Non-invasive measurement of aortic pressure in patients: Comparing pulse wave analysis and applanation tonometry.

OBJECTIVE: The aim of the present study was to validate and compare novel methods to determine aortic blood pressure non-invasively based on Oscillometric Pulse Wave Velocity (PWV) measurement using four limb-cuff pressure waveforms and two lead Electrocardiogram (ECG) with a validated tonometric pulse wave analysis system in patients.
MATERIALS AND METHODS: After receiving the consent, in 49 patients with hypertension, coronary artery disease, diabetes mellitus, PWV, and central blood pressures were recorded in a randomised manner using both the oscillometric and tonometric devices. All recordings were performed 10 minutes after the patient lying comfortably in a noise-free temperature-controlled room. The test was performed between 09 am and 10 am after overnight fast. A minimum of three measurements were performed by the same skilled and trained operator. From the raw data obtained with two devices, software calculated the final vascular parameters.
RESULTS: A total of 49 patients (8 women and 41 men), of mean age 40.5 years (range: 19-81 years) participated in the present study. After transforming the brachial pressures into aortic pressures, the correlation coefficient between the Aortic Systolic Pressure (ASP) values obtained with two methods was 0.9796 (P<0.0001). The mean difference between ASP with two methods was 0.3 mm Hg. Similarly, Aortic Diastolic Pressure (ADP) values obtained with two methods also correlated significantly with correlation coefficient of 0.9769 (P<0.0001). The mean difference of ADP was 0.2 mm Hg. In case of Aortic Pulse Pressure (APP), the mean difference was 0.1 mm Hg. All parameters of central aortic pressures obtained with two methods correlated significantly.
CONCLUSION: The new method of transforming the Carotid Femoral PWV (cfPWV) and brachial blood pressure values into aortic blood pressure values seems to be reasonably good. The significant correlation between the values obtained by tonometric device and oscillometric PWV method shows that the latter can be used non-invasively in patients to find the aortic pressure.

Introduction

Recently, many studies have indicated that arterial stiffness is an independent predictor of increased morbidity and mortality for Cardiovascular Diseases (CVD) such as atherosclerosis[1] in general, and coronary artery disease,[2-4] peripheral artery disease,[5] chronic kidney disease,[67] hypertension,[89] cerebral stroke,[10] in particular. Being a reliable measure of arterial stiffness, Pulse Wave Velocity (PWV) has been suggested as a better marker of arterial stiffness.[11] With the awareness of arterial stiffness, clinicians have started adding measurement and treatment of underlying arterial stiffness in their clinical practice.[12] International healthcare societies like the European Society of Hypertension (ESH) and European Society of Cardiology (ESC) have included PWV measurement in their 2003 guidelines for management of hypertension.[1314]

The pulse pressure wave travels to the peripheral arteries and reflects back from the vasculature. In normal subjects, the reflected pulse pressure wave reaches the origin (aortic root) during the diastole phase of the cardiac cycle. This reflected pressure wave increases the Diastolic Pressure (ADP) gradient at the aortic root, which facilitates the Coronary Artery Blood Flow (CBF). But in the presence of arterial stiffness, the PWV is increased. As a consequence of this increased PWV, the pulse pressure wave is reflected prematurely and much early in a cardiac cycle. This leads to an elevation of the Aortic Systolic Pressure (ASP), decrease in ADP, and a rise in the Aortic Pulse Pressure (APP).

The rise in systolic pulse pressure is called as an augmentation pressure and its ratio to the APP is called as Augmentation Index (AIx). This augmentation of ASP leads to increased afterload and gives rise to the Left Ventricular Hypertrophy (LVH). Studies have shown that AIx correlates with the left ventricular mass in hypertensive patients.[15] Its association with an increased risk of stroke has also been suggested. As the role of increased in ASP and AIx is multifold, its measurement can be used as a surrogate and supplementary measure for arterial stiffness.[16]

Device and method, which can non-invasively measure both the PWV and Central blood pressure, will be of immense help for carrying out large population-based studies on arterial stiffness.

Since the aortic pressure values are deemed to be an outcome of the brachial pressure as well as the PWV, and PeriScope, which has a provision for recording brachial blood pressures and PWV, it is suggested that PeriScope may be capable of finding aortic pressures. In a unpublished pilot study at cardiac catheter laboratories during angiography, the values of invasive aortic pressures found by a fluid-filled catheter method were compared with brachial pressures and cfPWV obtained by PeriScope using multivariate statistical analysis. There was a good correlation in ASP and ADP and the regression equations applied were found to convert the brachial blood pressure and PWV values into equivalent aortic pressures. Thus, the aim of the present study was to validate a novel method determining aortic systolic, diastolic, pulse, augmentation pressures along with AIx based on an oscillometric method using four limb-cuff pressure waveforms and two lead ECG (PeriScope) with a validated tonometric Pulse Wave Analysis System (SphygmoCor, AtCor Medical Pty Ltd, West Ryde, Australia).

Materials and Methods

Study Population

Patients (with hypertension, CAD, and DM) attending the outpatient department of Nizam's Institute of Medical Sciences, Hyderabad, India, were enrolled in the study. The institutional ethics committee had approved the study protocol and informed consent was obtained before the measurements. Patients with atrial fibrillation or unstable clinical presentation were excluded from the study. A total of 49 patients participated in the present study; Table 1 presents their demographic data.

Table 1

Demographic characteristics of study participants

In a randomised manner, PWV and central blood pressures were recorded using PeriScope as described earlier[17] and with Sphygmocor as per the procedure described in the instruction manual. The international recommendations for the measurement of arterial stiffness were respected.[18]

Periscope Method in Brief

PeriScope (Genesis Medical Systems, Hyderabad, India) is a validated[17] and tested medical device used to measure Brachial Ankle PWV (baPWV) and derives the Carotid Femoral PWV (cfPWV), which is equivalent to aortic PWV. This device uses four blood pressure cuffs to be tied around four limbs of the subjects. The cuffs are used to determine blood pressure values at the brachium and ankle of the respective limb by the oscillometric method. This method also records arterial pressure waveforms at predetermined cuff pressures. It has two-channel ECG leads to record ECG simultaneously. Based on simultaneous ECG and pressure waveform recording, PeriScope calculates the pulse wave velocities.

This method is devoid of any operator bias since it is fully automated and no operator handling of any probe is involved to record the waveforms. It does not require any special operative skills and can be carried out in any clinical setup just like an ECG recording. In a unpublished pilot study at cardiac catheter laboratories during angiography, the values of invasive aortic pressures found by a fluid-filled catheter method were compared with brachial pressures and cfPWV obtained by PeriScope using multivariate statistical analysis.

There was good correlation (ASP): r=0.9526, P<0.0001; ADP: r=0.9168, P<0.0001). Regression equations applied were found to convert the brachial blood pressure and PWV values into equivalent aortic pressures. These regression equations were introduced into the software of PeriScope to determine the aortic pressure values.

Recordings were taken 10 minutes after the patient lying comfortably in a noise-free temperature-controlled room. All tests were performed after overnight fast between 09 am and 10 am. On each occasion, a minimum of three measurements were taken by the same skilled and well-trained operator. From the raw data obtained by the two devices, the software calculated the final vascular parameters.

Statistics

The regression analysis was carried out to find the correlation between the measurements from the two devices. All values (demographics) are stated as mean and Standard Deviation (SD). Vascular data are also analysed using the methods presented by Bland-Altman.[19] Bland-Altman plots represent the data graphically and analyse the reproducibility of measurements according to different methods and the resulting corrections for SD. For statistical analysis, the GraphPad Instat software (Graphpad Inc, USA) was used.

Results

A total of 49 patients (8 women and 41 men), with mean age 40.5 years (range: 19-81 years) participated in the present study. Medical treatment was not withheld for the measurements. The mean blood pressure was systolic 134.6 ± 22.4 mm Hg and diastolic 76.3 ± 13.8 mm Hg.

Comparison of ASP, ADP, and APP

In the present study, after transforming the brachial into aortic pressures, the correlation between the ASP values obtained with two methods was highly significant: Correlation coefficient of 0.9796, P<0.0001. The mean difference between ASP with two methods was 0.3 mm Hg with SD of ±4.8 mm Hg [Figure 1a].

Figure 1

The correlation of AoSysP and AoDiaP between the two methods is extremely significant

The AAMI 1992 standard[20] allows mean difference 5.0 mmHg with an SD of +/–8.0 mmHg between the standard method whereas Grade ‘A’ BHS protocol for Blood pressure measurements stipulates >60% readings in +/–5 mmHg, > 85% in +/–10 mmHg and >95% +/–15 mmHg.[21]

ADP values obtained by both methods correlated significantly with correlation coefficient of 0.9769 and P<0.0001. The mean difference of ADP was 0.2 mm Hg, with SD of ±2.87 mm Hg [Figure 1b].

Estimated values of ASP and ADP by the two methods did not differ much and the difference from the mean was within 2 SD. The difference spread for Bland–Altman plots in Figures 2a and 2b for ASP and ADP show a satisfying spread of residues, respectively.

Figure 2

Bland–Altman plots of AoSysP and AoDiaP from both the methods show a good confirmation with each other

Comparison of Augmentation Pressure and AIx

For Aug P, the mean difference was –0.08 mm Hg, with an SD of ±4.9 mm Hg, and for AIx, the mean difference was –0.8% with SD ±11%. With AIx, being a ratio of Augmentation Pressure to pulse pressure multiplied by 100, and also sensitive to small changes, a higher SD is observed. Figure 3a and 3b illustrate the uniform distribution of the residuals.

Figure 3

Bland–Altman plots of AugP and AIx from both the methods show an even spread of residuals

Discussion

In our earlier study, using a non-invasive PWV measurement with an oscillometric method, we have established the role of arterial stiffness in various CVD.[11] Aortic pressure is a surrogate marker for arterial stiffness. It is important to treat patients with elevated arterial stiffness. Additional estimation of aortic blood pressure with this device will significantly enhance its usage in clinical practice. Since there is a relationship between arterial stiffness, brachial blood pressure, and central pressure, it is logical to explore the possibility of adding central blood pressure parameters to the existing device, which was successfully implemented.

Pulse wave velocity, which is a relevant indicator of arterial stiffness, can be measured non-invasively with a variety of automatic devices. A novel index for estimating arterial stiffness as ‘QPV interval’ Was determined by means of surface ECG and Doppler ultrasound of the brachial artery simultaneously .[22] This method is complex and requires a skilled operator.

The present method was further validated by comparing it with an established method of Generalized Transfer Function (GTF) using tonometry. The results obtained by the new PeriScope method were compared with that obtained by SphygmoCor, which is a widely used device with tonometry and GTF. Our study showed that the results obtained by these two methods correlated well, with no significant differences. The PeriScope method is able to calculate the central aortic pressure non-invasively which can be applied in routine clinical setting. Being simple and operator-independent, the new device can thus be useful in large population-based studies for identifying patients with a high risk of CVD mortality and morbidity associated with increased arterial stiffness. Since PeriScope is an automated device without any user intervention, the variations are significantly low.

The Bland–Altman analysis shows proper spreading of residues, and therefore no dependence on mean and difference. The analysis showed that the measurements were well within the tolerance band set up by the Association for the Advancement of Medical Instrumentation (AAMI). However, in case of AIx, though the mean difference is low, the SD is high. This could be attributed to the formula for AIx, in which the ratio of Augmentation Pressure to pulse pressure is multiplied by 100.

i.e. Augmentation Index=100*(Augmentation Pressure/Aortic Pulse Pressure), where Augmentation Pressure is the increase in Aortic Systolic pressure due to arterial stiffness. This determination of the AIx gives large scope of variation even with a small change in Augmentation Pressure or APP value.

Conclusion

With the new method of transforming the cfPWV and brachial blood pressure values into aortic pressure values appears reasonable. The good correlation between the results obtained by tonometric GTF method and oscillometric PWV method shows that the latter can be used non-invasively to find the aortic pressure in patients. However, more studies are warranted in large patient population.