessen Blood Pressure Lowering for Primary and Secondary Prevention of Stroke Hypertension, August 1, 2006; 48(2): 187 - 195. [Full Text] [PDF] |
||||