Journal Logo

EDITORIAL COMMENTARIES

Accuracy of noninvasive central blood pressure estimation: still a long ‘wave’ to go

Pucci, Giacomoa,b; Vaudo, Gaetanoa,b; Picone, Dean S.c

Author Information
doi: 10.1097/HJH.0000000000002610
  • Free

Despite a rollercoaster-like trend caused by a series of scientific confirmations and confutations related to its clinical usefulness, the interest upon the noninvasive estimation of central blood pressure (cBP) is still on the rise. There are many convincing reasons supporting the clinical application of cBP. Pressure waveform changes between the aorta and periphery have an important relationship with the structure and the function of large arteries and may therefore be considered a marker of vascular ageing, a parameter which is increasing in prominence in daily clinical practice [1]. Another reason is the possibility to use cBP to directly relate indexes of myocardial performance, such as stroke work, left ventricular contractile reserve and myocardial energetic efficiency, to aortic, not to brachial, pressure load [2,3].

Measurement of cBP from upper-arm cuff (brachial)-based oscillometric devices, instead of classical tonometry methods, is more attractive for three further reasons. First, cuff devices are familiar in clinical practice, promoting the easy-use of cBP. Second, these devices are suitable for 24-h ambulatory BP monitoring [4]. Third, these devices measure the peripheral waveform and BP for waveform calibration at the same site. This final reason should not be underestimated. Indeed, for all devices that reconstruct the central arterial waveform, calibration of the peripheral waveform is a fundamental step to obtain absolute values of cBP.

The most common process of peripheral waveform calibration to generate the central waveform is based on identification of its two extremes (SBP and DBP), which then allows mean arterial pressure (MAP) to be found by integrating the area under the waveform curve. MAP, obtained by waveform integration, and DBP therefore can be applied to waveforms to calculate cSBP values following the rationale that, as opposite to SBP (and pulse pressure – PP), MAP and DBP remain relatively constant along the entire arterial tree [5].

Despite the crucial role of MAP in the central waveform calibration, there is no reason to believe that waveform calibration to MAP/DBP is more accurate than calibration to SBP/DBP [6]. Indeed, MAP/DBP is useful for calibration of a peripheral waveform taken from an arterial site where SBP/DBP cannot be measured, such as the radial or carotid site. This is not the case for brachial BP because, when MAP is integrated from the brachial SBP/DBP calibrated waveform, calibration of the brachial waveform to SBP/DBP or MAP/DBP will generate practically identical cBP results [7].

Derivation of MAP via integration of a peripheral waveform (e.g. brachial), calibrated to SBP/DBP taken at the same anatomical site of waveform measurement, is not insensitive to measurement errors. Any inaccuracy of the brachial SBP/DBP, representing the two extremes of the waveform, will inevitably translate to an inaccurate integrated MAP and inaccurate waveform calibration. This is an important point because inaccuracy of brachial oscillometric SBP and DBP for measurement of invasive brachial BPs still is the major limiting factor to the accurate noninvasive estimation of cBP [8]. On the other hand, the calculation of MAP by fixed ratio equations (such as 33 or 40% PP) cannot be used as a surrogate of MAP integrated from peripheral waveforms, because it was recently shown not to accurately predict true invasive MAP [9].

To overcome some of the above limitations, the Mobil-o-Graph (I.E.M., Stolberg, Germany), a brachial cuff-based device, proposed a novel method to calibrate peripheral waveforms using the brachial MAP directly from the oscillometric pressure signal as the point corresponding to the maximum level of oscillation during cuff deflation when transmural pressure approximates the zero value. Indeed, by substituting ‘oscillometric’ brachial MAP to the SBP/DBP calibration (the latter is identical to using ‘integrated’ MAP/DBP for calibration of the brachial waveform), a higher level of accuracy in estimating central SBP and PP [10], a closer association with intermediate markers of cardiac organ damage [11], and a higher prognostic value [12] than classical SBP/DBP calibration of the brachial waveform were all demonstrated. This interesting approach has solid physiological background [13] but does not solve the problem of inaccuracy of oscillometric brachial DBP measurement which could negatively influence central DBP and PP estimation.

Therefore, the key question is whether oscillometric MAP is a more accurate measurement of invasive brachial MAP than oscillometric SBP is of invasive brachial SBP, and thus more accurate calibration method.

In the present volume of the Journal, a study from Gotzmann et al.[14] provides further insights on this topic. In a selected population of 159 individuals undergoing cardiac catheterization, they evaluated the accuracy of two brachial cuff-based devices, the SphygmoCor Xcel device (AtCor Medical, Sydney, Australia) and the Mobil-o-Graph, for the noninvasive assessment of cBP compared with invasive BP values obtained during coronary angiography. More than a simple comparison of the performance of two commercially available devices, the current study offers the opportunity to compare results when different methods of calibration are applied to the same brachial waveform, specifically when two methods for brachial MAP assessment are used. This was only performed with the Mobil-o-Graph device, because the SphygmoCor Xcel does not usually provide the oscillometric MAP.

The main results from the brachial SBP/DBP (identical to using ‘integrated’ MAP/DBP) calibration of the brachial waveform showed there was, as expected, a degree of underestimation of invasive cSBP (on average by 4 mmHg for both devices) and central PP (cPP) (8% for the SphygmoCor Xcel and 12% for Mobil-o-Graph). This cPP underestimation is mainly as a consequence of the cSBP underestimation. When oscillometric MAP replaced the ‘integrated’ MAP for calibration of the Mobil-o-Graph, both cSBP and cPP were overestimated by the same degree in absolute terms (4 mmHg). Significantly, when analyzing the association between estimated cSBP and invasive cSBP, the authors found that calibration to oscillometric MAP/DBP was associated with more dispersion around the mean than calibration to SBP/DBP. This finding highlights that similar absolute mean differences may hide wider dispersion of data, and raise the question whether large limits of agreement suggest imprecision or conceal clinically and prognostically meaningful information [15]. Results from the current study appear to be in favor of the first hypothesis.

The current study does not provide data related to invasive brachial MAP, leaving some doubts about the exact sources of measurement error. It should also be noticed that the processes of oscillometric MAP measurement and brachial waveform acquisition are not perfectly simultaneous but stepwise. This introduces a further potential source of error which may become clinically relevant especially in people with extreme and rapid BP fluctuations.

Overall, results of the current study support the notion that different calibration methods provide different results for each different device. This point is emphasized by other recent data that suggests the SphygmoCor Xcel, when calibrated with oscillometric MAP, significantly underestimated cSBP [16]. This means that the accuracy of cBP devices should be assessed rigorously with all available calibration modes to help identify the method that gives the most accurate cBP for each device. In doing so, the accuracy of oscillometric SBP, MAP and DBP measures used for calibration should also be tested for accuracy compared with peripheral (brachial) SBP, DBP and MAP invasive values. This would help provide important insights into the mechanisms of error in estimated cBP from oscillometric MAP calibration.

In conclusion, the present findings should stimulate further research and better understanding of brachial oscillometric signals, to find a way to obtain the ‘missing numbers’ needed to optimize the precious information contained in the peripheral wave profile. This perspective is intriguing, although there is still a long ‘wave’ to go….

ACKNOWLEDGEMENTS

Conflicts of interest

There are no conflicts of interest.

REFERENCES

1. Mayer CC, Climie R, Hametner B, Bruno RM. The European COST action VascAgeNet fostering innovation – when industry comes to science. Artery Res 2020; [Epub ahead of print] doi:10.2991/artres.k.200430.001.
2. Mancusi C, Losi MA, Izzo R, Canciello R, Manzi MV, Sforza A, et al. Effect of diabetes and metabolic syndrome on myocardial mechano-energetic efficiency in hypertensive patients. The Campania Salute Network. J Hum Hypertens 2017; 31:424.
3. Torres MAR, Texeira TF, Camarozano AC, Bellagamba CCA, Quevedo NM, Heidemann Junior AI, et al. The value of a simplified approach to end-systolic volume measurement for assessment of left ventricular contractile reserve during stress-echocardiography. Int J Cardiovasc Imaging 2019; 35:1019–1026.
4. Pucci G, Battista F, Crocetti A, Tilocca G, Boschetti E. How to measure 24-h central blood pressure and its potential clinical implications. High Blood Press Cardiovasc Prev 2017; 24:141–148.
5. Pauca AL, Wallenhaupt SL, Kon ND, Tucker WY. Does radial artery pressure accurately reflect aortic pressure? Chest 1992; 102:1193–1198.
6. Papaioannou TG, Karageorgopoulou TD, Sergentanis TN, Protogerou AD, Psaltopoulou T, Sharman JE, et al. Accuracy of commercial devices and methods for noninvasive estimation of aortic systolic blood pressure a systematic review and meta-analysis of invasive validation studies. J Hypertens 2016; 34:1237–1248.
7. Picone DS, Schultz MG, Peng X, Black JA, Dwyer N, Roberts-Thomson P, et al. Intra-arterial analysis of the best calibration methods to estimate aortic blood pressure. J Hypertens 2019; 37:307–315.
8. Shih YT, Cheng HM, Sung SH, Hu WC, Chen CH. Quantification of the calibration error in the transfer function-derived central aortic blood pressures. Am J Hypertens 2011; 24:1312–1317.
9. Schultz MG, Picone DS, Armstrong MK, Black JA, Dwyer N, Roberts-Thomson P, et al. The influence of SBP amplification on the accuracy of form-factor-derived mean arterial pressure. J Hypertens 2020; 38:1033–1039.
10. Weber T, Wassertheurer S, Rammer M, Maurer E, Hametner B, Mayer CC, et al. Validation of a brachial cuff-based method for estimating central systolic blood pressure. Hypertension 2011; 58:825–832.
11. Nakagomi A, Okada S, Shoji T, Kobayashi Y. Crucial effect of calibration methods on the association between central pulsatile indices and coronary atherosclerosis. Am J Hypertens 2017; 30:24–27.
12. Wassertheurer S, Baumann M. Assessment of systolic aortic pressure and its association to all cause mortality critically depends on waveform calibration. J Hypertens 2015; 33:1884–1888.
13. Geddes LA, Combs W, Denton W, Whistler SJ, Bourland JD. Indirect mean arterial pressure in the anesthetized dog. Am J Physiol 1980; 238:H664–H666.
14. Gotzmann M, Hogeweg M, Bauer F, Seibert FS, Rohn BJ, Mügge A, et al. The impact of calibration approaches on the accuracy of oscillometric central aortic blood pressure measurement. J Hypertens 2020; 38:2154–2160.
15. Wassertheurer S, Hametner B, Mayer CC, Hafez A, Negishi K, Papaioannou TG, et al. Aortic systolic pressure derived with different calibration methods: associations to brachial systolic pressure in the general population. Blood Press Monit 2018; 23:134–140.
16. Schultz MG, Picone DS, Armstrong MK, Black JA, Dwyer N, Roberts-Thomson P, et al. Validation study to determine the accuracy of central blood pressure measurement using the Sphygmocor Xcel cuff device. Hypertension 2020; 76:244–250.
Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.