98

Devices and technology

Blood pressure variability of two ambulatory blood pressure monitors Radhakrishna R. Kallema, Kevin E.C. Meyersa,b, Andrew J. Cucchiaraa, Deirdre L. Sawinskia and Raymond R. Townsenda Objective There are no data on the evaluation of blood pressure (BP) variability comparing two ambulatory blood pressure monitoring monitors worn at the same time. Hence, this study was carried out to compare variability of BP in healthy untreated adults using two ambulatory BP monitors worn at the same time over an 8-h period. Methods An Accutorr device was used to measure office BP in the dominant and nondominant arms of 24 participants. Simultaneous 8-h BP and heart rate data were measured in 24 untreated adult volunteers by Mobil-O-Graph (worn for an additional 16 h after removing the Spacelabs monitor) and Spacelabs with both random (N = 12) and nonrandom (N = 12) assignment of each device to the dominant arm. Average real variability (ARV), SD, coefficient of variation, and variation independent of mean were calculated for systolic blood pressure, diastolic blood pressure, mean arterial pressure, and pulse pressure (PP). Results Whether the Mobil-O-Graph was applied to the dominant or the nondominant arm, the ARV of mean systolic (P = 0.003 nonrandomized; P = 0.010 randomized) and PP (P = 0.009 nonrandomized; P = 0.005 randomized) remained significantly higher

Introduction Mean blood pressure (BP) is a risk factor for cardiovascular (CV) events. However, additional epidemiological evidence suggests that instability and variability in BP are also important as they are associated with the development of target organ damage [1–3]. Ambulatory blood pressure monitoring (ABPM) provides information about mean BP, daytime changes in BP, and within-period BP changes. ABPM variability includes both short term and circadian components that can be estimated by the SD of the BP values over a defined period of the day or by the night-to-day BP ratio, respectively, and by the average real variability (ARV) of the BP. However, BP variability is not confined to 24-h studies but can also be determined from a single office visit with multiple readings or from office visit to office visit, with BP variability determined both over the short term (days, months) and over the long term (months, years). All supplementary digital content is available directly from the corresponding author. c 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins 1359-5237

than the Spacelabs device, whereas the ARV of the mean arterial pressure was not significantly different. The average BP readings and ARVs for systolic blood pressure and PP obtained by the Mobil-O-Graph were considerably higher for the daytime than the night-time. Conclusion Given the emerging interest in the effect of BP variability on health outcomes, the accuracy of its measurement is important. Our study raises concerns about the accuracy of pooling international ambulatory blood pressure monitoring variability data using different c 2014 Wolters devices. Blood Press Monit 19:98–102 Kluwer Health | Lippincott Williams & Wilkins. Blood Pressure Monitoring 2014, 19:98–102 Keywords: ambulatory blood pressure monitoring, average real variability, comparison, diastolic blood pressure, Mobil-O-Graph, Spacelabs 90207, systolic blood pressure, variability a Department of Medicine, Renal, Electrolyte and Hypertension Division, University of Pennsylvania School of Medicine and bDivision of Nephrology, Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania, USA

Correspondence to Radhakrishna R. Kallem, MD, MPH, Department of Medicine, Renal, Electrolyte and Hypertension Division, University of Pennsylvania School of Medicine, 122 Founders, 3400 Spruce Street, Philadelphia, PA 19104, USA Tel: + 1 215 614 0423; fax: + 1 215 662 3459; e-mail: [email protected] Received 1 May 2013 Revised 7 October 2013 Accepted 19 November 2013

There are little or no data on the evaluation of BP variability comparing two ABPM monitors worn simultaneously. As variability is a parameter used to assess CV and other outcomes related to the BP, we aimed to determine whether two ABPM monitors worn at the same time would provide similar information about variability. Hence, this study was carried out to compare variability of BP in healthy adults using two ambulatory BP monitors worn at the same time over an 8-h period. In addition, one of the monitors [Mobil-O-Graph (MOG)] was worn for an additional 16 h so that the day–night variability could be assessed in these participants.

Methods Nonrandomized study

Twelve adult healthy volunteers were recruited at the Hospital of the University of Pennsylvania over a period of 2 months. Written informed consent for the study, approved by the Institutional Review Board of the University of Pennsylvania, was obtained from all participants. DOI: 10.1097/MBP.0000000000000019

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

Blood pressure variability Kallem et al. 99

BP and heart rate (HR) data, reported elsewhere [4], were collected over an 8-h period beginning at 8:30 a.m.± 60 min. Demographic data, including date of birth, sex, and ethnicity were collected before the start of the monitoring period (Table 1). The height and weight of each participant were recorded using the Seca 242 digital stadiometer (Seca, Hanover, Maryland, USA) and Scaletronix 5002 stand-on scale (Scaletronix, White Plains, New York, USA). Three standard seated office BP recordings were obtained using an Accutorr (Mindray, Mahwah, New Jersey, USA) device before and after the 8-h monitoring interval.

24 h as part of a separate study protocol that prespecified its application to the nondominant arm. Thus, we did not randomize the designated arm for the monitor in this study.

The Spacelabs 90207 (Spacelabs Medical, Issequah, Washington, USA) and MOG (I.E.M., Stolberg, Germany) monitors were then applied in the sitting position to record 8 h ambulatory BP and HR data from the brachial artery of each arm once every 15 min. As shown in Fig. 1, the Spacelabs monitor was worn for 8 h on the dominant arm, whereas the MOG was worn on the nondominant arm for

Mid-arm circumference was used to determine the cuff size for the measurement of all BP readings using Accutorr, MOG, and Spacelabs monitors. The Accutorr and Spacelabs monitors are calibrated against the mercury sphygmomanometer annually in our institution. We relied on the manufacturer’s calibration for the MOG monitor.

Randomized study

Another 12 healthy adults who provided written informed consent for a separate IRB-approved study were recruited for the application of the same monitors (MOG and Spacelabs) for the same duration (8-h interval). However, this time, the arm to which each monitor was applied was determined randomly by a coin flip at the time of application.

Data Table 1

Baseline characteristics N = 24

Demographics Sex [n (%)] Male Female Race [n (%)] White Asian African American Age (years) [mean (range)] Height (cm) [mean (range)] Weight (kg) [mean (range)] BMI (kg/m2) [mean (range)]

Nonrandomized (N = 12)

Randomized (N = 12)

7 (58.3) 5 (41.7) 9 (75) 3 (25) 0 (0) 38 (24–57) 175.75 (156–198) 84.4 (63–109) 27.39 (20–39)

6 (50) 6 (50) 5 (41.7) 5 (41.7) 2 (16.6) 38 (21–64) 166.1 (153–180) 70.9 (49–107) 25.5 (20–35)

The summary BP and HR readings for both the nonrandomized and the randomized studies have been reported elsewhere [4]. The ABPM manufacturers’ presets were used to obtain the data with no further editing. The ARV and SD of 8 h summary BP within each participant were calculated for both the nonrandomized and the randomized studies, and reported as mean±SEM. Paired t-testing (two tailed) was performed to determine the statistical significance of differences within the participants. Also, the ARV and SD of day-time and night-time BP and HR readings recorded by the MOG device are calculated for each participant, compared using a paired t-test to evaluate difference between day and night (P < 0.05). Finally, we calculated the coefficient of variation (COV = SD/mean

Fig. 1

Nonrandomized cohort N = 12 Spacelabs 8h 24 h Mobil-O-Graph

Accutorr interarm BP difference (3 readings)

Randomized cohort N = 12 Accutorr interarm BP difference (3 readings)

Spacelabs 8h 8h Mobil-O-Graph

Graphical representation of the study design. BP, blood pressure.

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

100 Blood Pressure Monitoring 2014, Vol 19 No 2

SBP) and subsequently computed the variance independent of mean according to the method described by Rothwell and colleagues [1,5].

Results Randomization resulted in the Spacelabs and MOG being applied to the dominant arms on 5 and 7 of the 12 participants, respectively. MOG recorded an average of 24 (75%) successful and valid readings during an 8-h period, with about eight (25%) readings being discarded because of technical failures, whereas the Spacelabs monitor obtained an average of 30 (89%) successful and valid readings, with about four (11%) readings being discarded because of technical failures for each participant. In the nonrandomized cohort, the MOG device recorded higher mean systolic ARV [11.3±1.0 mmHg (P = 0.003)] and mean pulse pressure (PP) ARV [12.3±1.5 mmHg (P = 0.009)] for the 8-h interval compared with the mean

systolic ARV of 7.5±0.6 mmHg and mean PP ARV of 7.8±0.5 mmHg measured by the Spacelabs monitor as shown in Fig. 2a. In contrast, close agreement was observed between the devices for measured ARV of mean arterial (P = 0.277) and diastolic (P = 0.116) pressures, with MOG measuring 7.6±0.8 and 8.0±0.9 mmHg and Spacelabs measuring 6.6±0.5 and 6.7±0.5 mmHg, respectively. In the randomized study, MOG measured a higher mean systolic ARV [9.2±0.6 mmHg (P = 0.010) as shown in Table 2], mean PP ARV [11.0±1.0 mmHg (P = 0.005)] and mean diastolic ARV [8.3±0.7 mmHg (P = 0.031)], compared with the mean systolic ARV (7.2±0.4 mmHg), mean PP ARV (8.1±0.6 mmHg) and mean diastolic ARV (6.5±0.5 mmHg) using the Spacelabs monitor (Supplementary Tables 1–3). These results are also shown in Fig. 2b. Whether the MOG was applied to the dominant or the nondominant arm, the mean systolic and PP ARV readings remained significantly higher than those obtained

Fig. 2

Nonrandomized study: ARV

(a)

MAP

DBP

PP

P = 0.008 P = 0.003

SBP

0

2

4

6

8

10

12

14

16

Randomized study: ARV

(b) MAP

P = 0.031

DBP PP

P = 0.005

SBP

P = 0.01 0

2

4

6 8 ARV (mmHg) Spacelabs

10

12

14

Mobil-O-Graph

(a, b) The mean average real variability over 8 h in SBP, PP, MAP, and DBP as bars from each monitor in the nonrandomized (a) and randomized studies (b). The error bars represent the SEM. ARV, average real variability; DBP, diastolic blood pressure; MAP, mean arterial pressure; PP, pulse pressure; SBP, systolic blood pressure.

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

Blood pressure variability Kallem et al. 101

Table 2

SBP randomized cohort Number of readings

Participant 1 2 3 4 5 6 7 8 9 10 11 12 Mean

MOG

Slabs

MOG

Slabs

8 h SBP

8 h range

8 h ARV

8 h SBP

8 h range

8 h ARV

31 23 25 24 14 20 25 25 30 24 26 32 25

34 31 33 32 28 27 30 34 35 32 33 34 32

117 126 114 115 104 140 115 117 143 134 112 119 121

31 50 48 64 27 80 37 47 82 73 51 40 53

7 10 10 11 9 12 7 6 11 12 10 7 9

122 129 127 120 116 147 129 131 147 135 121 128 129

38 37 25 38 29 42 29 51 40 36 37 34 36

9 7 6 7 7 10 5 8 8 7 6 7 7

ARV, average real variability; MOG, Mobil-O-Graph; SBP, systolic blood pressure.

with the Spacelabs device, whereas the ARV of the mean arterial pressure (MAP) was not significantly different.

SBP data are not necessarily interchangeable between different monitors.

The average systolic blood pressure (SBP) (P < 0.000), diastolic blood pressure (DBP) (P < 0.000), MAP (P < 0.000), and PP (P < 0.044) readings obtained by the MOG were considerably higher for the day-time than the night-time readings. Similarly, the day-time ARVs (P = 0.015) and SDs (P = 0.005) for SBP and ARV (P < 0.000) and SD (P < 0.000) for PP were significantly higher than the night-time ARVs and SDs.

Despite the lower average SBP readings recorded by the MOG device, the variability was higher when compared with the Spacelabs device. This may seem counterintuitive because of the higher average SBP obtained by the Spacelabs device. Therefore, we computed a coefficient of variation (MOG: r = 0.181, P = 0.398; slabs: r = – 0.215, P = 0.312) and the variance independent of mean (MOG: r = – 0.031, P = 0.885; slabs: r = – 0.002, P = 0.991) for SBP for both devices to observe their association with the mean BP. These showed regression slopes, which are not significantly different from zero, indicating that the variation in the devices was independent of mean SBP and not linked to it, supporting the validity of our observations. We also computed the BP variability between the two ABPM monitors using SD and compared it with the ARV to observe whether they differ. Our results suggest that the SD of the SBP (P = 0.004) and the PP (P = 0.009) measured by MOG were higher than that of the Spacelabs monitor, whereas the SD of the DBP and the MAP, although numerically higher, was not statistically different.

Discussion In this study, we observed that variability in systolic pressure, as assessed by either ARV or SD, was consistently higher with the MOG compared with the Spacelabs device. In the MOG, the daytime variability by ARV in SBP was higher than the night-time values. These relationships also held true for PP, which is principally governed by the systolic pressure. Interestingly, we did not observe a difference in variability around the MAP with either device, nor did we observe a difference in variability around the MAP comparing daytime with night-time values with the MOG. This is notable because the devices actually measure the MAP, and estimate the SBP and DBP values using proprietary algorithms. We observed that the MOG recorded a higher SBP variability compared with the Spacelabs monitor while recording lower levels of SBP [4]. As noted above, this could be because of the internal algorithms used by the monitors to calculate the systolic and diastolic pressures from the MAPs measured. This is further supported by our finding that the ARV of MAPs, although numerically different, was not significantly different. That the observed ARV differences could have resulted from application of the monitor to the dominant versus nondominant arms is unlikely because this finding was noted again in the randomized cohort. Our studies therefore make the important observation that ARV of

Hansen and colleagues reported on 8938 participants from 11 populations followed for a mean of 11 years. Their findings suggested that BP variability assessed from 24-h ambulatory recordings contributed only about 1% to risk stratification over and beyond 24-h BP information. However, on close analysis, ARV was a better indicator of CV outcome than SD in their study [6]. Pierdomenico and colleagues, in a study of 1280 sequential hypertensive patients older than 40 years of age, found that high ARV of daytime systolic BP was an independent predictor of CV risk whereas high SD did not predict CV risk. Given the potential importance of ARV as a risk factor for CV outcomes, and as a better measure of BP variability [7], it was noteworthy that a higher ARV with a lower average SBP was obtained by the MOG compared with the Spacelabs device.

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

102 Blood Pressure Monitoring 2014, Vol 19 No 2

As we have noted previously [4], it is important to interpret a BP metric within the constraint of the measurement technique used, particularly when using oscillometric methods. There is very little literature comparing withinparticipant variability in BP using two different contemporary ambulatory measurement methods. We suggest that our study makes a modest, but valuable contribution toward this important aspect of clinical medicine. Determination of systolic pressure by oscillometric devices is performed by an internal algorithm. Technical factors such as bleed down rates, for example, are different between devices and could account for some of the differences in measurements. In the absence of specific details, as the algorithms are proprietary, this remains speculative. We acknowledge several limitations in our study. Our numbers are relatively small, and we cannot say which monitor is a better indicator of variability as we can no longer use a mercury standard, and we did not perform an invasive (intra-arterial) pressure measurement to support our findings. Our participants were young, healthy, and unmedicated; thus, our observations may not apply to an older population with comorbidities or to those on antihypertensive medications. It is important that further studies evaluating this are carried out on larger, more diverse cohorts.

confirmed in large cohorts, raise concerns about the accuracy of pooling international ABPM variability data when different devices are used.

Acknowledgements These studies were carried out in the Clinical and Translational Research Center, which is funded by Grant number UL1RR024134 from the National Center For Research Resources. Conflicts of interest

There are no conflicts of interest.

References 1

2 3

4

5

Conclusion

We noted statistically significant differences in the variability of systolic and PP measurements, but not in MAP measurement during a period of 8 h using two different ABPM devices. Given the increasing interest in the influence of BP variability on health outcomes, the accuracy of its measurement is important. Our findings, if

6

7

Rothwell PM, Howard SC, Dolan E, O’Brien E, Dobson JE, Dahlo¨f B, et al. Prognostic significance of visit-to-visit variability, maximum systolic blood pressure, and episodic hypertension. Lancet 2010; 375: 895–905. Mancia G. Prognostic value of long-term blood pressure variability: the evidence is growing. Hypertension 2011; 57:141–143. Muntner P, Shimbo D, Tonelli M, Reynolds K, Arnett DK, Oparil S. The relationship between visit-to-visit variability in systolic blood pressure and all-cause mortality in the general population: findings from NHANES III, 1988 to 1994. Hypertension 2011; 57:160–166. Kallem RR, Meyers KE, Sawinski DL, Townsend RR. A comparison of two ambulatory blood pressure monitors worn at the same time. J Clin Hypertens (Greenwich) 2013; 15:321–325. Rothwell PM, Howard SC, Dolan E, O’Brien E, Dobson JE, Dahloff B, et al. Effects of beta blockers and calcium-channel blockers on within-individual variability in blood pressure and risk of stroke. Lancet Neurol 2010; 9:469–480. Hansen TW, Thijs L, Li Y, Boggia J, Kikuya M, Bjo¨rklund-Bodega˚rd K, et al. Prognostic value of reading-to-reading blood pressure variability over 24 hours in 8938 subjects from 11 populations. Hypertension 2010; 55:1049–1057. Pierdomenico SD, Di Nicola M, Esposito AL, Di Mascio R, Ballone E, Lapenna L, et al. Prognostic value of different indices of blood pressure variability in hypertensive patients. Am J Hypertens 2009; 22:842–847.

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

Blood pressure variability of two ambulatory blood pressure monitors.

There are no data on the evaluation of blood pressure (BP) variability comparing two ambulatory blood pressure monitoring monitors worn at the same ti...
302KB Sizes 0 Downloads 0 Views