Validation of the Noninvasive Assessment of Central Blood Pressure by the SphygmoCor and Omron Devices Against the Invasive Catheter Measurement

Transcription

1 original contributions nature publishing group Validation of the Noninvasive Assessment of Central Blood Pressure by the SphygmoCor and Omron Devices Against the Invasive Catheter Measurement Feng-Hua Ding 1,2, Wang-Xiang Fan 1, Rui-Yan Zhang 2, Qi Zhang 2, Yan Li 1 and Ji-Guang Wang 1 Background To investigate the accuracy of the SphygmoCor and Omron HEM9-AI devices in the estimation of central blood pressure (BP) in comparison with the simultaneous invasive catheter measurement. Methods The radial arterial pulse was sequentially recorded by the use of the Omron and SphygmoCor devices in 33 patients, with the calibration of the brachial oscillometric BP, to derive central BPs, which were also measured simultaneously with a catheter-based fluid-filled manometer system. The procedure was repeated three times to obtain 99 pairs of noninvasive and invasive measurements. Results The noninvasive central systolic BP estimations were significantly (P <.1) associated with the invasive measurement at the ascending aorta, with a correlation coefficient of.91 and.9 for the SphygmoCor and Omron devices, respectively. However, both devices underestimated central systolic BP with a difference of 15 mm Hg (95% confidence interval (CI), 17 to 13 mm Hg, P <.1) for SphygmoCor and 2 mm Hg (95% CI, 4 to mm Hg, P <.5) for Omron. In comparison with the invasive catheter measurement at the brachial artery, the oscillometric Omron device underestimated brachial systolic BP by 19 mm Hg (95% CI, 23 to 15 mm Hg, P <.1). Conclusions Both devices underestimated central systolic BP, with a larger deviation by SphygmoCor. Nonetheless, these noninvasive estimations of central BP closely correlate with the invasive measurements, and can still be properly used, on the condition that device specific diagnostic thresholds become available. Keywords: applanation tonometry; blood pressure; catheterization; central aortic blood pressure; hypertension; validation American Journal of Hypertension, advance online publication 6 October 211; doi:1.138/ajh Emerging evidence supports that central hemodynamics might be superior to brachial blood pressure (BP) in the prediction of cardiovascular events and target organ damage. 1 3 The issue is how to measure central hemodynamics noninvasively and accurately. The SphygmoCor device (AtCor Medical, Sydney, Australia) has been commercially available for more than a decade, and used extensively in epidemiological 2,4,5 and clinical studies. 6,7 It directly records radial pulse waveform using applanation tonometry, employs a generalized transfer function to obtain central aortic pressure waveform, and then with the calibration of brachial BP and the assumption of constant mean arterial pressure over large arteries derives central systolic BP and pulse pressure Because, in the use of the SphygmoCor device, radial pulse waveforms are calibrated with brachial BPs measured with a 1 Centre for Epidemiological Studies and Clinical Trials, The Shanghai Institute of Hypertension, Ruijin Hospital, Shanghai Jiaotong University School of Medicine, Shanghai, China; 2 Department of Cardiology, Ruijin Hospital, Shanghai Jiaotong University School of Medicine, Shanghai, China. Correspondence: Ji-Guang Wang (jiguangw@gmail.com) Received 18 March 211; first decision 16 April 211; accepted 2 July American Journal of Hypertension, Ltd. sphygmomanometer, errors of the cuff measurement would also be incorporated into the estimation of central BP. 12,13 In comparison with the intra-arterial measurement at the brachial artery, the cuff measurement underestimates systolic BP. 14 In the last years, several studies challenged the use of the SphygmoCor device in the estimation of central BP because of a significant deviation from pressures measured with the invasive technique, but did not properly acknowledge the cuff measurement as the major source of bias. 15,16 The recent introduction of another device (Omron HEM- 9AI; Omron Healthcare, Kyoto, Japan) into the market further complicated the issue. 17 The Omron device also records radial pulse waveform, with the calibration of brachial BP, derives the pressure at the second peak of systole, and then calculates central systolic BP using a regression equation with second peak of systole as a major independent variable. 17 The central systolic BP provided by the Omron device under many instances is higher than brachial systolic BP, 18 which contradicts the fact that systolic pressure in general increases from the central aorta to the brachial artery. 1,19 To the best of our knowledge, until recently, this device has not been validated against the invasive catheter measurement by an independent research group. 136 december 211 VOLUME 24 NUMBER AMERICAN JOURNAL OF HYPERTENSION

2 Noninvasive Assessment of Central Pressure original contributions In the present study, we investigated the SphygmoCor and Omron devices in the estimation of central systolic BP in comparison with the invasive catheter measurement. Methods Subjects. Patients hospitalized in the Department of Cardiology in Ruijin Hospital (Shanghai, China) would be potentially eligible for inclusion in the present study, if they were in the age range from 4 to 85 years, and underwent coronary angiography and/or percutaneous coronary intervention, if they did not have cardiac arrhythmia, pacemaker implantation, severe valvular heart disease, acute coronary syndrome within 7 days, or left ventricular ejection fraction less than 5%, and if the difference in systolic BP between the two arms was not more than 5 mm Hg. The study protocol was approved by the Ethics Committee of Ruijin Hospital, Shanghai Jiaotong University School of Medicine (Shanghai, China). All patients gave written informed consent. Of the 44 enrolled participants, 11 were excluded because of the failure to measure central BP by the Omron device (n = 3), or because of frequent atrial or ventricular ectopic beats (n = 3), severe valvular heart disease (n = 3), or left ventricular ejection fraction less than 5% (n = 2), identified during the invasive or noninvasive measurements. Thus, the present analysis included 33 participants, which as planned, was exactly in accordance with the number of participants required by the European Society of Hypertension International Protocol for the validation of BP measuring devices in adults. 2 Hypertension was defined as a BP of at least 14 mm Hg systolic or 9 mm Hg diastolic, or as the use of antihypertensive drugs for controlling BP. Diabetes mellitus was defined as a plasma glucose level of at least 126 mg/dl fasting or at least 2 mg/dl at 2 hours after glucose load, or as the use of antidiabetic drugs or insulin. Dyslipidemia was defined as a serum concentration of triglycerides, total cholesterol, or low-density lipoprotein cholesterol of at least 2 mg/dl, 24 mg/dl or 16 mg/dl, respectively, a serum concentration of high-density lipoprotein cholesterol of 35 mg/dl or lower, or as the use of lipid lowering drugs. Coronary artery disease was diagnosed visually if luminal diameter narrowing was equal to or greater than 5% on at least one of the major epicardial coronary arteries. Noninvasive assessment of central and brachial BPs. The noninvasive measurement of central BPs was performed in the supine position using the SphygmoCor and Omron devices, while the patient was lying on the catheterization table with the left arm stretching out and the palm up. The left radial arterial pulse wave was recorded by the Omron device after a brachial BP was measured oscillometrically with the same device. Then, the radial arterial pulse wave was also recorded by the SphygmoCor device. The same brachial BP measured with the Omron device was used for the calibration of radial pulse waveform. For each patient, the above described procedure was repeated for at least three times to obtain three paired estimations of central BP by both devices. Invasive measurement of central and brachial BPs. Simultaneously with the noninvasive recording of the radial arterial pulse wave, BP was measured invasively with a 6 Fr diagnostic catheter (Infiniti Angiographic Catheter; Cordis, Miami, FL) attached to a fluid-filled manometer system (Mac-Lab Hemodynamic Recording System; GE Healthcare, Milwaukee, WI). The transducer was kept at the level of midaxillary line during the examination and zero calibrated to atmosphere before catheterization. The catheter (1 cm long, 1.45 mm of inner diameter, and.28 mm of wall thickness) was inserted through the right radial artery with its tip steadily positioned in the ascending aorta. The frequency characteristic of the manometric system was determined by the flush test. 21 The natural frequency was confirmed to be greater than 2 Hz, and the damping coefficient was not less than.3. The invasive right brachial BP was measured from the right arm via a 6 Fr radial introducer sheath (Glidesheath; Terumo Medical, Tokyo, Japan) attached to the same manometer. The radial sheath (16 cm long and 2.8 mm of tip inner diameter) was advanced from the right radial artery with its tip at the end of brachial artery. The pressure waves were recorded in the ComboMap Pressure and Flow System (Volcano, San Diego, CA) and analyzed offline with the ComboMap software. Statistical analysis. Data analysis was performed using the SPSS software (version 19.; SPSS, Chicago, IL). The intraclass Table 1 Characteristics of the study subjects Characteristics Overall (n = 33) Age (years) 6.1 ± 8.7 Male, n (%) 21 (63.6) Body height (cm) ± 6.8 Body weight (kg) 71.6 ± 8.3 Body mass index (kg/m 2 ) 25.5 ± 2.9 Systolic blood pressure (mm Hg) 132 ± 16 Diastolic blood pressure (mm Hg) 78 ± 8 Pulse rate (beats/minute) 77 ± 8 Current smoking, n (%) 12 (36.4) Hypertension, n (%) 23 (69.7) Diabetes mellitus, n (%) 8 (24.2) Dyslipidemia, n (%) 14 (42.4) Coronary artery disease, n (%) 17 (51.5) Left ventricular ejection fraction (%) 66.8 ± 4.5 Use of medications, n (%) Angiotensin-converting enzyme inhibitors 14 (42.4) Angiotensin receptor blockers 1 (3.3) β-blockers 18 (54.5) Calcium channel blockers 17 (51.5) Diuretics 2 (6.1) Statins 2 (6.6) Values are mean ± s.d. or number of subjects (%). For the definition of hypertension, diabetes mellitus, dyslipidemia, and coronary artery disease, see Methods. AMERICAN JOURNAL OF HYPERTENSION VOLUME 24 NUMBER 12 december

3 original contributions Noninvasive Assessment of Central Pressure y =.83x r =.91 n = y =.97x r =.9 n = 99 csbpsphy (mm Hg) csbpom (mm Hg) csbpcath (mm Hg) 3 csbpsphy csbpcath (mm Hg) 2 1 Mean + 2 s.d. 1 Mean 2 csbpom csbpcath (mm Hg) csbpcath (mm Hg) 3 Mean + 2 s.d. 2 1 Mean 1 2 Mean 2 s.d. 3 3 Mean 2 s.d (csbpsphy + csbpcath)/2 (mm Hg) (csbpom + csbpcath)/2 (mm Hg) Figure 1 Correlation (upper panels) and Bland Altman (bottom panels) plots for central systolic blood pressure measured with the SphygmoCor (left panels) and Omron (right panels) devices against the catheter measurement. For the correlation analysis, the regression equation and correlation coefficient are given, and the lines of identity (dotted lines) and regression (solid lines) were drawn. For the Bland Altman plot, the lines at the mean and plus and minus 2 standard deviations were drawn. csbpsphy, csbpom, and csbpcath indicate central systolic blood pressure derived from the SphygmoCor, Omron devices and the catheter measurement, respectively. correlation analysis, the paired Student s t-test and Bland Altman plots were used to assess the agreement between the noninvasive and invasive BP measurements. Correlation analysis for the differences between and the mean values of the paired noninvasive and invasive measurements was performed with the Pearson s method. Results Characteristics of the study participants The 33 study participants included 21 men (63.6%), 23 hypertensive patients (69.7%), and 17 patients with coronary artery disease (51.5%, Table 1). Overall, age (± s.d.) averaged 6.1 (± 8.7) years and ranged from 45 to 83 years. Mean body mass index was 25.5 (± 2.9) kg/m 2. All patients took BP lowering drugs, and 2 (6.6%) took a statin. Comparison between noninvasive and invasive central BPs The central systolic BPs estimated with both noninvasive devices were significantly (P <.1) and closely correlated with the catheter measurement at the ascending aorta, with an intraclass correlation coefficient of.91 and.9 for the SphygmoCor and Omron devices, respectively (Figure 1). However, both devices significantly underestimated central systolic BP with an average difference of 15 mm Hg (95% CI, 17 to 13 mm Hg, P <.1) and 2 mm Hg (95% CI, 4 to mm Hg, P <.5) for the SphygmoCor and Omron devices, respectively (Table 2 and Figure 1). The correlation coefficient for the differences between and the mean values of the noninvasive and invasive central systolic BPs was.1 (P <.5) and.8 (P =.9) for the SphygmoCor and Omron devices, respectively. Because the Omron device does not provide central diastolic BP estimation, we only compared the noninvasive estimations of diastolic and pulse pressures by the SphygmoCor device against the invasive measurement. Diastolic BP and pulse pressure estimated by the SphygmoCor device were also significantly (P <.1) and closely correlated with the invasive measurements, with an intraclass correlation coefficient of.74 and.93, respectively. The mean difference between the noninvasive and invasive measurements was small for diastolic BP (P =.91), and of similar amplitude as systolic BP for pulse pressure (P <.1, Table 2). The correlation coefficient for the differences between and the mean values of the noninvasive and invasive pulse pressures was.18 (P <.1). Comparison between the noninvasive and invasive brachial BPs The oscillometrically measured brachial BP was significantly (P <.1) and closely correlated with the invasive catheter measurement at the brachial artery, with an intraclass correlation coefficient of.89,.84, and.87 for systolic, diastolic, and pulse pressure, respectively (Figure 2). However, the 138 december 211 VOLUME 24 NUMBER 12 AMERICAN JOURNAL OF HYPERTENSION

4 Noninvasive Assessment of Central Pressure original contributions Table 2 Brachial and central blood pressures measured with the SphygmoCor and Omron devices in comparison with the invasive catheter measurement Central pressure (n = 99) SphygmoCor Noninvasive devices Invasive catheter a Difference from the catheter (95% confidence interval) P value Systolic blood pressure (mm Hg) 123 ± ± ( 17 to 13) <.1 Diastolic blood pressure (mm Hg) 78 ± ± 1 ( 2 to +1).91 Pulse pressure (mm Hg) 45 ± 14 6 ± ( 16 to 14) <.1 Omron Systolic blood pressure (mm Hg) 136 ± ± 22 2 ( 4 to ) <.5 Brachial pressure (n = 33) Systolic blood pressure (mm Hg) 137 ± ± ( 23 to 15) <.1 Diastolic blood pressure (mm Hg) 77 ± ± 13 2 ( 5 to ).6 Pulse pressure (mm Hg) 6 ± ± ( 2 to 12) <.1 Values are mean ± s.d., unless otherwise indicated. a Blood pressure was invasively measured three times in each of the 33 participants in the central aorta but only once in the brachial artery. The comparison was therefore performed in 99 pairs for central blood pressure and in 33 pairs for brachial blood pressure. bsbposci (mm Hg) y =.89x r =.88 n = 33 of the oscillometric and invasive measurements was.28 (P <.1) for systolic BP and.41 (P <.1) for pulse pressure. We then computed systolic pressure amplification from central aorta to brachial artery by subtracting central from brachial systolic BP for the invasive catheter technique, and for the SphygmoCor and Omron devices, separately. The mean (s.d.) systolic pressure amplification was 18 (13), 12 (6), and 1 (1) mm Hg, respectively. bsbposci bsbpcath (mm Hg) bsbpcath (mm Hg) Mean + 2 s.d. Mean 4 Mean 2 s.d (bsbposci + bsbpcath)/2 (mm Hg) Figure 2 Correlation (upper panel) and Bland Altman (bottom panel) plots for brachial systolic blood pressure measured oscillometrically against the catheter measurement. bsbposci and bsbpcath indicate brachial systolic blood pressure measured oscillometrically and invasively, respectively. For further information, see legend to Figure 1. noninvasive oscillometric method significantly (P <.1) underestimated systolic and pulse pressure by 19 mm Hg (95% CI 23 to 15 mm Hg) and 16 mmhg (95% CI 2 to 12 mm Hg), respectively (Table 2 and Figure 2). The correlation coefficient for the difference between and the mean values Discussion To the best of our knowledge, our study was the first that simultaneously validated the SphygmoCor and Omron devices in the estimation of central BP against the invasive catheter measurement. The noninvasive and invasive central systolic BPs were highly correlated. However, neither SphygmoCor nor Omron was able to provide accurate assessment of real aortic pressure values, although the underestimation was less substantial by Omron than SphygmoCor. The greater underestimation by SphygmoCor was apparently introduced from the cuff measurement of brachial BP. Our finding on the comparison between SphygmoCor and Omron is in line with the results of a recent study that compared these two devices. 18 In 33 subjects recruited in the United Kingdom, Richardson et al. found that the estimated central systolic BP was 12 mm Hg lower with the SphygmoCor than Omron device. Of course, without invasive catheter measurements in this study, it is difficult to judge which device is more accurate. These similar findings across populations suggest that the difference in the estimation of central systolic BP is inherent with their technological differences, and that central BP values are not exchangeable between the two devices, and device specific reference values would be required for clinical decisions, if any. Our finding on the comparison between the SphygmoCor device and the catheter measurement is also in keeping with AMERICAN JOURNAL OF HYPERTENSION VOLUME 24 NUMBER 12 december

5 original contributions Noninvasive Assessment of Central Pressure the results of previous studies in two different populations. 15,16 Whenever the sphygmomanometrically measured brachial BP was used for the calibration, the SphygmoCor device substantially underestimated central systolic BP. The differences varied from 13 mm Hg in Caucasians 15 to 4 mm Hg in Chinese. 16 As these previous studies 15,16 and present research clearly indicated, and O Rourke repeatedly argued, 13 it is the brachial BP measured with cuff rather than the generalized transfer function, which is the criminal source of this systemic error. However, the paradoxical point is that brachial BP measured with cuff is still the core in the definition and management of hypertension, and hence indispensable now and in the foreseen future. The way forward for the SphygmoCor device, therefore, is either to give up the dependence on the use of cuff pressure for calibration, which is very unlikely, to use parameters largely independent of the level of cuff pressure, such as augmentation index or pressure amplification from central aorta to the brachial arteries, which is the difference or ratio between the brachial and central pressures, or simply to continue with the current recommendations, 1 since superiority of central pressure over brachial cuff pressure in monitoring therapeutic effects 6,22 and predicting cardiovascular events 2,3 has been well established on the basis of the SphygmoCor technique. The manufacturer of the Omron device was aware about the influence of the cuff pressure on the estimation, and accounted for the difference between the cuff and catheter brachial BPs in the development of their algorithm for the estimation of central BP. 17 To avoid using generalized transfer function, the Omron device estimates central systolic BP using a regression equation that includes the second peak of systole in the radial artery, while accounting for several major confounding variables, such as sex and age. These technological approaches bridged to some extent the gap between the noninvasive and invasive central blood measurements, but also introduced more random errors, as suggested by the more prominent dispersity in the differences between the noninvasive and invasive measurements. Because of the technological differences between SphygmoCor and Omron, the Omron device apparently requires its own reference values as diagnostic thresholds. Finally, the fact that central BP provided by the Omron device is often higher than brachial cuff BP may also be disturbing. Our study has to be interpreted within the context of its strengths and limitations. One of major strengths of our study is that two devices have been simultaneously validated against the invasive catheter measurement. In addition, the oscillometrically and invasively measured brachial BP were also compared. However, all of our study subjects were treated with drugs that may influence central BP measurements. One cannot determine whether the noninvasive measures are confounded to different extents by various drug treatments. The absence of an untreated normotensive group is a limitation of our study. In addition, we used a fluid-filled catheter to measure central BPs, which is known to have the occasional phenomena of excessive damping and attenuation resulting suboptimal BP waveforms. 23 However, it has the advantage of a stable external zero, while a catheter-mounted sensor records an offset which depends on the vertical position of the sensor in the body. In conclusion, the noninvasive assessment of central systolic BP is highly correlated with the invasive catheter measurement, but inaccurate in the estimation of real aortic pressure values, because of systemic errors from the inaccurate sphygmomanometer for SphygmoCor, and because of random errors from the regression equation for Omron. Nonetheless, these devices can be properly used in the research setting, on the condition that diagnostic thresholds specific for a certain device become available. In the clinical setting, where we accept brachial cuff pressure as the current standard in the care of hypertension, we need to recognize that this correlates with, but actually differs from intra-arterial brachial pressure measurement. Acknowledgments: The authors gratefully acknowledge the participation of all study participants, and the technical assistance of Miss Jie Wang. The study was partially supported by a grant from Omron Healthcare (China), which is a branch of the manufacturer of the Omron HEM-9AI device, Omron Healthcare, Kyoto, Japan. Disclosure: J.-G.W. reports receiving funding from the National Natural Science Foundation of China (grants ) and the Ministry of Science and Technology (grant 26BAI1A3 and a grant from China-European Union Collaborations (111)), Beijing, China, the Shanghai Commissions of Science and Technology (grant 7JC1447) and Education (grant 7ZZ32), and the Shanghai Shenkang Hospital Development Centre (grant SHDC127318) and the European Union (grants LSHM-CT and HEALTH-F ) and consulting and lecture fees from GSK, Omron, Pfizer, Sankyo, Sanofi-Aventis, and Servier. The other authors have no conflict of interest. 1. Agabiti-Rosei E, Mancia G, O Rourke MF, Roman MJ, Safar ME, Smulyan H, Wang JG, Wilkinson IB, Williams B, Vlachopoulos C. Central blood pressure measurements and antihypertensive therapy: a consensus document. Hypertension 27; 5: Roman MJ, Devereux RB, Kizer JR, Lee ET, Galloway JM, Ali T, Umans JG, Howard BV. Central pressure more strongly relates to vascular disease and outcome than does brachial pressure: the Strong Heart Study. Hypertension 27; 5: Wang KL, Cheng HM, Sung SH, Chuang SY, Li CH, Spurgeon HA, Ting CT, Najjar SS, Lakatta EG, Yin FC, Chou P, Chen CH. Wave reflection and arterial stiffness in the prediction of 15-year all-cause and cardiovascular mortalities: a communitybased study. Hypertension 21; 55: Li Y, Staessen JA, Li LH, Huang QF, Lu L, Wang JG. Reference values for the arterial pulse wave in Chinese. Am J Hypertens 28; 21: Chen X, Li Y, Sheng CS, Huang QF, Zheng Y, Wang JG. Association of serum uric acid with aortic stiffness and pressure in a Chinese workplace setting. Am J Hypertens 21; 23: Williams B, Lacy PS, Thom SM, Cruickshank K, Stanton A, Collier D, Hughes AD, Thurston H, O Rourke M. Differential impact of blood pressure-lowering drugs on central aortic pressure and clinical outcomes: principal results of the Conduit Artery Function Evaluation (CAFE) study. Circulation 26; 113: McEniery CM, Yasmin, McDonnell B, Munnery M, Wallace SM, Rowe CV, Cockcroft JR, Wilkinson IB. Central pressure: variability and impact of cardiovascular risk factors: the Anglo-Cardiff Collaborative Trial II. Hypertension 28; 51: Karamanoglu M, O Rourke MF, Avolio AP, Kelly RP. An analysis of the relationship between central aortic and peripheral upper limb pressure waves in man. Eur Heart J 1993; 14: Pauca AL, O Rourke MF, Kon ND. Prospective evaluation of a method for estimating ascending aortic pressure from the radial artery pressure waveform. Hypertension 21; 38: Gallagher D, Adji A, O Rourke MF. Validation of the transfer function technique for generating central from peripheral upper limb pressure waveform. Am J Hypertens 24; 17: O Rourke MF, Adji A. An updated clinical primer on large artery mechanics: implications of pulse waveform analysis and arterial tonometry. Curr Opin Cardiol 25; 2: december 211 VOLUME 24 NUMBER 12 AMERICAN JOURNAL OF HYPERTENSION

6 Noninvasive Assessment of Central Pressure original contributions 12. Smulyan H, Siddiqui DS, Carlson RJ, London GM, Safar ME. Clinical utility of aortic pulses and pressures calculated from applanated radial-artery pulses. Hypertension 23; 42: O Rourke MF, Adji A. Clinical use of applanation tonometry: hope remains in Pandora s box. J Hypertens 21; 28: Ochiai H, Miyazaki N, Miyata T, Mitake A, Tochikubo O, Ishii M. Assessment of the accuracy of indirect blood pressure measurements. Jpn Heart J 1997; 38: Cloud GC, Rajkumar C, Kooner J, Cooke J, Bulpitt CJ. Estimation of central aortic pressure by SphygmoCor requires intra-arterial peripheral pressures. Clin Sci 23; 15: Zuo JL, Li Y, Yan ZJ, Zhang RY, Shen WF, Zhu DL, Gao PJ, Chu SL. Validation of the central blood pressure estimation by the SphygmoCor system in Chinese. Blood Press Monit 21; 15: Takazawa K, Kobayashi H, Shindo N, Tanaka N, Yamashina A. Relationship between radial and central arterial pulse wave and evaluation of central aortic pressure using the radial arterial pulse wave. Hypertens Res 27; 3: Richardson CJ, Maki-Petaja KM, McDonnell BJ, Hickson SS, Wilkinson IB, McEniery CM. Comparison of estimates of central systolic blood pressure and peripheral augmentation index obtained from the Omron HEM-9AI and SphygmoCor systems. Artery Research 29; 3: Van Bortel LM, Struijker-Boudier HA, Safar ME. Pulse pressure, arterial stiffness, and drug treatment of hypertension. Hypertension 21; 38: O Brien E, Atkins N, Stergiou G, Karpettas N, Parati G, Asmar R, Imai Y, Wang J, Mengden T, Shennan A. European Society of Hypertension International Protocol revision 21 for the validation of blood pressure measuring devices in adults. Blood Press Monit 21; 15: Gardner RM. Direct blood pressure measurement dynamic response requirements. Anesthesiology 1981; 54: de Luca N, Asmar RG, London GM, O Rourke MF, Safar ME. Selective reduction of cardiac mass and central blood pressure on low-dose combination perindopril/ indapamide in hypertensive subjects. J Hypertens 24; 22: Ercole A. Attenuation in invasive blood pressure measurement systems. Br J Anaesth 26; 96: AMERICAN JOURNAL OF HYPERTENSION VOLUME 24 NUMBER 12 december