Ambulatory Blood Pressure Monitoring for Every Hypertensive Patient: It's About Time!

THE ARGUMENT FOR—Lawrence R. Krakoff, MD, Professor of Medicine, Mount Sinai School of Medicine and Chief of Medicine, Englewood Hospital and Medical Center, Englewood, NJ

Since the end of the 19th century, measurement of arterial blood pressure (BP) as a basis for diagnosis and treatment of hypertension (HTN) has, in most clinics, relied on the methods described by Riva‐Rocci and Korotkov, namely the stethoscope and mercury column (or equivalent). ¹ Yet clinic BPs, even carefully taken, are not representative of usual or average pressures during ordinary activity. Ayman and Goldshine recognized in 1940 that measurements of pressure at home, outside the clinic, might provide a more accurate assessment of patients' usual BP. ²

Development of new portable technologies for measurement and recording of BP during ordinary activities has been achieved over the past 30 years. These technologies have been used in prospective studies that now establish the superiority of ambulatory BP monitoring (ABPM), compared to clinic pressures, for predicting risk of cardiovascular (CV) disease. ³ , ⁴ , ⁵ The most recent study indicates that a 24‐hour average BP <135 mm Hg systolic pressure is associated with a very low future risk of CV disease, even when adjusted for higher or lower clinic pressures. ⁵ ABPM has been shown to guide the diagnosis of refractory HTN ⁶ and, on occasion, assist in the detection of secondary HTN.

ABPM provides average pressures during the entire day (24 hours) and during various distinct periods within the day, wakefulness, sleep, or upon awakening, when there is a rapid increase in pressure or "surge." This latter period has been associated with greater risk of CV events. Normally, sleep pressures are <10% ³ lower than awake pressures. A less than normal fall in pressure during sleep has been associated with greater risk of stroke in prospective studies. ⁴ In a small series of type 1 diabetic participants, a diminished fall in pressure during sleep was prospectively associated with increased risk of diabetic nephropathy. ⁷ Thus use of ABPM provides valuable diagnostic and prognostic information that can guide management of HTN and related high‐risk conditions, particularly diabetes. These facts alone imply that there is a strong case for making ABPM available to many, if not all, hypertensive patients.

Most of those who manage hypertensive patients are primary care providers, cardiologists, and nephrologists. They may not yet be familiar with ABPM. The technology is new, yet little different and simpler than such widely used tests as Holter electrocardiographic (ECG) monitoring. For ABPM to be implemented, quality control and appropriate training and supervision of personnel are necessary; however, ABPM is now recognized in the United States by the Center for Medicare and Medicaid Services (CMMS) as a legitimate diagnostic procedure that is reimbursable. The primary indication for ABPM is detection of white coat HTN, i.e., HTN that is isolated to the office or clinic. After nearly 30 years of development, ABPM is no longer just a research activity.

Concern about potential costs for ABPM, if included for care of hypertensive patients, has been raised in the past. The CMMS approved limits for charges for ABPM. Actual experience indicates that the total individual charges and reimbursements will be in the range of $55–$95 ⁸ (information provided by Center for Medicare several days to weeks, offers promise. ¹² , ¹³ However, home pressure measurement does not include pressures during sleep and thus omits information that may have value. Use of recording devices that take multiple measurements in the clinic and calculate short‐term average pressures may be helpful for reducing the white coat effect (and potential for measurement error) when compared to a few pressures taken by a physician or nurse. Such devices are being evaluated and compared to home pressures and ABPM. ¹⁴ It is, however, most likely that ABPM will maintain its predominance as the gold standard for diagnosis of HTN until alternative methods have proven their value.

HTN is a disorder of arterial BP that is clearly linked to reversible CV risk. The diagnosis of HTN should be as accurate and precise as modern technology can provide. ABPM has matured from a research exploration to applicable and evidence‐supported technology for improved management of HTN in developed countries. ABPM should be made available to many, if not all, of those with HTN for optimal health care.

THE ARGUMENT AGAINST—William B. White, MD, Professor and Chief, Section of Hypertension & Clinical Pharmacology, Pat and Jim Calhoun Cardiology Center, University of Connecticut School of Medicine, Farmington, CT

Given the considerable consensus in the scientific literature and clinical community and the guidance thereof, hypertension (HTN) diagnosed by ambulatory blood pressure (BP) recordings has been recently considered as a superior tool to clinical measurements in clinical trials of antihypertensive drugs as well as for the assessment of hypertensive patients in clinical practice. ¹ While numerous studies have demonstrated that ambulatory BP is the best means to identify patients with suspected white coat HTN, the method is not necessarily required for the evaluation of every general hypertensive patient.

Methods of clinical BP measurement improve precision and reduce the need for ambulatory BP monitoring in many patients

Poor Measurement Technique

One of the pitfalls in BP measurement in the clinic is a lack of attention to important details. Small modifications in measuring technique could lead to a substantial improvement in precision. For example, it is imperative that the BP be measured in a supported arm at the level of the heart to avoid the effects of isometric exercise (from the patient) and hydrostatic pressure (from improper arm position) on the arterial pressure. ² Additionally, neither the patient nor the observer should be talking or moving during the actual measurement as it will raise the BP and could give the impression of an elevated BP in a patient who is otherwise nor‐motensive or well‐controlled on antihypertensive therapy. The patient should be seated comfortably with both feet on the floor. The area of the upper arm where the cuff and bladder are placed should not be clothed, nor should it have a tight band of material from a shirt or blouse obstructing venous flow. Attention to these simple details will result in a more accurate BP in the medical environment.

Terminal Digit Bias

Another concern for accurate office BP measurement is terminal digit bias. Numerous studies have shown that clinical observers will conveniently round off both the systolic and diastolic BP to zero rather than use 2, 4, 6, or 8 mm Hg as the terminal digit. This phenomenon of "zero bias" may occur in as many as 70%–90% of BP readings in the office setting and is due to a lack of attention and forgetfulness on the part of the observer. In a recent retrospective evaluation in our own HTN practice, ³ terminal digit preference favored 0, 2, and 8 mm Hg, but there was only a 30% incidence of zero digit bias (Figure 1). Nevertheless, the terminal digit was not randomly distributed, so even HTN specialists can improve in their technique!

Figure 1.

Terminal digit bias in a hypertension faculty practice based on results from 103 patients. Adapted from J Hum Hypertens. 2003;17:819‐822. ³

DATA SUPPORTING THE USEFULNESS OF AMBULATORY BP MONITORING IN CERTAIN PATIENTS

Ambulatory BP values have been shown to provide reasonably accurate information regarding a patient's risk of cardiovascular (CV) events as compared with office BP measurements alone. None of these studies, however, has been able to demonstrate at precisely what values events are negligible or clinically acceptable. In the instance of evaluating cardiac target organ disease, Grandi et al. ⁴ showed that left ventricular mass index could be mildly elevated in patients with white coat HTN even when the 24‐hour BP was quite normotensive (Table). Admittedly, the findings in this particular study are contrary to numerous cross‐sectional trials published in the past, but they do give the clinician concern about the robustness and validity of analyses that evaluate patients at just one point in time.

Prospective Outcome Studies

The most well‐designed studies in this field are from Verdecchia et al., ⁵ , ⁶ , ⁷ , ⁸ which are, in general, used as the basis for most of the guidelines and/or recommendations by specialty societies. While these studies have substantial value, none of them had randomization of patients to a group in whom management or treatment decisions were made based on ambulatory BP as compared with clinical BP. Verdecchia et al. have shown in a prospective cohort study design that men with BP <136/87 and women with BP <131/86 on ambulatory BP monitoring did not have increased CV risk, as compared with patients with normal office BPs. ⁵ The follow‐up period in this study, however, was <'5 years. Fagard et al. ⁶ studied 353 subjects in a randomized, double‐blind, placebo‐controlled trial in which he demonstrated that subjects with non‐sustained HTN were at a lower risk of a poor outcome than subjects with sustained HTN, but a greater risk than normotensive individuals. The study was too small to determine at which level of BP one could safely avoid antihypertensive therapy.

The Office vs. Ambulatory study (OvA) ⁸ was a prospective cohort study in =2000 treated hypertensive patients that was initiated in 1988 and completed in 2002. The purpose of this study was to evaluate whether CV events were more likely to be predicted by the 24‐hour BP as compared with the clinic BP. Physicians treating the patients were not guided by the results of the ambulatory BP values. As is shown in Figure 2, the ambulatory BP was a superior predictor of CV outcomes when compared with the office (or clinic) BP. Nevertheless, as the level of the clinic BP increases, it is noteworthy that the event rates increase even when the 24‐hour BP is <135/85 mm Hg, which suggests that this value is probably too high and that the goal of therapy should be significantly lower, especially in patients with underlying CV morbidities or diabetes mellitus.

Certainly if a patient has a 24‐hour ambulatory BP>135/85, he/she is likely to be at increased risk and should be treated with antihypertensive therapy. Little is known, however, about ambulatory BP values between 125‐135/75‐85 mm Hg. Thus treatment may still be needed (Figure 3), and the ambulatory BP values have not reduced the likelihood of drug therapy but have increased the cost of care.

Figure 3.

Use of ambulatory blood pressure (ABP) in hypertension management. Algorithm for use of ABP in clinical practice by the author. BP=blood pressure; ABPM=ambulatory blood pressure monitoring. Adapted from N Engl J Med. 2003;348:2377–2378. ¹

Thus, while ambulatory BP monitoring will provide more information about a patient's BP behavior in a variety of situations, CV risk in many patient populations, such as individuals with type 2 diabetes mellitus, kidney disease coronary heart disease, and cerebrovascular disease, have not been determined based on ambulatory BP vs. other forms of BP measurement.

INSURANCE COVERAGE OF AMBULATORY BP MONITORING

Unfortunately, coverage of ambulatory BP monitoring in clinical practice is relatively sparse by the various insurance programs in the United States. Medicare has been covering the study since April of 2002 for beneficiaries who have a suspected diagnosis of white coat HTN. ⁹ A national coverage decision on other indications is not available at this time and probably will not be until more data become available as to how ambulatory BP monitoring affects patient management. The one situation that should be covered is for evaluation of the high‐risk patient whose office BP is variable despite complex antihypertensive therapy. These patients require tight BP control; the ambulatory BP recording is helpful in making the diagnosis of uncontrolled HTN, especially toward the end of a drug's dosing period.

WHO NEEDS AN AMBULATORY BP MONITORING STUDY AND WHO DOES NOT?

Since the 24‐hour BP measurement does give an enhanced prediction of CV risk for some patients, ambulatory BP monitoring clearly has a role in the diagnosis of patients with: 1) white coat HTN in the untreated state; 2) patients with marked variability in the BP in both the office and home setting, making diagnosis extremely difficult; and 3) selected, treated patients with significant comor‐bidities who would be at great risk if the true BP was not optimal. In contrast, newly‐diagnosed patients with severe HTN, especially those with overt target organ involvement, diabetes, and coronary or cerebrovascular events, need antihypertensive drug therapy, which should not be delayed. Patients who have had prior ambulatory BP studies in which the office or self‐monitored BP values are similar to the daytime BP values can be followed by the more simple and less expensive office‐based measurement. Finally, we do not believe that patients whose office and home BPs are elevated to a similar level require ambulatory BP recordings since it is likely that the 24‐hour BP is abnormally high as well.