Design
Editorial
Answers are grounded in this paper and cite their sources.
This commentary highlights significant discrepancies in blood pressure readings between different ambulatory blood pressure monitors and emphasizes the need for rigorous validation against mercury standards.
For treating patients with hypertension, organizations such as the UK's National Institute for Health and Clinical Excellence (NICE) state that ambulatory blood pressure monitoring (ABPM) should be offered to patients with suspected hypertension.1 ABPM has been shown to provide a better predictive value for target organ damage2 and vascular events compared with clinic blood pressure (BP) measurement3 although the reproducibility of nocturnal BP fall by ABPM seems limited.4 In this month's issue, Kallem and colleagues5 reported their comparison of ambulatory BP (ABP) readings from two monitors (Mobil-O-Graph [I.E.M., Stolberg, Germany] and Spacelabs 90207 [Spacelabs Medical, Issequah, WA]) worn at the same time by 12 untreated adults. The Spacelabs monitor recorded a 7.9-mm Hg higher systolic pressure compared with the Mobil-O-Graph in the randomized condition outside of the confines of a clinical setting. In 1989, Santucci and colleagues6 compared the data obtained with two different ABP monitors from 17 normotensive and 28 hypertensive patients. Both devices were satisfactory and showed good agreement in a clinical setting.6 Although these two studies5, 6 used different conditions for the BP measurements, we recognize that the results reported by Kallem and associates5 bring up an important issue regarding the accuracy of ABPM devices. Several points should be taken into account when interpreting the results of the Kallem study. The first point is the manner in which the number of patients was determined. The study population (12 adults in each nonrandomized or randomized condition) was small, and thus the results of the study might be overinterpreted. The second point is that no data regarding the mercury standard or intra-arterial BP were reported in the study. It is thus not possible to discern which of the two ABPM devices was more accurate. Few studies have compared an ABP monitoring sequence using two monitors worn simultaneously by adult patients in a randomized study design. The data presented by Kallem and colleagues thus make an important contribution, provided they are considered within the context of the study's limitations. Validation studies have been conducted for Mobil-O-Graph7 and Spacelabs 90207,8 but it has not been confirmed whether those validations were accurate. For example, most validation protocols that have been introduced for automated BP measuring devices recommend procedures carried out at rest only.9 O'Brien and colleagues9 indicated that many factors could influence the ABPM validation procedure, including compliance to the protocol being employed; the accuracy of the standard; establishing precisely the model being validated; the influences of BP level, age, and exercise on device accuracy; the provisions necessary for special populations, such as pregnant women, the elderly, and children; the influence of oscillometric vs Korotkoff sound detection and electrocardiographic gating on comparative measurements; the assessment of performance as distinct from accuracy; and the relevance of general factors such as the algorithm being employed and computer compatibility. In fact, Czarkowski and coworkers10 determined that venous blood redistribution affects the accuracy of the oscillometric method of BP measurement (Spacelabs 90207) using lower-body negative pressure to simulate changes of body position. They found that although the Spacelabs 90207 remained in the A class according to British Hypertension Society (BHS) protocol, the device's accuracy was altered by venous blood redistribution, invoking the question of whether such an alternation constitutes an additional limitation of oscillometric ABPM. In addition, Rossen and colleagues11 pointed out an important error in the Spacelabs Report Management System 92506 software. The numerical differences between the printout and on-screen hourly averages were considerable, ranging from −37 mm Hg to 18 mm Hg systolic BP and from −16 mm Hg to 10 mm Hg diastolic BP. There are now numerous ABPM devices available, and it is more important than ever that the devices for measuring ABPM are accurately validated. In a recent systematic review, Hodgkinson and coworkers12 assessed methodological quality by determining whether validation protocols had been used and followed correctly. From 5420 journal articles identified, 108 met the inclusion criteria for the review. Excluding studies that assessed monitors that are no longer in use, 40 relevant studies of 21 different monitors were reviewed. Thirty-eight (95%) studies used a validation protocol, and 28 of these studies assessed a monitor in the general population. The protocols “passed” the validation in 24 of the 28 studies, but 12 of the 24 (50%) reported a difference of at least 5 mm Hg systolic BP between the test device and the reference standard for ≥30% of the readings. In addition, only a minority of the 28 ABPM validation studies conducted in the general population have correctly reported the validation according to the relevant protocol, and half of these included major protocol violations likely to have biased the results.12 Taken together, the results described above suggest that assessing the evidence of monitor accuracy is not straightforward. Furthermore, most monitors that “passed” validation showed significant variation in BP from the reference standard, highlighting the inadequacies in older validation protocols. Overall, the pass rate in studies using the European Society for Hypertension International Protocol (ESH-IP)13 would have dropped from 83% to 50% if the ESH-IP2 criteria14 had been used instead. In fact, in 3 studies the Spacelabs 9020715-17 and in two studies the Mobil-O-Graph18, 19 passed validation but would have failed if they were assessed using the ESH-IP2 accuracy criteria. What's done is done. As it stands, the paper by Kallem and colleagues5 (admittedly derived from a small study population) adds to the literature raising serious questions about the accuracy of pooling ABPM data obtained from a variety of ambulatory devices. To resolve the issues regarding the accuracy of ABPM devices, studies of large numbers of patients in which mercury standard measurements are also performed are needed. The authors declare no conflict of interest.
No takes yet. Share an insight, caveat, or question.
Nagai et al. (2013) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: