Antecedent rest may not be necessary for automated office blood pressure at lower treatment targets

Abstract

In SPRINT (Systolic Blood Pressure Intervention Trial), use of the Omron 907XL blood pressure (BP) monitor set at 5 minutes of antecedent rest to record BP produced an automated office BP value 7/6 mm Hg lower than awake ambulatory BP at 27 months. The authors studied the impact on automated office BP of setting the Omron 907XL to 0 minutes instead of 5 minutes of rest in patients with readings in the lower normal BP range, similar to on‐treatment BP in the SPRINT intensive therapy group. Patients (n = 100) in cardiac rehabilitation were randomized to three BP readings at 1‐minute intervals using an Omron 907XL BP device set for 5 or 0 minutes of antecedent rest. Mean (±standard deviation) automated office BP (mm Hg) after 5 minutes of rest (120.2 ± 14.6/66.9 ± 8.6 mm Hg) was lower (P < .001/P < .01) than without rest (124.2 ± 16.4/67.9 ± 9.1 mm Hg). When target BP is in the lower normal range, automated office BP recorded without antecedent rest using an Omron 907XL device should be higher and closer to the awake ambulatory BP, compared with readings taken after 5 minutes of rest.

1. INTRODUCTION

Automated office blood pressure (AOBP) is becoming the preferred method for office blood pressure (BP) measurement, with its advantages over conventional office BP being recognized in recent guidelines.1, 2 The principles of AOBP are based on the premise that there is no reason to have medical personnel present when multiple office BP readings are being recorded with a fully automated, electronic sphygmomanometer. Leaving the patient alone eliminates any possibility of conversation and tends to reduce patient‐related anxiety, which might increase BP, resulting in a white‐coat effect and overdiagnosis of hypertension.

The use of AOBP or its equivalent to identify persistent hypertension on drug therapy in SPRINT (Systolic Blood Pressure Intervention Trial) has raised concern that BP measurement with this technique might preclude application of the study's results to clinical practice.3, 4 Critics have focused on the findings in the SPRINT ABPM Substudy,5 which compared office BP on antihypertensive therapy at 27 months with the awake ambulatory BP (ABP). Whereas office BP is generally considered to be higher than the awake ABP, in the intensive treatment group in SPRINT, the mean AOBP (119.7/65.9 mm Hg) on treatment at 27 months was 6.8/6.1 mm Hg lower than the awake ABP (126.5/72.0 mm Hg). Subsequent examination of the actual method of office BP measurement in SPRINT and a review of other studies comparing office BP with awake ABP in the lower range of normal have shown that office BP is consistently lower than the awake ABP, regardless of the measurement technique.6

In view of these findings, we decided to compare AOBP as recorded in SPRINT using the Omron 907XL BP monitor set for 5 minutes of antecedent rest with AOBP obtained without any rest, in the BP range achieved in patients in the intensive therapy group, whose target systolic BP was <120 mm Hg. The specific objective of the study was to determine the extent to which changing the setting of the Omron 907 device for the rest period before the readings from 5 to 0 minutes would increase AOBP. Depending on the magnitude of the increase, the 0‐minute setting might be more appropriate for recording AOBP when systolic BP readings are close to target in the low normal BP range.

2. METHODS

The two techniques for recording AOBP, with and without 5 minutes of rest before taking three BP readings, were compared using a randomized, crossover, study design involving a convenience sample of patients enrolled in an urban outpatient cardiac rehabilitation program. Criteria for enrollment included men and women 50 years and older and current attendance in the cardiac rehabilitation program during recovery from myocardial infarction, percutaneous coronary intervention, or cardiac surgery (coronary bypass surgery and/or valve surgery). Study approval was obtained from the University Health Network/Toronto Rehabilitation Institute research ethics board. Patients who were unable to provide informed consent and individuals attending other programs were excluded.

Cardiac rehabilitation patients were invited to participate during one of their on‐site classes. Once informed consent was obtained, eligible participants were instructed not to exercise before their next scheduled weekly visit and to avoid caffeine 1 hour before the BP measurements. During this visit, AOBP was recorded using the Omron 907XL sphygmomanometer as follows:

• Technique 1: Omron 907XL set for 5 minutes of rest followed by three AOBP readings recorded at 1‐minute intervals (method used in SPRINT).

• Technique 2: Omron 907XL set for 0 minutes of rest followed by three AOBP readings recorded at 1‐minute intervals.

The order of performing techniques 1 and 2 was randomized to avoid introducing any bias caused by an “order effect.”

Patients were taken to a quiet clinic room where the Omron 907XL sphygmomanometer was attached with the correctly sized cuff matched to their upper arm circumference. The patients were tested in a seated position, with their back supported, legs uncrossed, and arms supported at the level of the heart. The recorder was set to automatically time the period of rest (0 or 5 minutes) with 1‐minute intervals between the three BP readings. After activating the device, the research assistant left the room, returning after the three readings were completed. The same device was used for each within‐person visit. Demographic data and clinical characteristics were then obtained followed by the second set of AOBP readings after the alternate period of 0 or 5 minutes of rest.

The sample size calculation of 100 participants was based on a 90% α power at a β of 0.05 to detect differences in systolic AOBP of 2 to 3 mm Hg between the two techniques. The sample size and planned analyses were consistent with similar studies.7, 8 Statistical analysis was performed using IBM SPSS (version 22.0) and MedCalc Statistical Software (version 16.8.4). The primary outcome, systolic and diastolic BP, was calculated for each set of AOBP readings. Shapiro‐Wilk tests were completed in the initial inspection of data to determine normality. Bland‐Altman plots were completed to examine the agreement between the AOBP measurements using the two techniques and to detect any evidence of bias related to the absolute levels of BP. Differences between mean systolic and diastolic AOBP values were assessed using paired t tests for both “order effect” and “no order effect” measurements. Sociodemographic characteristics were self‐reported and clinical characteristics were extracted from medical records. Statistical significance was defined as P < .05.

3. RESULTS

A total of 108 patients were screened for possible enrollment, with 100 consenting to participate in the study. The demographic characteristics of the patients are shown in Table 1. Participants were aged a mean (±standard deviation) of 68.4 ± 8.6 years and had a mean body mass index of 28.7 ± 6.0. Most patients were receiving at least one drug that could lower BP. The research assistant measured each patient's upper arm circumference and BP readings were recorded using the following cuff sizes with number of patients shown in parentheses: small (2), standard (76), and large (22).

Table 1.

Demographic characteristics of the study population

Participants (N = 100)	Baseline measures
Men, No.	70
Women, No.	30
Age, mean±standard deviation, y (range)	68.4 ± 8.6 (50–87)
Body mass index, kg/m2	28.7 ± 6.0
Ethnicity
White	74
Asian	11
Other	15
Cardiac history
Percutaneous coronary intervention	42
Myocardial infarction	17
Aortocoronary bypass	16
Aortic valve replacement	9
Other	16
Risk factors
Hyperlipidemia	78
Hypertension	46
Diabetes mellitus	20
Blood pressure–lowering medications
β‐Blocker	65
Angiotensin‐converting enzyme inhibitor	50
Calcium channel blocker	27
Angiotensin receptor blocker	5

In the cohort of 100 patients, mean systolic AOBP recorded with 0 minutes of rest was 124. 2 ± 16.4 mm Hg compared with a mean systolic AOBP of 120.2 ± 14.6 mm Hg recorded after 5 minutes of rest (Table 2), with the 4.0‐mm Hg difference being statistically significant (P < .001). Similarly, for diastolic BP, the mean reading recorded after 0 minutes of rest (67.9 ± 9.1 mm Hg) was significantly (P < .01) higher than the mean reading after 5 minutes of rest (66.9 ± 8.6 mm Hg). There was no significant difference in heart rate between the two sets of readings (Table 2).

Table 2.

BP and heart rate values after 0 vs 5 minutes of rest

All patients (N = 100)	0 Min of rest	5 Min of rest	P values
Systolic BP, mean±SD, mm Hg	124.2 ± 16.4	120.2 ± 14.6	<.001
Diastolic BP, mean±SD, mm Hg	67.9 ± 9.1	66.9 ± 8.6	<.01
Heart rate, mean±SD, beats per min	63.3 ± 10.3	62.6 ± 9.3	.14
AOBP technique (with order effect)	5 Min of rest first	0 Min of rest second
Systolic BP, mean±SD, mm Hg	121.4 ± 13.7	122.7 ± 14.4	.22
Diastolic BP, mean±SD, mm Hg (n = 46)	67.6 ± 9.6	67.5 ± 9.7	.86
AOBP technique (with order effect)	0 Min of rest first	5 Min of rest second
Systolic BP, mean±SD, mm Hg	125.6 ± 17.9	119.2 ± 15.3	<.001
Diastolic BP, mean±SD, mm Hg (n = 54)	68.2 ± 8.5	66.3 ± 7.8	<.001

AOBP indicates automated office blood pressure; BP, blood pressure; SD, standard deviation.

P values denote significant differences between 5 minutes of rest vs 0 minutes of rest.

A significant order effect on the systolic AOBP values was present for the 5 minutes of rest setting, with readings 2.2 mm Hg lower (P < .001) when 5 minutes of rest was second. Mean systolic AOBP was 2.9 mm Hg lower when 0 minutes of rest was second. A significant order effect was also seen for diastolic BP using the 5 minutes of rest setting (1.3 mm Hg; P < .001), whereas the difference was only 0.7 mm Hg for the 0 minutes of rest (Table 2).

Systolic BP data were plotted according to the method of Bland and Altman (Figure). The anticipated differences between the two sets of office BP readings were seen. There was a negative bias of −4.0 mm Hg for the 0 minutes vs 5 minutes of rest BP readings, but there was no evidence of any bias related to the absolute level of BP. The differences in systolic BP were similarly distributed in lower and higher ranges of BP.

Figure 1.

Bland‐Altman plot of differences between the 5‐minute rest period and the 0‐minute rest period for systolic blood pressure (BP) readings (y axis) vs the mean of the two measurements (x axis) for all 100 patients. The solid line indicates the mean difference between 5‐minute and 0‐minute values, and the dashed lines indicate limits of agreement (±1.96 standard deviation).

4. DISCUSSION

In the SPRINT intensive therapy group, mean AOBP was 6.8/6.6 mm Hg lower than the awake ABP,5 which was an unexpected finding, since office BP is generally considered to be higher than awake ABP, especially in the high‐normal BP or hypertensive range. It is now apparent6 that all recognized techniques for office BP measurement may produce readings that are lower than the awake ABP when readings are below a systolic BP of 130 mm Hg. This finding suggests that prolonged periods of antecedent rest when using automated sphygmomanometers, such as the Omron 907 XL, may not be required if office BP at target is to approximate the awake ABP, a recognized gold standard for predicting the risk of future cardiovascular events in relationship to an individual's BP status. It is generally recognized that recording BP without any antecedent rest results in higher readings. However, there are limited data on the difference in readings when AOBP is recorded using the Omron 907XL device with and without 5 minutes of rest, especially when mean BP is below the newer target systolic BP value <130 mm Hg. Moreover, it is now important to investigate AOBP readings recorded with the Omron 907XL device, since the manufacturer of the BpTRU device is no longer in business. One might anticipate seeing greater use of the Omron 907XL monitor and other similar devices specifically designed for recording AOBP.

Accordingly, we sought to determine whether performing AOBP using the Omron 907XL device with the 0 minutes of rest setting might result in a higher mean BP than with the setting for 5 minutes of rest. The results of the present study confirm this hypothesis, in that mean BP was 4.0/1.0 mm Hg higher without any antecedent rest. There are several reasons why this finding is important. At readings below target BP, the 0 minutes of rest technique should lead to AOBP readings that are closer to the awake ABP. Also, AOBP without a rest period can be performed in less than half the time it took to obtain the readings in SPRINT. Moreover, AOBP using the 0 minutes setting also records BP in less time than it takes for attended manual or electronic BP, which requires 5 minutes of rest before at least two readings are recorded.

A crossover study design was used in this study to account for possible differences in mean BP when readings were taken first vs second. In this instance, mean BP was slightly lower when readings were taken second, but this factor was accounted for by the randomization process.

In a previous study,7 office BP readings recorded with the Omron 907XL were compared with AOBP readings obtained using a BpTRU sphygmomanometer. When readings were recorded in 50 patients with the Omron 907XL device set for three readings at 1‐minute intervals with 0 minutes of antecedent rest, the mean AOBP was 141 ± 24/73 ± 14 mm Hg compared with a mean AOBP of 139 ± 23/74 ± 14 mm Hg with the BpTRU set for five readings at 1‐minute intervals without rest. Thus, the mean systolic AOBP values for the two devices were similar, although the systolic BP using the Omron 907XL was slightly higher, which is consistent with its smaller number of readings. Even though this difference is only 2 mm Hg, it somewhat ameliorates the lower AOBP readings seen in patients with treated hypertension at target BP.7

5. STUDY LIMITATIONS

One limitation of our study is that 24‐hour ABP was not performed. However, since the relationship between AOBP and awake ABP in this lower normal BP range has already been described,6 the findings in the present study should be relevant to clinical practice. In addition, even though not all of our study population had hypertension, the majority of the patients were receiving drugs that lower BP, with some individuals taking these for other cardiac conditions. However, there is no obvious reason why the 0 minutes of rest setting of the Omron 907XL would not result in higher readings than with 5 minutes of rest in the lower normal range of a systolic BP in different types of patients receiving antihypertensive medications.

6. CONCLUSIONS

When screening for possible hypertension, the Omron 907XL should be used with 5 minutes of rest before three readings are recorded at 1‐minute intervals, as was performed in SPRINT. However, when readings in treated patients decrease to a systolic BP <130 mm Hg, the Omron 907XL may be used without any antecedent rest, if the mean reading is to reflect more closely the awake ABP. Doing so would reduce the time required to record the office BP and remove the unfounded concern that AOBP takes too long to record BP.

CONFLICT OF INTEREST

There are no conflicts of interest to disclose on behalf of the authors.

Colella TJF, Tahsinul A, Gatto H, Oh P, Myers MG. Antecedent rest may not be necessary for automated office blood pressure at lower treatment targets. J Clin Hypertens. 2018;20:1160–1164. 10.1111/jch.13319