Reply to Dr Myers' commentary on the use of automated blood pressure machines in office blood pressure measurements.

Abstract

We reviewed the commentary1 by Dr Myers wherein he suggests that casual automated office blood pressure (AOBP) should replace resting manual office blood pressure (MOBP), and that casual readings taken with AOBP are similar to awake ambulatory blood pressure monitoring (ABPM).1 Dr Myers suggests that the absence of resting for AOBP readings in the Conventional Versus Automated Measurement of Blood Pressure in the Office (CAMBO) trial2 accounts for the discrepancy between those results and our data.3 Indeed, a recent study by Nikolic and colleagues4 reported that office BP does decrease with resting time (by −4.1 mm Hg from 5 to 10 minutes). However, we would like to point out that the resting period did not change the precision, which was remarkably poor, in both the CAMBO trial and our study.2, 3 Correlation coefficients and average bias data do not paint a complete picture when one is comparing two different methods.5, 6 The primary outcome of our study was the Bland-Altman analysis, and we reported wide limits of agreement (−31, +33 mm Hg) between AOBP and ABPM,3, 5 which were quite similar to that reported by Dr Myers in the CAMBO trial2 (−31.9, +33.6 mm Hg) despite the difference in the AOBP resting period. The scatter plots in our study (Figure 1)3 and in the CAMBO study2 (Figure 2) are indeed strikingly similar; however, with different conclusions being drawn. We therefore find it curious that Dr Myers has chosen to ignore the wide limits of agreement between AOBP and ABPM as assessed by Bland-Altman analysis and continues to promote AOBP as a surrogate for daytime ABPM.1 While clearly not a surrogate for ABPM when subjected to rigorous statistical testing, AOBP is endorsed by a number of national professional organizations in their guidelines for the diagnosis and management of office hypertension. According to these guidelines (National Institute for Health and Clinical Excellence [NICE],7 European Society of Hypertension [ESH],8 and Canadian Hypertension Education Program [CHEP],9 to name just a few), AOBP should be measured in a similar manner as MOBP. We, therefore, take issue with Dr Myers' claim that AOBP measurements should not be preceded by a period of rest ostensibly because it led to an underestimation of daytime ABPM in those studies where proper office resting technique was used.1 Here, again, the reader needs to be reminded that just because the average bias was less with nonresting AOBP, this does not mean that there was any improvement in diagnostic accuracy. Except for data from one study, which was a post hoc analysis in a substudy of a larger trial using unconventional definitions,10 the published literature suggests that compared with MOBP, the mean BP from a series of readings by an AOBP device is lower not only among individuals with white-coat hypertension, but also among individuals who do not have the white-coat effect.3, 4, 11, 12 This downward bias means that the price for a partial elimination of white-coat effect is an increased number of missed patients with masked hypertension.3 Rather than changing time-tested guidelines for the diagnosis of office hypertension, we suggest that a more accurate understanding of the role of AOBP be promoted.