NDT Advance Access published online on November 7, 2007
Nephrology Dialysis Transplantation, doi:10.1093/ndt/gfm470
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Reproducibility of pulse-wave analysis and pulse-wave velocity determination in chronic kidney disease
1Department of Nephrology, Herlev Hospital and 2Rigshospitalet, University of Copenhagen, Denmark
Correspondence and offprint requests to: Dr M. Frimodt-Møller, Laboratory of Nephrology 5403, Herlev Hospital, Herlev Ringvej 75, 2730 Herlev, Denmark. Email: marfri01{at}heh.regionh.dk
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Background. Indices of central arterial stiffness, derived by use of applanation tonometry, have shown to be strong independent predictors of cardiovascular morbidity and mortality in patients with chronic kidney disease (CKD). The objective of this study was to evaluate the intra- and inter-observer and day-to-day reproducibility of pulse-wave analysis (PWA) and pulse-wave velocity (PWV) in pre-dialysis patients with CKD stages 3–5 using applanation tonometry with the SphygmoCor® software and hardware.
Methods. Double recordings of the radial pressure waveform and the aortic and brachial PWV were performed under standardized conditions in 19 CKD patients with a mean GFR 25.3 ml/min/1.73 m2 (range 9.9–42.2) by two trained observers and repeated by one of the observers within a week.
Results. The mean inter-observer and day-to-day differences (±2 SD) for the augmentation index (AIx) were 0.9 ± 15.8% and 2.6 ± 11.2%, for subendocardial viability ratio (SEVR) –0.9 ± 15.5% and –0.4 ± 24.7%, for aortic pulse pressure (PP) 1.4 ± 13.3 mmHg and 0.3 ± 20.9 mmHg and for aortic PWV 0.3 ± 3.2 m/s and –0.7 ± 1.9 m/s, respectively. Intra-observer differences were calculated for each of three sets of double measurements and showed good reproducibility as well. Calculations on sample size needed in a clinical trial showed a limited number of patients needed in a clinical study over time.
Conclusions. PWA and PWV based on applanation tonometry using the SphygmoCor® software and hardware are highly reproducible in pre-dialysis patients with CKD with the day-to-day variation being in accordance with the intra- and inter-observer variation. Thus, applanation tonometry using the SphygmoCor® system is a simple, non-invasive method to assess central haemodynamics in clinical trials in patients with pre-dialysis CKD with only a limited number of patients needed to detect significant differences.
Keywords: augmentation index; blood pressure; cardiovascular disease; chronic kidney disease; pulse-wave velocity; reproducibility
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionAcknowledgementsReferences |
|---|
There is a high cardiovascular morbidity and mortality among patients with chronic kidney disease (CKD) [1] and strategies to reduce this increased risk of cardiovascular disease are therefore of major importance. When studying the effect of therapeutic interventions on cardiovascular disease, the most relevant endpoints are death and cardiovascular morbidity. Markers of arterial stiffness such as aortic pulse-wave velocity (PWV) and augmentation index (AIx), obtained by pulse-wave analysis (PWA) have, in recent studies, been shown to be strong independent predictors of cardiovascular morbidity and mortality in patients with CKD [2–4]. These parameters can be assessed non-invasively by use of applanation tonometry [5–7]. To evaluate PWA and PWV in long term studies the reproducibility is critical. In healthy individuals, intra- and interobserver reproducibility of the method has been shown to be high [8–11], but in patients with CKD information on reproducibility is limited [12,13]. Reproducibility has been evaluated in patients receiving haemodialysis [12] and in a mixed group of healthy controls, pre-dialysis, dialysis treated and renal transplanted patients [13]. Reproducibility has never been systematically evaluated in a group consisting of solely pre-dialysis patients. The aim of the present study was to assess intra-and interobserver as well as day-to-day reproducibility of PWA and PWV using the SphygmoCor ® hard- and software in pre-dialysis patients with CKD stage 3-5.
|
Subjects and methods |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Subjects
All subjects were recruited from the outpatient clinic at the Department of Nephrology, Herlev Hospital. Nineteen patients (11 male and 8 female), mean age 57 years (range 28–80) took part in the study. Inclusion criteria were age between 18 and 80 years, CKD with plasma-creatinine between 150 and 450 µM and a written informed consent. Seven patients had diabetic nephropathy, three had chronic glomerulonephritis, three had hypertensive nephropathy, two had adult polycystic kidney disease, one had tubulo-interstitial nephropathy and three had chronic nephropathy of unknown aetiology. The study was approved by The Ethical Committee of Copenhagen County. The demographic data are shown in Table 1.
|
Study protocol
Examinations were made from 9 a.m. on two different days, within a week, after a minimum 10 min rest in supine position in a quiet room at a constant temperature of 23–25°C. No consumption of alcohol was allowed for 24 h before examination, and no tea, coffee or smoking was allowed from 8 h before examination. All subjects were fasting for 8 h except those having diabetes mellitus, who had a light meal before examination. Morning medication of the participants was postponed until after the examinations with the exception of anti-diabetic medicine which was taken as usual.
Pulse-wave analysis and pulse-wave velocity
All measurements of PWA and PWV were performed using an applanation tonometer (Millar, SPT-301B, Houston, TX, USA) and the SphygmoCor® hardware and software (version 7.0, Atcor Medical, Sydney, Australia) as double-recordings by two trained observers at day 1 in random order. Recordings were repeated by one of the observers at day 2. The brachial blood pressure was the average of the last two out of three blood pressure recordings measured by use of a mercury manometer. The method of PWA and PWV measurements has been described in detail elsewhere [5,6]. Briefly PWA is based on a mean of 10 s of tonometer-recorded radial pulse-wave forms and a brachial blood pressure. From this the estimated central aortic pulse-waveform can be derived, using an algorithm, the ‘generalized transfer function’. PWV is determined as the difference in travel time of the pulse-wave between two different recording sites and the heart, divided by the travel distance of the pulse-wave form. An electrocardiogram (ECG) is used to determine the start of the pulse-wave. The mean of 10 s of tonometer recorded pulse-waves at either the radial and carotid artery (brachial PWV) or the femoral and carotid artery (aortic PWV) is used to determine the arrival of the pulse-wave at the peripheral recording site. The distance is measured by a tape measure between the recording sites and the suprasternal notch: when determining the aortic PWV the distance from the carotid location to the suprasternal notch is subtracted from the distance between the suprasternal notch and the femoral recording site. Likewise, the distance from the carotid recording site to the suprasternal notch is subtracted from the distance between the suprasternal notch and the radial sample site, when determining brachial PWV.
Quality control
PWA
Visually acceptable recordings of a peripheral pulse-waveform were only accepted if the variations in pulse height, diastole and pulse length were equal or less than 5% and the mean pulse height was above 80 mV as expressed by a quality index (%) provided by the software. An index greater than 80% was accepted.
PWV
Visually acceptable pulse-waveforms and a mean pulse height above 80 mV were required and the time difference between the ECG-signal and the signal from the recording sites should have an SD of less than 10% of the mean value.
Arterial haemodynamics
The following parameters were calculated (Figure 1); (i) aortic augmentation index without and with correction for a heart rate of 75 (AIx, AIx@HR75) calculated as the difference between the first and the second systolic shoulders divided by the pulse pressure, (ii) time to reflection (TR) defined as the total travel time for the incident pulse wave to the periphery and its return, (iii) subendocardial viability ratio (SEVR), calculated as the ratio of the diastolic pressure time index and the systolic pressure time index, (iv) the estimated aortic blood pressures (v) aortic and brachial PWV (vi) heart rate (HR) and finally (vii) aortic pulse pressure (PP) calculated as the difference between the estimated aortic systolic and diastolic BP.
|
Statistics
The day-to-day and intra- and inter-observer reproducibility have been evaluated by assessing the agreement between measurements by (i) looking for statistically significant bias using a paired two-tailed t-test analysis and (ii) calculating limits of agreement using Bland-Altman's plots. In these plots the differences between the studied parameters were plotted against their mean values. Limits of agreement were considered as being within 2 SDs of the mean differences, thus expressing the expected variation in 95% of the cases [14]. Furthermore the day-to-day reproducibility was evaluated by looking at the reliability of measurements in terms of the intraclass correlation coefficient [15]. The intraclass correlation coefficient expresses the ability of a measurement to discriminate between different groups of subjects/patients by calculating the contribution of patient variance to total variance. Sample size needed in a clinical study was calculated by the statistical program, ‘R. Foundation for statistical computing, 2006’. The data are presented as mean ± 2 SD unless otherwise stated. A P-value of less than 0.05 was considered as significant. Data were analysed by use of a statistical computer program (Statistica 6.1, StatSoft, Inc., OK, US).
|
Results |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionAcknowledgementsReferences |
|---|
All pulse-wave recordings in the study were within our quality standard, as described above in the Subjects and Methods section. The mean quality index was 95% (range 81–100%, SD 5.4%).
Intra-observer variation
Calculation of intra-observer reproducibility was based on each of the three sets of double recordings of each patient: double recordings performed by both observers at day 1 and repeated by one of the observers at day 2. Bland-Altman plots were constructed for all data. An example of a Bland–Altman plot for the intra-observer reproducibility is shown in Figure 2. None of the plots showed any sign of the mean differences being dependent on the underlying mean values. Significant differences between the first and second recordings performed by the same observer were found for observer 1 in aortic PWV and HR and for observer 2, day 2 in aortic diastolic BP (Table 2). In general the mean differences and variations were less for observer 2 than observer 1, which is probably due to a larger experience in performing pulse-wave measurements of observer 2 than observer 1.
|
|
Inter-observer variation
Calculation of inter-observer variation was based on the mean of 19 double recordings performed by observer 1 and by observer 2 at day 1. Mean differences were all close to zero with acceptable variations and without any significant differences (Table 3). Bland–Altman plots were constructed for all data. An example of such a plot for the inter-observer variation is shown in Figure 3. None of the parameters showed any tendency of the mean being dependent on the underlying mean value.
|
|
Day-to-day variation
Calculation of day-to-day variation was based on a comparison between the mean of the 19 double recordings performed by observer 2 at day 1 and at day 2 (Table 4). Mean differences were all close to zero with acceptable variations. An example of one of the constructed Bland–Altman plots for the day-to-day reproducibility is shown in Figure 4. There was no tendency of the mean differences being dependent on the underlying mean values. All parameters except aortic PWV were without any significant day-to-day difference. The intra-class correlation coefficient of aortic PWV was 94% and the rest of the parameters had intra-class correlation coefficients above 87% except brachial PWV, which had 72% (Table 4). The Bland-Altman plot for the brachial PWV is shown to illustrate the presence of an outlier, contributing to the reduced correlation (Figure 5).
|
|
|
Sample size in a clinical study
Based on the variations of the mean differences between recordings performed at day 1 and day 2 (Table 4), the number of subjects needed in a clinical trial on vascular stiffness in CKD stages 3–5 can be calculated using a paired sample t-test. Studied by double recordings, our data indicates that around 10 subjects in each intervention group will give an 80% chance of detecting a difference of 1 m/s in aortic PWV and brachial PWV at a significance level of 5%. Around three subjects per group are needed to give an 80% power of detecting a difference in AIx@HR75 (%) of 10%, at the same significance level. A clinical trial will always include more patients. Thus if each treatment group consists of 20 patients the power of detecting a difference of 1 m/s in aortic and brachial PWV and a difference of 10% in AIx@HR75, at a 5% significance level will be
100%. |
Discussion |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Patients with CKD are known to have an increased arterial stiffness [2] and this might decrease the reproducibility of tonometric pulse-wave recordings. In the present study, however, reproducibility was good. Our studies were done in pre-dialysis CKD patients with a mean GFR of 25 ml/min/1.73 m2 under standardized examination conditions. It was found that intra- and inter-observer as well as the day-to-day reproducibility of PWA and PWV determination were high. The day-to-day variation was comparable with the intra- and inter-observer variation, which is an important finding with respect to the use of the method in long-term clinical trials. Furthermore, the intra- and inter-observer reproducibility was comparable with the reproducibility in our previous study in healthy individuals—even though all indices of arterial stiffness were increased in the present study, indicating an increased arterial stiffness in patients with CKD as expected [8].
The mean differences in the intra-observer study were all minimal although significant differences were found in recordings performed by the two different observers. A difference in aortic diastolic BP of 0.3 mmHg and in HR of 1.4 beats/min is hardly of any clinical relevance, but a difference in aortic PWV of –1.0 m/s could have a clinical impact. We have no explanation of this bias, but it supports our previous findings and recommendation of double-recordings to improve the reproducibility [8].
PWA and PWV had a good inter-observer reproducibility judged by the Bland–Altman plots as well as the test of differences between measurements. These analyses were based on a comparison between the mean of two double recordings, which probably accounts for the better reproducibility.
There are no previous studies of day-to-day reproducibility of indices of PWA and PWV in patients with CKD, but our data on intra- and inter-observer variation are in accordance with previous studies using the SphygmoCor® soft- and hardware. Savage et al. [13] conducted a study on reproducibility of PWA in 188 subjects including 23 healthy controls, 71 patients with CKD (plasma creatinine range 77–1106 µmol/l), 67 dialysis patients and 27 patients with a renal transplant. Good reproducibility was found in AIx, TR, SEVR and aortic mean BP in the intra-observer study in all 188 subjects and in the inter-observer study in 11 healthy controls and 24 patients. Repeated PWA after 2–16 weeks in 31 renal transplanted patients showed good agreement. Determination of PWV was also repeated, but using another equipment [13]. In 10 patients on haemodialysis Covic et al reported good intra-observer reproducibility of AIx based on an analysis of intra-observer errors [12].
The day-to-day variation in PWA and PWV in our study was minimal and comparable with the intra-and inter-observer variation. However a significant bias of –0.7 m/s was found in aortic PWV between measurements performed at day 1 and 2. Blacher et al. [2] showed that an increase in aortic PWV of 1 m/s is associated with a relative risk of all-cause mortality of 1.39, which indicate, that our bias of –0.7 m/s could be of clinical relevance. At the same time a high intraclass correlation coefficient of aortic PWV of 94% is present. The discrepancy between this and the found bias could be due to a high population variability, contributing to a higher intraclass correlation coefficient [15]. All subjects have been examined under standardized conditions and we have no explanation of this bias. But, we do see a change in BP between day 1 and 2 in the same direction as aortic PWV, which could explain the difference in aortic PWV.
In general, the correlation was high for the majority of the parameters with intra-class correlation coefficients being above 87%. Brachial PWV had a slightly lower correlation, which could be partly explained by the obvious outlier seen in Figure 5. The good day-to-day reproducibility is in agreement with the mentioned long-term study of Savage et al. [13], but is contradicted by two other studies in healthy people showing significant day-to-day variation in AIx, augmentation pressure and the ratio of central aortic over peripheral systolic BP in one study [9] and in peripheral and central BP in the other study [10].
Clinical perspectives and conclusion
Increasing attention has been paid to PWA after publication of the recent CAFÉ study in which the method of applanation tonometry was used. A discrepancy between peripheral and central haemodynamics was found with two different antihypertensive treatment regimes despite similar peripheral blood pressure [16]. The results are in agreement with previous studies [17,18] indicating a different impact on central haemodynamics by different antihypertensive drugs. Central blood pressures have been shown to be more useful than brachial blood pressures in predicting cardiovascular structural damage and clinical outcomes [19–21]. This has been suggested as explanation for the differences seen in cardiovascular outcome in several clinical trials despite similar brachial blood pressures [18]. Patients with CKD have, in general, a high risk of CVD. Pre-dialysis patients might have a more beneficial effect from pharmacological intervention than patients receiving dialysis, as the dialysis group is expected to have more progressive CVD. These observations stress the importance of more long-term intervention studies assessing central haemodynamics including patients with CKD and especially pre-dialysis patients.
The variation in our day-to-day study indicated that a very low number of patients with CKD are needed per group in a clinical study over time. Wilkinson et al. [22] has performed a similar analysis of AIx evaluating within-observer and between-observer reproducibility in a mixture of hypertensives, diabetics and healthy controls without any remarks on kidney function. They found that around 25 patients were needed per group to detect a difference in AIx of 10% with a power of 80% and at a significance level of 5%, where we found a need of only around three patients to detect the same difference in AIx@HR75 with double recordings performed. Thus, if each treatment group consists of 20 patients the power of detecting a difference of 10% in AIx@HR75, at a 5% significance level, will be 100%. It is not clear how Wilkinson et al. made their calculations and which variations they are based on, so it is not possible to comment on the differences.
In conclusion, we have found that PWA and PWV based on applanation tonometry using the SphygmoCor® soft- and hardware are highly reproducible in a population of pre-dialysis patients with the day-to-day variation being in agreement with the intra- and inter-observer variation. Only a limited number of CKD patients are required in an intervention study to detect significant clinically relevant differences in the parameters afforded by this method. Thus, applanation tonometry, with the SphygmoCor® system is a simple, non-invasive method to assess central haemodynamics in long-term clinical trials in patients with CKD.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionAcknowledgementsReferences |
|---|
We express our gratitude for the funding of this study by the Danish Society of Nephrology, the Danish Kidney Foundation (Nyreforeningen), Aase Bays Foundation and the Simon Fourner Hartmann Foundation. We also thank laboratory technician Bodil Hellstrøm for her meticulous work during the study and statisticians Tobias Wirenfeldt Klausen and Thomas Scheike for their statistical advice. We acknowledge our gratitude to the patients who participated in this study.
Conflict of interest statement. None declared.
|
References |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionAcknowledgementsReferences |
|---|
- Foley RN, Parfrey PS, Sarnak MJ. Clinical epidemiology of cardiovascular disease in chronic renal disease. Am J Kidney Dis (1998) 32:S112–S119.[Web of Science][Medline]
- Blacher J, Guerin AP, Pannier B, Marchais SJ, Safar ME, London GM. Impact of aortic stiffness on survival in end-stage renal disease. Circulation (1999) 99:2434–2439.
[Abstract/Free Full Text] - London GM, Blacher J, Pannier B, Guerin AP, Marchais SJ, Safar ME. Arterial wave reflections and survival in end-stage renal failure. Hypertension (2001) 38:434–438.
[Abstract/Free Full Text] - Pannier B, Guerin AP, Marchais SJ, Safar ME, London GM. Stiffness of capacitive and conduit arteries: prognostic significance for end-stage renal disease patients. Hypertension (2005) 45:592–596.
[Abstract/Free Full Text] - Nichols WW, O’Rourke MF. McDonald's Blood Flow in Arteries. Theoretical, Experimental and Clinical Principles (1998) London: Arnold.
- O’Rourke MF, Gallagher DE. Pulse wave analysis. J Hypertens Suppl (1996) 14:S147–S157.[Medline]
- Wilkinson IB, Hall IR, MacCallum H, et al. Pulse-wave analysis: clinical evaluation of a noninvasive, widely applicable method for assessing endothelial function. Arterioscler Thromb Vasc Biol (2002) 22:147–152.
[Abstract/Free Full Text] - Frimodt-Moller M, Nielsen AH, Kamper AL, Strandgaard S. Pulse-wave morphology and pulse-wave velocity in healthy human volunteers: examination conditions. Scand J Clin Lab Invest (2006) 66:385–394.[CrossRef][Web of Science][Medline]
- Siebenhofer A, Kemp C, Sutton A, Williams B. The reproducibility of central aortic blood pressure measurements in healthy subjects using applanation tonometry and sphygmocardiography. J Hum Hypertens (1999) 13:625–629.[CrossRef][Web of Science][Medline]
- Filipovsky J, Svobodova V, Pecen L. Reproducibility of radial pulse wave analysis in healthy subjects. J Hypertens (2000) 18:1033–1040.[CrossRef][Web of Science][Medline]
- Ter AE, Holewijn S, Stalenhoef AF, de GJ. Variation in non-invasive measurements of vascular function in healthy volunteers during daytime. Clin Sci (2005) 108:425–431.[CrossRef][Web of Science][Medline]
- Covic A, Goldsmith DJ, Panaghiu L, Covic M, Sedor J. Analysis of the effect of hemodialysis on peripheral and central arterial pressure waveforms. Kidney Int (2000) 57:2634–2643.[CrossRef][Web of Science][Medline]
- Savage MT, Ferro CJ, Pinder SJ, Tomson CR. Reproducibility of derived central arterial waveforms in patients with chronic renal failure. Clin Sci (2002) 103:59–65.[Web of Science][Medline]
- Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet (1986) 1:307–310.[CrossRef][Web of Science][Medline]
- Armitage P, Berry G, Matheews JNS. intraclass correlation. In: Statistical Methods in Medical Research (2002;) Oxford: Blackwell. 704–707.
- Williams B, Lacy PS, Thom SM, et al. 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 (2006) 113:1213–1225.
[Abstract/Free Full Text] - London GM, Asmar RG, O’Rourke MF, Safar ME. Mechanism(s) of selective systolic blood pressure reduction after a low-dose combination of perindopril/indapamide in hypertensive subjects: comparison with atenolol. J Am Coll Cardiol (2004) 43:92–99.
[Abstract/Free Full Text] - Hirata K, Vlachopoulos C, Adji A, O’Rourke MF. Benefits from angiotensin-converting enzyme inhibitor ‘beyond blood pressure lowering’: beyond blood pressure or beyond the brachial artery? J Hypertens (2005) 23:551–556.[Web of Science][Medline]
- Jankowski P, Kawecka-Jaszcz K, Bryniarski L, et al. Fractional diastolic and systolic pressure in the ascending aorta are related to the extent of coronary artery disease. Am J Hypertens (2004) 17:641–646.[CrossRef][Web of Science][Medline]
- Boutouyrie P, Bussy C, Lacolley P, Girerd X, Laloux B, Laurent S. Association between local pulse pressure, mean blood pressure, and large-artery remodeling. Circulation (1999) 100:1387–1393.
[Abstract/Free Full Text] - Safar ME, Blacher J, Pannier B, et al. Central pulse pressure and mortality in end-stage renal disease. Hypertension (2002) 39:735–738.
[Abstract/Free Full Text] - Wilkinson IB, Fuchs SA, Jansen IM, et al. Reproducibility of pulse wave velocity and augmentation index measured by pulse wave analysis. J Hypertens (1998) 16:2079–2084.[CrossRef][Web of Science][Medline]
Accepted in revised form: 21. 6.07
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. E. Sharman, J. Brown, D. J. Holland, G. Macdonald, K. Kostner, and T. H. Marwick Influence of Altered Blood Rheology on Ventricular-Vascular Response to Exercise Hypertension, November 1, 2009; 54(5): 1092 - 1098. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. G. Papaioannou, K. S. Stamatelopoulos, G. Georgiopoulos, C. Vlachopoulos, S. Georgiou, M. Lykka, I. Lambrinoudaki, C. M. Papamichael, and C. I. Stefanadis Arterial Wave Reflections During the Menstrual Cycle of Healthy Women: A Reproducibility Study Hypertension, November 1, 2009; 54(5): 1021 - 1027. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Hornum, P. Clausen, J. Kjaergaard, J. M. Hansen, E. R. Mathiesen, and B. Feldt-Rasmussen Pre-diabetes and arterial stiffness in uraemic patients Nephrol. Dial. Transplant., October 21, 2009; (2009) gfp558v1. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||