The Fallacy of Attended Automated Office Blood Pressure Measurement

In 2016, the Hypertension Canada guidelines recommended automated office blood pressure (AOBP) as the preferred method for recording blood pressure (BP) in routine clinical practice. AOBP was also mentioned in the European hypertension guidelines as an option for recording BP in the office and the recent American College of Cardiology/American Heart Association guidelines have recognized the growing evidence supporting its use. The basic principles of AOBP include recording several readings with a fully automated sphygmomanometer while the patient is resting quietly and alone. However, a recent article from the Systolic Blood Pressure Intervention Trial (SPRINT) investigators1 and its accompanying editorial2 have questioned the advantages of patients being “alone” when AOBP readings are recorded. Before discussing the relative merits of being “alone,” it would be useful to review briefly the reasons why AOBP is under serious consideration as a replacement for current methods of recording BP in routine clinical practice. Compared to routine office BP readings, AOBP has a significantly stronger association with the awake ambulatory BP, is consistent from visit to visit, is relatively devoid of white coat effect, is associated with less digit preference (rounding off readings to the nearest zero value) and may have less masked hypertension.3,4 AOBP was also the measurement technique used in the only randomized controlled trial to show that screening for hypertension reduces future cardiovascular events. Based upon comparative BP data and on a clinical outcome study, AOBP has the same threshold for diagnosing hypertension as does awake ambulatory BP and home BP.3,4 According to the evidence, AOBP is the best method for screening patients for possible hypertension, mostly individuals with an office systolic BP of 130–150 mm Hg. However, AOBP may not be the best technique for determining if target systolic BP below 130 mm Hg has been achieved with drug therapy. Indeed, there is currently insufficient evidence to favor one type of BP measurement over another in this BP range.5 The recent retrospective survey of the actual methods used to record BP in SPRINT1 reported that only 4,082 of 9,361 subjects had proper AOBP measurements, with three readings being taken using the automated Omron 907XL recorder with the patient resting alone in a quiet place. Of the remainder, 2,247 had their BP recorded according to the AOBP method, with the only difference being that research staff remained in the room with the patient, without engaging in any conversation, a procedure now called “attended AOBP.” The retrospective analysis of the data from these two groups showed similar mean BP values and similar cardiovascular outcomes, with the implication being that that AOBP may not necessarily require the patient to be alone for accurate readings to be obtained. An apparent objective of the retrospective review of the actual BP methods used in SPRINT was to confirm that the results of the study were applicable to clinical practice, regardless of whether the readings were recorded using the new AOBP method or simply with an automated oscillometric sphygmomanometer. In other words, all that might be needed to apply the findings in SPRINT to patient care was for office staff in the community to record BP according to standard guidelines, including without conversation, but with office staff being present. However, there are reasons why this approach may not be appropriate for usual clinical practice. The American Heart Association first published standards for office BP measurement in 1939 and has been encouraging doctors and nurses to perform proper office BP measurements ever since. At most, these efforts have had limited success, considering that BP readings in routine clinical practice are on average 10/7 mm Hg higher than a research BP recorded according to the guidelines and they also correlate poorly with ambulatory BP, the gold standard for determining future cardiovascular risk according to an individual’s BP status.3 Neither the SPRINT investigators nor Stergiou et al.,2 the authors of the accompanying editorial, presented any new ideas about how a routine BP reading in the community could be transformed into a research quality BP. Instead, Stergiou et al. attempted to enhance their position favoring attended AOBP by including a table, listing five studies which purportedly showed that attended and unattended office BP readings were similar. However, the patients in the studies of Greiver et al.6 and Rinfret et al.7 had mean systolic BP readings of 121.1 and 126.7 mm Hg, respectively. Office readings in this range may be of interest for target BP with antihypertensive therapy, but are different from higher readings where hypertension is suspected, such as a systolic BP of 130–150 mm Hg, in that the low normal BPs are not subject to a white coat effect.5 Another two studies quoted in the editorial,2 Stergiou et al.8 and Al-Karkhi et al.,9 which supposedly compared attended versus unattended AOBP actually compared physicians’ readings with the patients’ own self-measured BP using a patient-activated, semi-automated sphygmomanometer. The self-measured and physicians’ mean BP readings may have been similar, but the assumption that these readings were equivalent to AOBP is faulty. In a variety of studies in different settings, mean AOBP was similar to mean awake ABP and home BP.3,4 In contrast, the mean self-measured BP in the patients of Stergiou et al.8 and Al-Karkhi et al.9 was 9/5 and 7/4 mm Hg higher than their mean home BP, making these self-measured readings the equivalent of standard research readings which, as noted by Stergiou et al.,2 are about 5/5 mm Hg higher than awake ambulatory BP. These self-measured readings were not the same as unattended AOBP. Thus, of the five studies quoted in the editorial,2 only that of Bauer et al.10 involving just 51 subjects actually supported attended AOBP as a possible method for hypertension screening in the office setting. However, it is questionable whether the ideal conditions under which the attended AOBP readings were recorded in this study would be duplicated in routine clinical practice. The editorial by Stergiou et al.2 concluded that “unattended (automated) office BP has no advantage over attended office BP, provided that the latter is taken with an automated device and there is no talking.” They also called this approach “feasible,” which is debatable. The only way to be certain that there is no conversation between office staff and the patient is to have the patient resting alone. Even if staff could be trained not to speak to patients, there is no guarantee that some patients would not speak to the staff immediately before or during the BP measurement, especially if they are nervous. The only data available on (attended) multiple oscillometric office BP readings versus awake ambulatory BP in routine clinical practice is from the Spanish ABPM Registry. In this report,11 the mean office BP in 27,211 hypertensive patients was 160/89 mm Hg compared to a mean awake ABP of 135/78 mm Hg. Replacing manual office BP with a mean of two oscillometric readings clearly did not abolish the white coat effect associated with routine office BP in this setting. Presumably the primary care physicians and nurses in Spain were aware of the standard guidelines for BP measurement, but chose not to follow them in their daily office routine. What is really needed is a shift in the paradigm, namely removing office staff to eliminate conversation, which is a recognized cause of a white coat effect. Furthermore, visits to the physician often involve changes in lifestyle or increases in drug therapy. Knowing this could conceivably lead to patients becoming more anxious in the presence of office staff compared to being alone. Instead of debating whether readings recorded in the presence or absence of office staff in a research study are the same as in clinical practice, we should create an optimum setting for office BP in the community, one which minimizes the likelihood that extraneous factors will affect a patient’s BP, leading to possible misdiagnosis and/or over-treatment of hypertension. The available evidence clearly supports unattended AOBP as the preferred method for office BP measurement. AOBP will always be feasible if compared to readings obtained using standard guidelines, which include 5 minutes of rest in a quiet place before BP is recorded and the use of an oscillometric sphygmomanometer, which can take multiple readings automatically. The current reality is that office BP is usually recorded only once, without much antecedent rest and often using a mercury or aneroid sphygmomanometer. Under these circumstances, some may consider AOBP unfeasible, but this method is not the appropriate comparator. Finally, nobody has ever presented valid reasons why it is better to have a doctor or nurse present in the room when an uncomplicated patient is having BP recorded in routine clinical practice. The idea that nurses and doctors performing attended AOBP will not speak to their patients remains a fallacy. The author declared no conflict of interest. Any views or opinions are those of the author(s) only and not of the AJH or its Editors.