Abstract
Guidelines recommend 1–2 minutes between repeated, automated office-based blood pressure (AOBP) measures, which is a barrier to broader adoption. Patients from a single hypertension center underwent a 3-day evaluation that included a 24-hour ambulatory BP monitor (ABPM) and one of two, non-randomized, unattended AOBP protocols. Half of patients underwent 3 AOBP measurements separated by 30 seconds and the other half underwent 3 BP measurements separated by 60 seconds. All measurements were compared to the average awake-time BP from ABPM as well as the first AOBP measurement. We used linear regression to assess whether the 30-second protocol was associated with individual or average AOBP measurements or awake-time ABPM and used an interaction term to determine whether interval modified the relationship between AOBP measurements (individual and mean) with awake-time ABPM. Among 102 patients (mean age 59.2±16.2 years, 64% women, 24% Black) the average awake-time BP was 132.5±15.6/77.7±12.2 mmHg among those who underwent the 60-second protocol and 128.6±13.6/76.5±12.5 mmHg for the 30-second protocol. Mean SBP/DBP was lower with the 2nd and 3rd AOBP measurement by −0.5/−1.7 mmHg and −1.0/−2.3 mmHg for the 60-second protocol versus −0.8/−2.0 mmHg and −0.7/−2.7 mmHg for the 30-second protocol; protocol did not significantly modify these differences. Differences between AOBP measurements (1st, 2nd, or 3rd) and awake-time ABPM were nearly identical across protocols. In conclusion, a 30-second interval between AOBP measurements was as accurate and reliable as a 60-second interval. These findings support shorter time intervals between BP measurements, which would make AOBP more feasible in clinical practice.
Keywords: unattended automated office-based blood pressure measurement, clinic-based blood pressure measurement, ambulatory blood pressure monitoring, hypertension
Blood pressure (BP) measurement is one of the most common clinical assessments performed during ambulatory visits.1 Automated office blood pressure (AOBP) with 3 replicate BP measurements is increasingly recognized as a superior approach to minimize human error and improve measurement precision during office visits.2,3 However, a major barrier to clinical adoption of AOBP is the time it takes to perform in clinic. In addition to a 5-minute period of rest prior to AOBP, guidelines recommend at least a 1–2 minute pause between each of the three measurements.4–6 As a result, AOBP requires a minimum of 7 minutes of rest time beyond cuff placement, instructions, and cuff inflation and deflation. Altogether, AOBP can add nearly 10 minutes to the BP measurement process. While this time requirement is comparable to well-performed auscultatory BP measurement,7 saving even 1 minute from this time requirement has implications for the scalability of AOBP in busy clinical practices.
Recommendations for intervals between BP measurements are largely informed by non-automated, auscultatory devices.8 One study of automated oscillometric home devices suggested that intervals shorter than 1 minute (i.e., 10-second) were less accurate than 1-minute intervals.9 However, this finding has not been replicated using automated oscillometric devices in the office setting. Moreover, few interval studies have evaluated the accuracy of measurement using 24-hr ambulatory blood pressure measurement (ABPM) as the gold-standard referent measurement.10,11 Notably the National Health and Nutrition Examination Survey (NHANES) use 30-seconds intervals between cuff inflations, although evidence for this approach is not described.12
In this series of patients from a single hypertension center, we compared a 30-second AOBP interval to a 60-second AOBP interval in relation to awake-time 24-hour ABPM measurements to assess its relative accuracy and its implications on the classification of stage 2 or uncontrolled hypertension (systolic BP [SBP] ≥ 140 mm Hg, diastolic BP [DBP] ≥ 90 mm Hg). We also compared the 2nd and 3rd AOBP to the first AOBP measurement to assess reliability. We hypothesized that measurement performance (accuracy, the classification of hypertension, and reliability) would be comparable with a shorter time interval.
Methods
The data supporting the findings of this study are available from the corresponding author upon reasonable request with institutional approval.
Population
We tested two different AOBP intervals among 103 patients of the Beth Israel Medical Center (BIDMC) Hypertension Center at Healthcare Associates between August 1st, 2019 and January 21, 2021. The Hypertension Center at Healthcare Associates is a multidisciplinary clinic of physicians, pharmacists, nurse practitioners, and licensed practical nurses embedded within the hospital-based academic general medicine ambulatory practice. Patients are referred from within the broader general medicine, cardiology, and geriatric practices for complex hypertension management and the performance and interpretation of ABPM. This project was determined by the BIDMC Institutional Review Board to be exempt research as a quality improvement initiative. Analyses were restricted to the 102 patients, (of the 103) who completed all BP assessments.
Ambulatory Blood Pressure Measurement
Patients at the Hypertension Center underwent a standardized measurement protocol over a three-day period. On Day 1, patients were brought to a private clinic exam room where the circumference of their arm was measured using a tape measure to determine cuff size. Cuff sizes were confirmed using the range indicators printed on the cuff when available (Spacelabs or Omron). In general, these assessments were performed using the non-dominant arm, unless there was a medical contraindication (tremor, prior mastectomy, continuous blood glucose monitor, etc).
Patients were then fitted with a 24-hr ABPM (SpaceLabs 90227 or 90217A-1 of Spacelabs Healthcare, Snoqualmie, WA, USA), which they wore over the ensuing 25 hours. Patients removed the monitor on Day 2 and returned with the device two days later for interpretation on Day 3. ABPMs were programmed to perform measurements every 20 minutes during the day and hourly during sleep. ABPMs were pre-programmed based on patients’ self-reported sleep and awake times to enable the device to change measurement frequency at night. In the process of active measurement, patients were asked to remain still and let their arm rest in their other arm, on an adjacent table, or across their heart. Participants were provided a log to record sleep and awake times as well as relevant activities (exercise, meals, timing of BP medications, driving periods) during the 24-hr period. The actual sleep and awake times were used to determine average awake ABPM. Only participants with a minimum of 20 successful awake ABPM values were included in this study consistent with guidelines.13
Automated Office Blood Pressure Measurement
AOBP was always obtained on the day the ABPM device was returned (Day 3), generally using the same arm. There were no changes made to patients’ hypertension regimen after ABPM was started and before AOBP was performed. Patients were assigned 1 of 2 protocols in consecutive fashion. The first group of patients underwent the 30-second interval protocol (August 1, 2019, to January 2, 2020), while the second group of patients underwent the 60-second interval protocol (January 9, 2020, to January 21, 2021). AOBP was generally performed during a morning visit prior to January 15, 2020 and during an afternoon visit after January 15, 2020 based on unrelated changes to the Hypertension Center schedule.
Using an Omron HEM907XL (Omron Healthcare Inc., Lake Forest, IL, USA), patients underwent an unattended, pre-programmed 5 minute delay followed by three BP measurements (designated M0-M2), separated by either a 30-second or 60-second interval between each measurement. Prior to measurement, they were positioned with their feet flat on the ground, their backs supported, and their palms facing upright. Their arms were supported and elevated at heart level on an adjacent table. We posted signage outside the clinic room to indicate the start and stop times for the BP measurement to avoid interruptions and maintain quiet. Patients were not able to see the BP monitor’s screen during measurements. Staff left the exam room after starting the monitor and returned after the third measurement was complete.
Other Covariates
Demographic information (age, sex, race) were confirmed during the patient visit or abstracted from the patient record. As part of the clinical care during the patient interview, patients were asked about prior diagnoses of diabetes, cardiovascular disease, and sleep apnea as well as current antihypertensive medications, alcohol intake (current status, number of days drinking per week and number of drinks per day), and smoking status (never, current, former). Body mass index and creatinine were abstracted from the most recent available data in the medical record. We categorized reasons for referral including: new diagnosis of hypertension, uncontrolled hypertension, white coat hypertension, orthostatic symptoms, or other.
Statistical Analysis
Population characteristics were determined with means and proportions overall and according to 60-second or 30-second protocols. The distributions of each AOBP measurement and awake-time ABPM for SBP and DBP were visualized using kernel density plots. We determined the mean and standard deviation for SBP and DBP by measurement type according to protocol; we also determined the prevalence of SBP ≥ 140 mm Hg, DBP ≥ 90 mm Hg, or either elevation (SBP ≥ 140 mm Hg or DBP ≥ 90 mm Hg, i.e. stage 2 or uncontrolled hypertension) according to protocol. All means and proportions were compared to the mean awake-time ABPM as a reference. Means were compared using paired t-tests and reported as mean differences, while proportions were compared using McNemar’s χ2 test. We also determined root mean square errors (RMSE) for each measurement.
We determined the difference between the first and subsequent AOBP assessments and plotted by protocol using generalized estimating equations (normal family, identity link, robust variance estimator). We also examined the association of the 30-second (versus the 60-second protocol) with each BP measurement adjusted for age, sex, and race (Model 1) or additionally adjusted with the difference between the first automated measurement and awake-time ABPM (Model 2), using linear regression. Using similar models, we examined the interaction between the 30-second (versus the 60-second) protocol and each BP measurement to assess for modifications to the association between each measurement and awake-time ABPM as the dependent variable. In sensitivity analyses, we additionally adjusted models for statin use and sleep apnea, patient characteristics with the largest magnitude imbalances between protocols at baseline.
All analyses were performed using Stata 15.1 (StataCorp LP, College Station, TX).
Results
Population
Of the 102 patients, 64% were women and 24% were Black with a mean age of 59.2 (SD, 16.2) years (Table 1). Over 80% were referred for evaluation of hypertension (new, uncontrolled, or white coat). Other characteristics were similar between protocols with the exception of sleep apnea (17.6% among the 30-second protocol versus 7.8% among the 60-second protocol) and statin use (29.4% among the 30-second protocol versus 47.1% among the 60-second protocol).
Table 1.
Baseline Characteristics
| Overall population |
60 second intervals |
30 second intervals |
||||
|---|---|---|---|---|---|---|
| N | Mean (SD) or % | N | Mean (SD) or % | N | Mean (SD) or % | |
| Age, yr, Mean (SD) | 102 | 59.2 (16.2) | 51 | 60.7 (16.7) | 51 | 57.7 (15.7) |
| Women, % | 102 | 63.7 | 51 | 64.7 | 51 | 62.7 |
| Black, % | 102 | 23.5 | 51 | 23.5 | 51 | 23.5 |
| Detailed race, % | ||||||
| White | 102 | 66.7 | 51 | 68.6 | 51 | 64.7 |
| Black or African American | 102 | 23.5 | 51 | 23.5 | 51 | 23.5 |
| Other | 102 | 9.8 | 51 | 7.8 | 51 | 11.8 |
| Hispanic or Latino, % | 102 | 6.9 | 51 | 3.9 | 51 | 9.8 |
| Body mass index, kg/m2, Mean (SD) | 102 | 27.9 (4.9) | 51 | 27.8 (5.1) | 51 | 27.9 (4.6) |
| Last creatinine, mg/dL, Mean (SD) | 98 | 0.9 (0.2) | 49 | 0.9 (0.2) | 49 | 0.9 (0.2) |
| Diabetes, % | 102 | 16.7 | 51 | 13.7 | 51 | 19.6 |
| Cardiovascular disease, % | 101 | 5.0 | 50 | 6.0 | 51 | 3.9 |
| Sleep apnea, % | 102 | 12.7 | 51 | 7.8 | 51 | 17.6 |
| Current statin use, % | 102 | 38.2 | 51 | 47.1 | 51 | 29.4 |
| Number of blood pressure medications | ||||||
| 0 | 102 | 51 | 51 | 52.9 | 51 | 49.0 |
| 1 | 102 | 17.6 | 51 | 19.6 | 51 | 15.7 |
| 2 | 102 | 14.7 | 51 | 11.8 | 51 | 17.6 |
| 3 | 102 | 10.8 | 51 | 9.8 | 51 | 11.8 |
| 4 or more | 102 | 5.9 | 51 | 5.9 | 51 | 5.9 |
| Smoking status, % | ||||||
| Never | 102 | 67.6 | 51 | 70.6 | 51 | 64.7 |
| Former | 102 | 28.4 | 51 | 27.5 | 51 | 29.4 |
| Current | 102 | 3.9 | 51 | 2.0 | 51 | 5.9 |
| Currently drinking alcohol, % | 102 | 46.1 | 51 | 37.3 | 51 | 54.9 |
| Average number of days drinking, Mean (SD) | 102 | 1.5 (2.1) | 51 | 1.5 (2.4) | 51 | 1.5 (1.9) |
| Average drinks per day, Mean (SD) | 101 | 0.8 (1.0) | 51 | 0.5 (0.7) | 50 | 1.0 (1.1) |
| Reason for referral, % | ||||||
| New hypertension | 102 | 30.4 | 51 | 31.4 | 51 | 29.4 |
| Uncontrolled hypertension | 102 | 19.6 | 51 | 19.6 | 51 | 19.6 |
| White coat hypertension | 102 | 31.4 | 51 | 35.3 | 51 | 27.5 |
| Orthostatic symptoms | 102 | 2.9 | 51 | 2.0 | 51 | 3.9 |
| Other | 102 | 15.7 | 51 | 11.8 | 51 | 19.6 |
Mean Blood Pressure and Proportion with Hypertension
The distributions of SBP and DBP AOBP and awake-time ABPM measurements are shown in Figure 1A–D. The distributions of SBP values across measurements were more consistent with the 60-second protocol than the 30-second protocol. In contrast, the distribution of DBP values were more consistent with the 30-second protocol than the 60-second protocol. Notably, for SBP, inconsistencies in distribution were mainly seen with M0, a measurement not influenced by interval.
Figure 1.
Kernel density plots of the distribution of (A) systolic blood pressure from the 60-second protocol, (B) systolic blood pressure from the 30-second protocol, (C) diastolic blood pressure from the 60-second protocol, and (D) diastolic blood pressure from the 30-second protocol. ABPM-awake (solid, black line) is the awake-time blood pressure from the 24-hour ambulatory blood pressure monitor. M0 (solid, non-black line) is the first automated office blood pressure measurement. M1 (dash line) is the second automated office blood pressure measurement. M2 (dotted line) is the third automated office blood pressure measurement.
We compared measurements and the prevalence of stage 2 hypertension from either protocol to daytime ABPM (Table 2). The mean SBP based on ABPM awake-time measurements among those on the 60-second protocol was 132.5 (SD, 15.6) mm Hg and the mean DBP was 77.7 (SD, 12.2) mm Hg (Table 2). For those on the 30-second protocol, the mean SBP based on ABPM awake-time measurements was 128.6 (SD, 13.6) mm Hg and the mean DBP was 76.5 (SD, 12.5) mm Hg. Greater differences in SBP were observed between AOBP and awake-time ABPM during the 30-second protocol compared to the 60-second protocol. This was evident for all measurements (M0-M2), including those not influenced by interval. Both variance estimates and root mean square errors were greater among SBP measurements obtained with the 60-second protocol versus the 30-second protocol, while the reverse was observed for DBP measurements.
Table 2.
Comparison of mean blood pressure measurements, N = 102
| Systolic blood pressure | Mean (SD) | Mean Absolute Difference | P | RMSE | % SBP≥140 | Difference | P |
|---|---|---|---|---|---|---|---|
| 60 second intervals, N=51 | |||||||
| M0 | 137.7 (20.3) | 5.2 | 0.04 | 13.2 | 37.3 | 7.8 | 0.29 |
| M1 | 137.2 (19.2) | 4.7 | 0.059 | 13.4 | 45.1 | 15.7 | 0.039 |
| M2 | 136.7 (19.8) | 4.2 | 0.084 | 13.1 | 37.3 | 7.8 | 0.34 |
| Average M1 & M2 | 136.9 (19.1) | 4.4 | 0.064 | 13.1 | 39.2 | 9.8 | 0.18 |
| Average M0-M2 | 137.1 (19.2) | 4.6 | 0.055 | 13.1 | 37.3 | 7.8 | 0.29 |
| ABPM awake | 132.5 (15.6) | Ref | Ref | Ref | 29.4 | Ref | Ref |
| 30 second intervals, N=51 | |||||||
| M0 | 135.2 (18.1) | 6.5 | 0.008 | 12.1 | 45.1 | 27.5 | 0.001 |
| M1 | 134.4 (17.4) | 5.8 | 0.017 | 12.3 | 39.2 | 21.6 | 0.007 |
| M2 | 134.5 (17.6) | 5.8 | 0.026 | 12.9 | 31.4 | 13.7 | 0.065 |
| Average M1 & M2 | 134.4 (16.9) | 5.8 | 0.017 | 12.5 | 31.4 | 13.7 | 0.065 |
| Average M0-M2 | 134.6 (17.1) | 6.0 | 0.012 | 12.3 | 33.3 | 15.7 | 0.039 |
| ABPM awake | 128.6 (13.6) | Ref | Ref | Ref | 17.6 | Ref | Ref |
| Diastolic blood pressure | Mean (SD) | Mean Absolute Difference | P | RMSE | % DBP≥90 | Difference | P |
| 60 second intervals, N=51 | |||||||
| M0 | 79.3 (14.0) | 1.5 | 0.30 | 8.9 | 15.7 | 0.0 | 1.00 |
| M1 | 77.6 (13.5) | −0.2 | 0.90 | 8.6 | 11.8 | −3.9 | 0.73 |
| M2 | 77.0 (13.5) | −0.7 | 0.63 | 9.1 | 13.7 | −2.0 | 1.00 |
| Average M1 & M2 | 77.3 (13.2) | −0.4 | 0.75 | 8.7 | 13.7 | −2.0 | 1.00 |
| Average M0-M2 | 78.0 (13.3) | 0.2 | 0.88 | 8.6 | 13.7 | −2.0 | 1.00 |
| ABPM awake | 77.7 (12.2) | Ref | Ref | Ref | 15.7 | Ref | Ref |
| 30 second intervals, N=51 | |||||||
| M0 | 79.7 (13.5) | 3.1 | 0.044 | 9.6 | 21.6 | 5.9 | 0.58 |
| M1 | 77.7 (14.1) | 1.1 | 0.43 | 8.8 | 17.6 | 2.0 | 1.00 |
| M2 | 76.9 (14.4) | 0.4 | 0.78 | 9.0 | 13.7 | −2.0 | 1.00 |
| Average M1 & M2 | 77.3 (14.0) | 0.8 | 0.58 | 8.8 | 17.6 | 2.0 | 1.00 |
| Average M0-M2 | 78.3 (13.7) | 1.7 | 0.23 | 9.0 | 25.5 | 9.8 | 0.27 |
| ABPM awake | 76.5 (12.5) | Ref | Ref | Ref | 15.7 | Ref | Ref |
| Stage 2 or uncontrolled hypertension | % SBP≥140 or DBP≥90 | Difference | P | ||||
| 60 second intervals, N=51 | |||||||
| M0 | - | - | - | - | 45.1 | 11.8 | 0.15 |
| M1 | - | - | - | - | 47.1 | 13.7 | 0.12 |
| M2 | - | - | - | - | 41.2 | 7.8 | 0.42 |
| Average M1 & M2 | - | - | - | - | 43.1 | 9.8 | 0.27 |
| Average M0-M2 | - | - | - | - | 43.1 | 9.8 | 0.23 |
| ABPM awake | - | - | - | - | 33.3 | Ref | Ref |
| 30 second intervals, N=51 | |||||||
| M0 | - | - | - | - | 49.0 | 23.5 | 0.012 |
| M1 | - | - | - | - | 43.1 | 17.6 | 0.064 |
| M2 | - | - | - | - | 31.4 | 5.9 | 0.58 |
| Average M1 & M2 | - | - | - | - | 35.3 | 9.8 | 0.30 |
| Average M0-M2 | - | - | - | - | 45.1 | 19.6 | 0.041 |
| ABPM awake | - | - | - | - | 25.5 | Ref | Ref |
Abbreviations: ABPM, ambulatory blood pressure; M0, the first automated office blood pressure measurement; M1, the second automated office blood pressure measurement; M2, the third automated office blood pressure measurement; RMSE, root mean square error
The prevalence of hypertension was generally greater with the 30-second protocol compared to the 60-second protocol. However, hypertension defined based on the average of measurements 2 and 3 resulted in a 9.8% higher prevalence of hypertension compared to ABPM awake measurements, regardless of measurement interval.
Association of Protocol with Measurement
We examined the association between protocol and BP measurements (Table 3). Compared to the 60-second protocol, the 30-second protocol was not associated with any of the AOBP measurements or the awake-time ABPM measurement. However, the 30-second protocol was associated with non-significant differences between the first systolic AOBP measurement and the ABPM awake SBP measurements (−3 mm Hg; P = 0.26), measurements not influenced by the interval protocol. A similar magnitude difference (~3 mm Hg lower during the 30-second protocol) was observed for both M1, M2, and the average of measurements.
Table 3.
Association of the 30-second protocol (versus the 60-second protocol) with blood pressure measurements, N=102
| 30-second versus 60-second protocol | ||||
|---|---|---|---|---|
| Model 1 |
Model 2 |
|||
| β (95% CI) | P | β (95% CI) | P | |
| Systolic blood pressure (mm Hg) | ||||
| M0 | −2.20 (−9.83, 5.43) | 0.57 | −3.23 (−8.90, 2.44)* | 0.26* |
| M1 | −2.35 (−9.56, 4.86) | 0.52 | −3.20 (−9.07, 2.68) | 0.28 |
| M2 | −1.86 (−9.30, 5.59) | 0.62 | −2.75 (−8.73, 3.23) | 0.36 |
| Average M1 & M2 | −2.10 (−9.24, 5.04) | 0.56 | −2.97 (−8.67, 2.72) | 0.30 |
| Average M0-M2 | −2.11 (−9.30, 5.08) | 0.56 | −3.03 (−8.59, 2.53) | 0.28 |
| ABPM awake | −3.56 (−9.44, 2.31) | 0.23 | −3.23 (−8.90, 2.44)* | 0.26* |
| Diastolic blood pressure (mm Hg) | ||||
| M0 | −0.49 (−5.73, 4.74) | 0.85 | −1.81 (−6.11, 2.49)* | 0.41* |
| M1 | −0.93 (−6.05, 4.19) | 0.72 | −1.96 (−6.53, 2.61) | 0.40 |
| M2 | −1.14 (−6.31, 4.02) | 0.66 | −2.25 (−6.78, 2.27) | 0.33 |
| Average M1 & M2 | −1.04 (−6.08, 4.01) | 0.68 | −2.11 (−6.54, 2.33) | 0.35 |
| Average M0-M2 | −0.65 (−5.71, 4.42) | 0.80 | −1.79 (−6.15, 2.56) | 0.42 |
| ABPM awake | −2.33 (−6.76, 2.09) | 0.30 | −1.81 (−6.11, 2.49)* | 0.41* |
Linear regression models. Model 1 is adjusted for age, sex, and race. Model 2 is adjusted for covariates in Model 1 and the difference between the first automated measurement and awake-time ABPM.
These numbers are identical in Model 2 after adjustment for the difference between them.
When compared to the first AOBP measurement, the mean BP measures obtained with the second and third measurements from the 30-second protocol were nearly identical to the 60-second protocol, both showing lower values (Figure 2).
Figure 2.
Mean difference in (A) systolic blood pressure and (B) diastolic blood pressure comparing the second automated office blood pressure measurement (M1) and the third automated office blood pressure measurement (M2) to the first automated office blood pressure measurement (M0) according to interval protocol: 60-seconds (diamond, dash line) or 30-second (circle, solid line). Vertical lines represent the 95% confidence intervals.
We also examined whether the 30-second protocol compared with the 60-second protocol modified relationships between each measurement and awake-time ABPM (Table 4). The 30-second protocol did not significantly modify any of the measurements’ associations with awake-time ABPM. Moreover, the magnitudes of the potential modifications by protocol were 0.2 mm Hg or less.
Table 4.
The effect of interval timing on blood pressure measurement association with awake-time ambulatory blood pressure monitoring, N=102
| Model 1 |
Model 2 |
|||
|---|---|---|---|---|
| β (95% CI) | P | β (95% CI) | P | |
| Systolic blood pressure (mm Hg) | ||||
| M0 | −0.06 (−0.34, 0.21) | 0.64 | - | - |
| M1 | −0.08 (−0.37, 0.22) | 0.60 | −0.02 (−0.19, 0.15) | 0.83 |
| M2 | −0.20 (−0.49, 0.10) | 0.18 | −0.08 (−0.26, 0.09) | 0.35 |
| Average M1 & M2 | −0.14 (−0.44, 0.16) | 0.37 | −0.02 (−0.18, 0.13) | 0.75 |
| Average M0-M2 | −0.10 (−0.40, 0.19) | 0.48 | −0.01 (−0.11, 0.10) | 0.88 |
| Diastolic blood pressure (mm Hg) | ||||
| M0 | −0.00 (−0.27, 0.26) | 0.97 | - | - |
| M1 | 0.02 (−0.24, 0.27) | 0.90 | −0.06 (−0.18, 0.07) | 0.36 |
| M2 | 0.04 (−0.23, 0.30) | 0.79 | −0.06 (−0.19, 0.08) | 0.40 |
| Average M1 & M2 | 0.02 (−0.24, 0.28) | 0.86 | −0.07 (−0.18, 0.04) | 0.22 |
| Average M0-M2 | 0.01 (−0.26, 0.27) | 0.96 | −0.05 (−0.13, 0.02) | 0.16 |
These linear regression models portray the relationship between each blood pressure measurement and awake-time ABPM with a measurement-by-protocol interaction term.
Model 1 is adjusted for age, sex, and race.
Model 2 is adjusted for Model 1 covariates and for the difference between the first automated office measurement and awake-time ambulatory blood pressure.
Sensitivity analyses additionally adjusting models for sleep apnea or statin use (covariates with the largest magnitude imbalances at baseline), did not meaningfully alter our findings (results not shown).
Discussion
In this study of AOBP measurement interval, a 30-second interval compared to a 60-second interval was not associated with differences in BP measures and did not modify the relationship between individual AOBP measurements and awake-time BP from ABPM. In fact, changes in mean BP from the first to the second and third measurements during the 30-second protocol were nearly identical to the 60-second protocol. Overall, we found no advantage to a greater wait time between AOBP measurements, which supports the implementation of more time-efficient AOBP in clinical practice.
Timely and accurate BP measurement is a clinical challenge. In fact, The National Heart, Lung, and Blood Institute’s recently convened working group on BP measurement, included research on aspects of the BP measurement protocol to simplify implementation in routine clinical practice among their “high impact” research priorities.5 Among the comprehensive list of items recommended for accurate BP measurement is that more than 1 measurement be performed during a single office visit.5 Multiple guidelines have cited that the time between measurements be 1–2 minutes,3,4 based on evidence showing little difference in accuracy between 1-minute and 2-minute intervals.10,11 However, there is limited evidence comparing shorter time periods. One of the few studies on this topic showed high reliability in measurements at 30 seconds and 60 seconds compared to 90 seconds and 120 seconds.8 Our study supports this work suggesting that the 30-second measurement is equally reliable as 60-second intervals.
Our study raises interesting considerations for the measurement of BP using AOBP. We noted that later BP measurements were more likely to differ mildly from the first AOBP measurement with both the 30-second and 60-second interval protocols. This is relevant when AOBP is based on the average of the last 2 BP measurements, a practice followed by many studies and found in prior guidelines.14 Later BP measures have been shown to more closely reflect awake-time ABPM.15 This is thought to be due to physiologic adaptation effects from the first BP measurement,16 which have been documented in several AOBP studies, where the first measurement differed from later measurements.17,18 Unfortunately, many device manufacturers do not allow for easy averaging of the latter two measurements. Nevertheless, our study confirms that this decrement in BP with subsequent measurement is not altered by shortening the intervening time interval. It should be noted that others have found less reliable BP measures with intervals beyond 60 seconds.8 It is possible that extending the BP interval might re-trigger physiologic adaptation phenomena in some patients or that more intervening time may allow for environmental factors or events to occur that influence BP. However, intervals beyond 60 seconds were not evaluated in our study.
Our study raises the question of whether the time interval can be reduced even further beyond 30 seconds. Indeed the Omron HEM907XL has a 5-second feature for repeat measurements. This question is beyond the scope of our current study. Nevertheless, Imamura and colleagues in their study of time interval found significant increases in diastolic BP and significant reductions in systolic BP with a 15-second interval that were not observed with 30-second or 60-second intervals.8 The biologic mechanisms for this are unclear, but even shorter time intervals would be worth exploring in subsequent studies.
Our study has limitations. First, the protocols were not performed in the same patient and patients were not randomized to different protocols. As a result, the observed associations could be confounded by unmeasured patient characteristics. Despite this, the characteristics assessed were fairly well-balanced between protocol groups. Second, there was a larger difference between the first AOBP measurement and the awake-time ABPM with the 30-second protocol, compared to the 60-second protocol, measures not influenced by the time interval. It is possible that differences in season and time of day by protocol may have contributed to these differences. As a result, differences in hypertension prevalence as well as the absolute difference in BP between measurements and daytime BP should be interpreted cautiously. Third, our study included patients referred to our hypertension center. As a result, they may not reflect a general ambulatory population. Finally, our study was small, limiting our ability to examine subgroups in our population. This also limits our ability to adjust for differences in the populations assigned the two protocols.
Our study also has strengths. First, we measured BP with one of the most common AOBP devices in use, a device employed in respected trials,19,20 and compared these BP measurements to ABPM, the gold-standard method for measuring BP. Second, our AOBP measurements were unattended and performed in private patient rooms, which minimizes environmental effects on measurement. Third, our population was diverse with each person’s own awake-time ABPM or AOBP serving as a reference. Finally, our study addresses a highly relevant issue of time constraints related to the implementation of AOBP in clinical practice.
Our study has important implications for clinical practice. There is growing recognition of the value of AOBP for accurate diagnosis and management of hypertension.21–23 However, the duration of time needed to implement AOBP continues to be a barrier for wider adoption in clinic settings. Our study shows that a 30-second time interval may be as accurate and reliable as a 60-second time interval. These time savings should serve to encourage greater uptake of AOBP in clinical practice. Further studies are needed to examine other features of high quality BP measurement that can further optimize the time efficiency of AOBP in clinical practice.
Perspectives
In conclusion, a shorter time interval between AOBP measurements reduced the time required to measure BP with no observed compromise in accuracy or reliability. This finding should reduce barriers toward implementing AOBP in clinical practice.
Novelty and Significance.
What is New?
Guidelines recommend a 60 to 120 second paused between automated office blood pressure measurements. Our study demonstrates similar accuracy and reliability between a 30 second and 60 second interval.
What is Relevant?
Time is a major barrier toward implementation of automated office blood pressure in clinical practice. Our study shows that 60 seconds may be eliminated from the measurement process by using a 30 second interval between measurements without compromising measurement performance.
Summary
Our findings support shorter time intervals between BP measurements, which would make automated office blood pressure more feasible in clinical practice.
Acknowledgments
The authors express gratitude to Healthcare Associates administration for providing space and resources for this quality improvement study as well as the medical residents and office staff who participated in the study’s implementation.
Sources of Funding
SPJ is supported by NIH/NHLBI grant 7K23HL135273–02.
Footnotes
Disclosures
The authors declare that there is no conflict of interest associated with this manuscript.
The authors have no conflicts to disclose.
References
- 1.US Preventive Services Task Force, Krist AH, Davidson KW, Mangione CM, Cabana M, Caughey AB, Davis EM, Donahue KE, Doubeni CA, Kubik M, Li L, Ogedegbe G, Pbert L, Silverstein M, Stevermer J, Tseng C-W, Wong JB. Screening for Hypertension in Adults: US Preventive Services Task Force Reaffirmation Recommendation Statement. JAMA 2021; 325:1650–1656. < 10.1001/jama.2021.4987>. [DOI] [PubMed] [Google Scholar]
- 2.Nerenberg KA, Zarnke KB, Leung AA, Dasgupta K, Butalia S, McBrien K, Harris KC, Nakhla M, Cloutier L, Gelfer M, Lamarre-Cliche M, Milot A, Bolli P, Tremblay G, McLean D, Padwal RS, Tran KC, Grover S, Rabkin SW, Moe GW, Howlett JG, Lindsay P, Hill MD, Sharma M, Field T, Wein TH, Shoamanesh A, Dresser GK, Hamet P, Herman RJ, Burgess E, Gryn SE, Grégoire JC, Lewanczuk R, Poirier L, Campbell TS, Feldman RD, Lavoie KL, Tsuyuki RT, Honos G, Prebtani APH, Kline G, Schiffrin EL, Don-Wauchope A, Tobe SW, Gilbert RE, Leiter LA, Jones C, Woo V, Hegele RA, Selby P, Pipe A, McFarlane PA, Oh P, Gupta M, Bacon SL, Kaczorowski J, Trudeau L, Campbell NRC, Hiremath S, Roerecke M, Arcand J, Ruzicka M, Prasad GVR, Vallée M, Edwards C, Sivapalan P, Penner SB, Fournier A, Benoit G, Feber J, Dionne J, Magee LA, Logan AG, Côté A-M, Rey E, Firoz T, Kuyper LM, Gabor JY, Townsend RR, Rabi DM, Daskalopoulou SS, Hypertension Canada. Hypertension Canada’s 2018 Guidelines for Diagnosis, Risk Assessment, Prevention, and Treatment of Hypertension in Adults and Children. Can J Cardiol 2018; 34:506–525. < 10.1016/j.cjca.2018.02.022>. [DOI] [PubMed] [Google Scholar]
- 3.Williams B, Mancia G, Spiering W, Agabiti Rosei E, Azizi M, Burnier M, Clement DL, Coca A, de Simone G, Dominiczak A, Kahan T, Mahfoud F, Redon J, Ruilope L, Zanchetti A, Kerins M, Kjeldsen SE, Kreutz R, Laurent S, Lip GYH, McManus R, Narkiewicz K, Ruschitzka F, Schmieder RE, Shlyakhto E, Tsioufis C, Aboyans V, Desormais I, ESC Scientific Document Group. 2018 ESC/ESH Guidelines for the management of arterial hypertension. Eur Heart J 2018; 39:3021–3104. < 10.1093/eurheartj/ehy339>. [DOI] [PubMed] [Google Scholar]
- 4.Whelton PK, Carey RM, Aronow WS, Casey DE, Collins KJ, Himmelfarb CD, DePalma SM, Gidding S, Jamerson KA, Jones DW, MacLaughlin EJ, Muntner P, Ovbiagele B, Smith SC, Spencer CC, Stafford RS, Taler SJ, Thomas RJ, Williams KA, Williamson JD, Wright JT. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults: Executive Summary: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Hypertension January 2017:HYP.0000000000000066. < 10.1161/HYP.0000000000000066>. [DOI]
- 5.Muntner P, Einhorn PT, Cushman WC, Whelton PK, Bello NA, Drawz PE, Green BB, Jones DW, Juraschek SP, Margolis KL, Miller ER, Navar AM, Ostchega Y, Rakotz MK, Rosner B, Schwartz JE, Shimbo D, Stergiou GS, Townsend RR, Williamson JD, Wright JT, Appel LJ, 2017 National Heart, Lung, and Blood Institute Working Group. Blood Pressure Assessment in Adults in Clinical Practice and Clinic-Based Research: JACC Scientific Expert Panel. J Am Coll Cardiol 2019; 73:317–335. < 10.1016/j.jacc.2018.10.069>. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Pickering TG, Hall JE, Appel LJ, Falkner BE, Graves J, Hill MN, Jones DW, Kurtz T, Sheps SG, Roccella EJ. Recommendations for blood pressure measurement in humans and experimental animals: part 1: blood pressure measurement in humans: a statement for professionals from the Subcommittee of Professional and Public Education of the American Heart Association Council on High Blood Pressure Research. Circulation 2005; 111:697–716. < 10.1161/01.CIR.0000154900.76284.F6>. [DOI] [PubMed] [Google Scholar]
- 7.Giles TD, Egan P. Pay (adequately) for what works: the economic undervaluation of office and ambulatory blood pressure recordings. J Clin Hypertens (Greenwich) 2008; 10:257–259. < 10.1111/j.1751-7176.2008.08084.x>. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Imamura M, Asayama K, Sawanoi Y, Shiga T, Saito K, Ohkubo T. Effects of measurement intervals on the values of repeated auscultatory blood pressure measurements. Clin Exp Hypertens 2020; 42:105–109. < 10.1080/10641963.2019.1583243>. [DOI] [PubMed] [Google Scholar]
- 9.Eguchi K, Kuruvilla S, Ogedegbe G, Gerin W, Schwartz JE, Pickering TG. What is the optimal interval between successive home blood pressure readings using an automated oscillometric device? J Hypertens 2009; 27:1172–1177. < 10.1097/hjh.0b013e32832a6e39>. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Myers MG, Valdivieso M, Kiss A, Tobe SW. Comparison of two automated sphygmomanometers for use in the office setting. Blood Press Monit 2009; 14:45–47. < 10.1097/MBP.0b013e32831e314f>. [DOI] [PubMed] [Google Scholar]
- 11.Myers MG, Valdivieso M, Kiss A. Optimum frequency of office blood pressure measurement using an automated sphygmomanometer. Blood Press Monit 2008; 13:333–338. < 10.1097/MBP.0b013e3283104247>. [DOI] [PubMed] [Google Scholar]
- 12.Centers for Disease Control. National Health and Nutrition Examination Survey (NHANES): Physician Examination Procedures Manual September 2011.
- 13.Yang W-Y, Melgarejo JD, Thijs L, Zhang Z-Y, Boggia J, Wei F-F, Hansen TW, Asayama K, Ohkubo T, Jeppesen J, Dolan E, Stolarz-Skrzypek K, Malyutina S, Casiglia E, Lind L, Filipovský J, Maestre GE, Li Y, Wang J-G, Imai Y, Kawecka-Jaszcz K, Sandoya E, Narkiewicz K, O’Brien E, Verhamme P, Staessen JA, International Database on Ambulatory Blood Pressure in Relation to Cardiovascular Outcomes (IDACO) Investigators. Association of Office and Ambulatory Blood Pressure With Mortality and Cardiovascular Outcomes. JAMA 2019; 322:409–420. < 10.1001/jama.2019.9811>. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Hackam DG, Quinn RR, Ravani P, Rabi DM, Dasgupta K, Daskalopoulou SS, Khan NA, Herman RJ, Bacon SL, Cloutier L, Dawes M, Rabkin SW, Gilbert RE, Ruzicka M, McKay DW, Campbell TS, Grover S, Honos G, Schiffrin EL, Bolli P, Wilson TW, Feldman RD, Lindsay P, Hill MD, Gelfer M, Burns KD, Vallée M, Prasad GVR, Lebel M, McLean D, Arnold JMO, Moe GW, Howlett JG, Boulanger J-M, Larochelle P, Leiter LA, Jones C, Ogilvie RI, Woo V, Kaczorowski J, Trudeau L, Petrella RJ, Milot A, Stone JA, Drouin D, Lavoie KL, Lamarre-Cliche M, Godwin M, Tremblay G, Hamet P, Fodor G, Carruthers SG, Pylypchuk GB, Burgess E, Lewanczuk R, Dresser GK, Penner SB, Hegele RA, McFarlane PA, Sharma M, Reid DJ, Tobe SW, Poirier L, Padwal RS. The 2013 Canadian Hypertension Education Program Recommendations for Blood Pressure Measurement, Diagnosis, Assessment of Risk, Prevention, and Treatment of Hypertension. Canadian Journal of Cardiology 2013; 29:528–542. < 10.1016/j.cjca.2013.01.005>. [DOI] [PubMed] [Google Scholar]
- 15.Juraschek SP, Ishak A, Mukamal KJ, Cohen ML, Beach JL. Impact of Clinic-Based Blood Pressure Approaches on Blood Pressure Measurement. Am J Hypertens July 2019. < 10.1093/ajh/hpz118>. [DOI] [PMC free article] [PubMed]
- 16.Stergiou GS, Skeva II, Zourbaki AS, Mountokalakis TD. Self-monitoring of blood pressure at home: how many measurements are needed? J Hypertens 1998; 16:725–731. [DOI] [PubMed] [Google Scholar]
- 17.Verberk WJ, Kroon AA, Kessels AGH, Lenders JWM, Thien T, van Montfrans GA, Smit AJ, de Leeuw PW. The optimal scheme of self blood pressure measurement as determined from ambulatory blood pressure recordings. J Hypertens 2006; 24:1541–1548. < 10.1097/01.hjh.0000239289.87141.b6>. [DOI] [PubMed] [Google Scholar]
- 18.Schulze MB, Kroke A, Bergmann MM, Boeing H. Differences of blood pressure estimates between consecutive measurements on one occasion: implications for inter-study comparability of epidemiologic studies. Eur J Epidemiol 2000; 16:891–898. [DOI] [PubMed] [Google Scholar]
- 19.ACCORD Study Group, Cushman WC, Evans GW, Byington RP, Goff DC, Grimm RH, Cutler JA, Simons-Morton DG, Basile JN, Corson MA, Probstfield JL, Katz L, Peterson KA, Friedewald WT, Buse JB, Bigger JT, Gerstein HC, Ismail-Beigi F. Effects of intensive blood-pressure control in type 2 diabetes mellitus. N Engl J Med 2010; 362:1575–1585. < 10.1056/NEJMoa1001286>. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.SPRINT Research Group, Wright JT, Williamson JD, Whelton PK, Snyder JK, Sink KM, Rocco MV, Reboussin DM, Rahman M, Oparil S, Lewis CE, Kimmel PL, Johnson KC, Goff DC, Fine LJ, Cutler JA, Cushman WC, Cheung AK, Ambrosius WT. A Randomized Trial of Intensive versus Standard Blood-Pressure Control. N Engl J Med 2015; 373:2103–2116. < 10.1056/NEJMoa1511939>. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Andreadis EA, Geladari CV, Angelopoulos ET. The optimal use of automated office blood pressure measurement in clinical practice. The Journal of Clinical Hypertension 2020; 22:555–559. < 10.1111/jch.13837>. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Myers MG, Campbell NRC. Unfounded concerns about the use of automated office blood pressure measurement in SPRINT. J Am Soc Hypertens 2016; 10:903–905. < 10.1016/j.jash.2016.10.003>. [DOI] [PubMed] [Google Scholar]
- 23.Pappaccogli M, Di Monaco S, Perlo E, Burrello J, D’Ascenzo F, Veglio F, Monticone S, Rabbia F. Comparison of Automated Office Blood Pressure With Office and Out-Off-Office Measurement Techniques. Hypertension 2019; 73:481–490. < 10.1161/HYPERTENSIONAHA.118.12079>. [DOI] [PubMed] [Google Scholar]