Automated Ambulatory Blood Pressure Devices and Self-Measured Blood Pressure Monitoring Devices: Their Role in the Diagnosis and Management of Hypertension

Technology Assessment

Scope of the Guideline

The following statement represents the American College of Physicians' policy concerning the use of automated ambulatory blood pressure monitoring devices and self-measured blood pressure monitors in the diagnosis and management of hypertension. It is a technology assessment, that is, a science-based evaluation of two blood pressure measuring devices to supplement or replace traditional office-based physician blood pressure measurements. The statement assumes knowledge about the traditional methods of blood pressure measurement and does not provide extensive details to physicians regarding the optimal techniques that should be followed when measuring blood pressure in the office (for example, measuring the blood pressure with the patient sitting, standing, lying down; measuring the blood pressure in both arms).

Hypertension continues to be a clinical and public health problem of impressive magnitude. According to the National Heart, Lung, and Blood Institute, more than 61 million Americans have hypertension. Elevated diastolic blood pressure (office-based blood pressure greater than 90 mm Hg) is a well-recognized independent risk factor for coronary heart disease, renal disease, and cerebrovascular disease. There is convincing evidence that treatment of individuals with elevated blood pressure reduces the number of deaths from stroke and, to a lesser extent, from coronary artery disease. Failing to detect and treat hypertension may thus result in severe consequences, and serious efforts should aim at minimizing the number of unrecognized or uncontrolled hypertensive patients.

On the other hand, false labeling of normotensive persons leads to drug therapy with potential adverse effects and unnecessary costs (1.3). Although the consequences of overdiagnosis are important, they are certainly not comparable with those of underdiagnosis.

Despite major advances in the field of hypertension, methods used for the diagnosis and for the identification of individuals at risk still have their limitations. There is evidence that office blood pressure measurements by physicians may misrepresent a person's usual blood pressure. Several studies have identified a phenomenon called white coat hypertension, whereby blood pressures obtained in an office setting are higher if measured by physicians as opposed to nonphysicians, and whereby physician-measured blood pressures in the office are higher than blood pressures measured outside the office [4.1-4.6]. In view of this phenomenon, a variety of methods for diagnosis and management of hypertension have been proposed to supplement or replace office-based readings by physicians: 1) measurement of blood pressure by technicians or nurses; 2) measurement of blood pressure outside the office by the patient using a standard blood pressure cuff; 3) measurement of blood pressure by automated ambulatory blood pressure monitoring devices; 4) detection of end-organ damage by noninvasive tests [for example, left ventricular hypertrophy by echocardiography or carotid wall thickness by ultrasonography]; 5) integration of information on other cardiovascular risk factors [for example, smoking, diabetes, cholesterol]; and 6) measurement of biochemical factors related to the occurrence of hypertension and its complications (for example, renin) (1.3).

It is clearly important that physicians recognize the existence of white coat hypertension and that effective strategies for the diagnosis and management of hypertensive persons at risk for complications be developed. About two thirds of all hypertensive persons have relatively modest elevations in their blood pressure. When blood pressures measured by physicians in the office are supplemented either by nurses' or technicians' measurements, by measurements made in other less stressful settings, or by using automated ambulatory monitoring, some people do not have sustained increases in their blood pressure (4.1-4.6).

Assessment of the Evidence

Accuracy and Validity of Blood Pressure Measurements

Self-Measured Blood Pressure Monitoring Devices (3.4, 3.5)

Several kinds of devices, using mercury, aneroid, or electronic systems, are commercially available for patients to measure their own blood pressure. Both the accuracy of these devices and the accuracy of patients' record of their blood pressure have been questioned: Many commercially available devices have been shown to provide unsatisfactory accuracy under controlled and field conditions (3.4, 3.5). The appropriate use of these devices has also been challenged: Studies have shown wide variations in patients' ability to accurately record their blood pressure. In addition, studies suggest that most patients require additional training, beyond the manufacturer's instructions. Because this training may require 20 minutes for the average patient, it is clearly important that patients be adequately trained in the self-measurement of blood pressure. Furthermore, whether self-measured blood pressure truly represents an individual's usual blood pressure remains unknown, given that the choice of timing of self-measurement is, in most cases, subjective and tends to be done for convenience, often on arising or in the evening (3.4, 3.5).

Automated Ambulatory Blood Pressure Monitoring Devices (3.6-3.11)

Automated ambulatory blood pressure monitoring uses either auscultatory or oscillometric techniques to measure blood pressure at frequent intervals during a 24-hour period and to record the results for subsequent analysis and interpretation. There are concerns about the accuracy, patient acceptability, and the mechanical reliability of automated ambulatory blood pressure monitoring devices. Quality standards for automated blood pressure monitors have not been clearly defined and are evolving. The collection of satisfactory data and the management and analysis of these data represent complex activities that also have not been standardized, preventing useful synthesis of results from available studies and hindering the systematic evaluation of the accuracy and the validity of automated ambulatory blood pressure monitoring devices. In addition, there is no consensus on the reference blood pressure value, as measured by automated ambulatory blood pressure devices, that would distinguish hypertension from normotension or that would distinguish individuals for whom treatment would significantly reduce morbidity and mortality risks. Indeed, therapeutic blood pressure thresholds have been defined using results from studies in which blood pressure was measured by physicians, nurses, or blood pressure technicians in the office. At this time there are no comparable automated ambulatory blood pressure monitoring data to guide clinicians in their diagnostic and management decisions.

Potential Benefits of Self-Measured Blood Pressure and of Automated Ambulatory Blood Pressure Monitoring Devices

Diagnosis of Hypertension

There is good evidence to suggest that physician-measured office blood pressure overestimates the actual blood pressure in patients, at least in hypertensive patients and perhaps in normotensive patients as well (4.2-4.6). Because they permit frequent monitoring of blood pressure during the course of a usual day of activities, self-measurement of blood pressure and automated ambulatory blood pressure monitoring devices could, in theory, offer a more valid measure of an individual's true blood pressure. In addition, although this has never been formally evaluated, self-measurement of blood pressure may improve compliance through active involvement of patients in their care. Unfortunately, there are no empiric data to guide clinicians in defining the average daily blood pressure at which the benefits of treatment outweigh the harms, or in defining when to measure blood pressure (for example, in early morning, at night, during physical activity). Furthermore, the therapeutic blood pressure threshold, the lowest blood pressure at which a significant advantage of treatment over nontreatment can be demonstrated, has been determined in studies using office-based blood pressure measurements. Similar evaluations of the effectiveness and benefits of treatment guided by either automated ambulatory blood pressures or self-measured blood pressures have not been done.

In summary, prospective analyses of the relationship between office-based blood pressure (obtained by physicians, nurses, or technicians), automated ambulatory blood pressure, or self-measured blood pressure, and the morbidity and mortality risks related to hypertension are needed to determine the specific role of these measuring devices in the diagnosis and management of hypertension.

Prognosis and Risk Prediction

Good evidence supports the association between automated ambulatory blood pressure measurements and left ventricular structural abnormalities. In fact, most studies have found closer correlation between automated ambulatory blood pressure and left ventricular mass than between traditional office-based blood pressure measurement methods and left ventricular mass (5.1-5.5). Caution should be exercised in the interpretation of these results because they are based on cross-sectional studies. Better causal inference about the relationship between blood pressure and left ventricular mass could be made with longitudinal, prospective studies, but this type of evidence does not exist. Only one study has compared the association of office blood pressure, self-measured blood pressure, and automated ambulatory blood pressure with left ventricular hypertrophy. The results showed that only self-measured blood pressure was significantly correlated with left ventricular mass (5.5). More studies are needed to confirm the strength of the association between automated ambulatory blood pressure or self-measured blood pressure and left ventricular hypertrophy.

Most of the evidence linking automated ambulatory blood pressure to cardiovascular morbidity and death is indirect and is inferred from the association between automated ambulatory blood pressure and left ventricular mass; left ventricular mass is an important predictor of cardiovascular morbidity and death in hypertensive patients (6.1). One study directly compared the prediction of cardiovascular risk by automated ambulatory blood pressure and office blood pressure in patients followed longitudinally (6.2). Although the results suggest that automated ambulatory blood pressure may be superior to office blood pressure as a predictor of cardiovascular risk, major methodologic problems threaten the validity of this conclusion: the lack of representativeness of the study population that consisted of patients referred to the investigator's laboratory; the lack of information on the extent of blood pressure control achieved by treatment; the lack of information about important cardiovascular risk factors such as smoking and serum cholesterol levels; the lack of information about the quality and frequency of office-based blood pressure and automated ambulatory blood pressure measurements; and the possibility of biases in ascertaining the outcomes measured. Studies about the relationship between self-measured blood pressure and cardiovascular morbidity and death have not been published.

The evidence linking automated ambulatory blood pressure to renal dysfunction and retinopathy is also inconclusive (5.8, 5.9). Indeed, the four studies that have examined the relationship between automated ambulatory blood pressure and renal dysfunction have yielded results that are inconsistent with each other. The findings from the two studies of the association between automated ambulatory blood pressure and retinopathy also conflict: One study showed a closer correlation between automated ambulatory blood pressure and retinopathy grade than with casual blood pressure, whereas the other study failed to show any difference in the strength of the association between blood pressure and retinopathy, according to the method of blood pressure measurement. Evidence linking self-measured blood pressure to renal dysfunction and retinopathy does not exist.

Management and Monitoring of Treatment Response

Support for the value of automated ambulatory blood pressure in monitoring treatment responses and in reducing the side effects of overtreatment stems from one small observational study and must therefore be considered to be highly tentative at this time (7.1). The daytime automated ambulatory blood pressure of 34 treated hypertensive patients whose office-based blood pressure exceeded 95 mm Hg on two consecutive clinic visits was monitored. These patients were treated to achieve office-based blood pressures of 90 mm Hg or less, the physicians being blinded to the automated ambulatory blood pressure measurements. The results showed that patients with initial automated ambulatory blood pressure less than 90 mm Hg did not experience significant changes in automated ambulatory blood pressure following treatment, whereas those with higher initial automated ambulatory blood pressures did. Such results may imply that patients with elevated office-based blood pressure but normal automated ambulatory blood pressure may be unnecessarily treated and subjected to the cost and side effects of treatment without its benefits. In addition to having an inadequate sample size, the study failed to follow patients for a period long enough to allow the evaluation of clinically critical outcomes such as morbidity or death. Five clinical trials have looked at the role of self-measured blood pressure in improving blood pressure control and medication adherence (7.2). Although the results suggest that this technique may result in reduction of blood pressure and in improved compliance, the studies were conducted on a small sample and their results are not consistent or conclusive. Data from randomized clinical trials or from well-designed prospective cohort studies are needed to evaluate and recommend the use of self-measured blood pressure and automated ambulatory blood pressure in the management of hypertension.

Potential Harms

The side effects of automated ambulatory blood pressure and self-measured blood pressure devices are generally minor: transient arm discomfort and the inconvenience and anxiety related to wearing the devices (8.1). The use of self-measured blood pressure or of automated ambulatory blood pressure devices, or both, may negatively affect health if it results in failure to identify hypertensive individuals who would benefit from treatment. Most randomized clinical trials have used office-based blood pressure measurements to make diagnoses, define treatment strategies, and measure the effectiveness of treatment in controlling blood pressure and in reducing the incidence of stroke and coronary heart disease. The risks of withholding treatment in patients with moderate elevation in office-based blood pressure, but normal-range automated ambulatory blood pressure are unknown. Using data from observational studies, one can only infer that the absolute risks, should one fail to detect and treat an individual with an office-based diastolic blood pressure between 90 and 99 mm Hg, are likely to be small (10.6).

Costs (9.1, 9.2)

The cost for 24 hours of automated ambulatory monitoring depends on the number of procedures that are done each year and on the cost of manpower required. The cost of an automated ambulatory blood pressure monitoring device varies from $2500 to $5000. In addition, fees for maintaining the equipment and fees for technicians and physicians' interpretation of the data need to be factored in. The estimated average cost for 24 hours of automated ambulatory monitoring is about $120. Patients are usually charged between $150 and $450. On the other hand, the cost for a one-time purchase of a self-measurement device is about $75. Patient training and follow-up supervision may add another $25 per year. Devices have an average lifespan of 3 years, so the total average costs per year are estimated to be approximately $50.

The net effects on expenditures of automated ambulatory blood pressure monitoring devices can only be speculative at this time. In view of the lack of effectiveness data, the cost-effectiveness of the use of these devices compared to the standard office-based approaches to hypertension, cannot be thoroughly evaluated. Even though direct costs of the technologies can be obtained, the cost savings or cost increases that could result from widespread adoption of automated ambulatory blood pressure monitoring devices can only be estimated crudely and can be based only on hypothetic assumptions.

Recommendations

Based on the available evidence concerning the effectiveness of self-measured and automated ambulatory blood pressure monitoring devices, the American College of Physicians makes the following recommendations.

Self-measured blood pressure and automated ambulatory blood pressure monitoring devices may, in theory, have a specific role in the diagnosis, prognosis, and management of hypertension. The evidence supporting the role of automated ambulatory blood pressure measurement in the diagnosis and treatment of hypertension is, for the most part, indirect. The major studies showing the benefits of treatment in decreasing the morbidity and mortality risks associated with hypertension have used office-based blood pressure measurements to make diagnoses and to treat and follow patients. Similar studies comparing treatment guided by self-measured blood pressures or automated ambulatory blood pressures to treatment guided by office-based blood pressures are required but have not been conducted. Therefore, the available evidence does not warrant widespread dissemination or routine use of automated ambulatory blood pressure measurement at this time. On the other hand, we support a more circumspect use of such devices for research and for the care of subgroups of hypertensive patients with specific clinical problems.

Self-measured blood pressure devices are increasingly being used by patients, and this practice should be encouraged. However, it should also be clear that there has not been sufficient formal evaluation of this method to warrant managing patients solely using blood pressure readings obtained with self-monitoring blood pressure devices. On the other hand, when used as adjunct to physicians' and nonphysicians' office-based measurements, self-measured blood pressures are an invaluable source of information for the management of hypertension. Patients and physicians electing to use these devices should become knowledgeable about their optimal use. This will involve ensuring that patients are taught how to measure their blood pressure and that they be given advice about when to measure it. The devices should also be checked regularly to ensure that measurements are accurate and reliable and that they correlate to blood pressure readings obtained in the office.

The background paper Automated Ambulatory Blood Pressure Monitoring and Self-Measurement in the Diagnosis and Management of Hypertension was initially commissioned by the Clinical Efficacy Assessment Subcommittee (CEAS). The authors, Lawrence J. Appel, MD, and William B. Stason, MD, submitted their first draft to the CEAS; a revised draft was then submitted to outside peer review to 15 cardiologists whose names were suggested by subspecialty societies. The manuscript was revised and resubmitted to the CEAS. The content of the paper was approved by the subcommittee, but the recommendations drawn from the review of the available scientific evidence differ between the authors and the CEAS. The position paper developed by the CEAS was approved by the Health and Public Policy Committee and the Board of Regents. The position paper and the background paper were both submitted, as individual documents, for publication in the Annals of Internal Medicine.