A meta-analysis helps to clarify the use of automated office blood pressure in clinical practice.

Abstract

The methodological details of the office blood pressure (OBP) measurement are crucial for the interpretation of blood pressure (BP) values and have been largely neglected in clinical research and in practice.1 Therefore, we thank Martin Myers for his comments on our meta-analysis, as he highlighted important methodological aspects of OBP measurement.2, 3 Our work aimed to examine the net effect of the “presence” of an observer during “automated” OBP measurement. We did not examine casual “auscultatory” OBP measurement, as this method induces too many observer-related issues which are not due to the “presence” of the observer per se. Therefore, our meta-analysis included only studies that used automated devices for both unattended and attended OBP, and the same device and measurement protocol for both, and performed all measurements in the same subjects. Indeed, in 4 of 10 studies included in our meta-analysis OBP readings were obtained using semi-automated BP monitors, which require pressing a button to start each measurement (fully automated monitors take several measurements with a single press of the button).4-7 In these four studies, the pooled difference between unattended and attended OBP was −0.3 (95% CI: −2.1, 1.6)/−0.1 (−1.2, 1.1) mm Hg (systolic/diastolic). In three of these studies,6, 7 home BP was also recorded providing lower values compared to unattended OBP. However, in all of these studies OBP was based on triplicate measurements. In a recent meta-analysis by Myers, the pooled difference between systolic daytime ambulatory BP and unattended OBP was 0.3 mm Hg although heterogeneity reached 90%.8 However, in most of the studies showing daytime ambulatory BP to be higher than unattended OBP, the latter was performed using the BpTRU device which performs six OBP measurements and calculates the average after discarding the first reading.8 The number of readings and the exclusion of the first reading influence the determination of OBP level. Studies using the BpTRU device have shown the first reading to be about 6 mmHg higher than the average of the succeeding readings even at systolic BP levels about 130 mm Hg.9 Interestingly, there was a trend toward higher attended than unattended OBP in studies with higher OBP levels.2 However, in clinical practice OBP differences are more likely to affect treatment decisions in subjects with OBP levels close to the diagnostic thresholds, rather those with higher OBP levels. Four studies have specifically addressed the impact of the presence of an observer on OBP by using the SPRINT study measurement protocol.10-13 Three of these studies showed minor differences in the range of <3 mm Hg,10-12 and only one showed attended systolic OBP to be higher by 8.6 mm Hg.13 In conclusion, we fully agree with Martin Myers that for the accurate diagnosis and optimal management of hypertension OBP must be carefully standardized and complemented by out-of-office BP measurement. Multiple factors may affect the OBP levels, including the rest period, observer presence, subject posture, device type, cuff size, number of readings, talking, discarding first reading, etc1 As Martin Myers mentioned, scientific organizations have been promoting guidelines for proper OBP measurement for decades. It is now time for scientific organizations to recommend automated BP monitors for the “screening” method for hypertension (OBP), as indeed they do for the “diagnostic” method (ambulatory and home BP monitoring). Attended automated OBP avoids most of the issues of casual OBP measurement, and unattended OBP is even more likely to ensure a standardized OBP evaluation. No conflict of interest.