If you wish to respond to a paper or other item already published in the BJA, please go to the abstract/full text version of that item and click on the link "E-Letters: Submit a response to the article".
Electronic Letters to:
|
|
E-letters: Electronic letters published:
-
![[Read E-letter]](/icons/misc/sectionDOWN.gif)
Re: Response to Payne, Webb and Maxwell: an indicator, yes; a measure, no.
- Rupert A Payne, David J. Webb, Simon R. Maxwell. (7 March 2006)
-
Response to Payne, Webb and Maxwell: an indicator, yes; a measure, no.
- Gordon B Drummond, Jean Bruce and Geoff Sharwood-Smith, (19 February 2006)
-
Assessment of pulse transit time to indicate cardiovascular changes during obstetric spinal anaesthe
- Rupert A Payne, David J. Webb, Simon R. Maxwell. (13 February 2006)
-
Re: Response to Assessment of pulse transit time to indicate cardiovascular changes during obstetric
- Gordon B Drummond, Geoffrey Sharwood-Smith and Jean Bruce (10 February 2006)
-
Response to Assessment of pulse transit time to indicate cardiovascular changes during obstetric spi
- Andrew RM Conway Morris (3 February 2006)
|
|
|||
|
Rupert A Payne, Clinical Lecturer Clinical Pharmacology Unit, Centre for Cardiovascular Science, The University of Edinburgh, David J. Webb, Simon R. Maxwell.
Send letter to journal:
|
We would like to respond to a few points raised by Drummond and colleagues in their reponse to our previous comments on the topic of transit time measurement as a marker of cardiovascular function. The accuracy of bioimpedance as a measure of cardiac output has been validated, but may well be subject to significant inaccuracy with drug administration. However, the use of bioimpedance as a measure of pre- ejection period is largely dependent on the timing of increased blood volume in the aorta, a direct consequence of mechanical ventricular ejection. Although the absolute impedance changes and the subsequent contour of the bioimpedance waveform may be prone to changes depending on other factors (electrode placement, pulmonary blood flow, etc.), thus affecting the accuracy of cardiac output assessment, the initial upstroke or "B-point" used to determine the pre-ejection period is relatively independent of such factors. Accuracy of bioimpedance measurement of systolic time intervals has been validated against invasive methods [1]. The sensitivity of the technique has indeed been questioned relative to other methods [2], although technological advances in signal processing may address this issue. Nonetheless, we feel that the pre-ejection period should be measured and reported whenever possible in studies employing transit time, and that bioimpedance provides a reasonably accurate means of achieving this. With respect to the site of measurement of blood pressure, it is well recognised that important differences exist between the brachial and radial artery sites [3,4]. However, pressure measurements at the two sites do tend to be closely correlated. Furthermore, mean pressure is relatively constant between these two sites (in contrast to systolic pressure which is prone to pulse amplification), and it is mean rather than systolic pressure that determines pulse wave velocity. Indeed, we have noted that changes in overall transit time tend to be accounted for by similar changes in the individual brachial and radial pulse wave velocities (unpublished observations). We did not measure the time relationship between arterial pressure and photoplethysmograph waveforms, as the former was subject to a significant delay introduced by the use of a fluid filled catheter, and the constancy of the delay was difficult to ascertain. We agree with Drummond and colleagues that the method used to measure the waveforms may result in significant timing inaccuracies. It is important to ensure that the system has a relatively flat frequency response to at least 25Hz, and that any AC coupling has a long time constant to avoid signal distortion. The analogue signal processing will also introduce a marked delay between physiological signal and electronic output waveform (around 20 to 70ms, depending on equipment), and it is essential to take this into account in timing measurements. Indeed it is worthwhile noting that ECG devices may also introduce a small delay, albeit usually around 5-10ms. We took great care to ensure the fidelity of the photoplethymograph waveform was adequate, and measured the signal delay using a simulated pulse wave. We sampled the waveform at 200Hz 12bit resolution, and interpolated the signal to obtain a timing resolution of 1kHz. We agree that determining the true foot of the wave in true real-time is difficult and indeed unreliable. However, the intersecting tangent time-point can be determined accurately as soon as the maximal slope is detected, and an accurate beat- to-beat measurement can still be reported during the same cardiac cycle. Reference List (1) Labibidi Z, Ehmke DA, Durnin RE et al. The first derivative thoracic impedance cardiogram. Circulation 1970; 41: 651-8 (2) Belz GG. Systolic time intervals: a method to assess cardiovascular drug effects in humans. Eur J Clin Invest 1995; 25(S1): 35- 41 (3) Verbeke F, Segers P, Heireman S et al. Noninvasive assessment of local pulse pressure: importance of brachial-to-radial pressure amplification. Hypertension 2005; 46: 244-8 (4) Bazaral MG, Welch M, Golding LA, Badhwar K. Comparison of brachial and radial arterial pressure monitoring in patients undergoing coronary artery bypass surgery. Anesthesiology 1990; 73: 38-45 Conflict of Interest:None declared |
|||
|
|
|||
|
Gordon B Drummond, Senior Lecturer Edinburgh University, Jean Bruce and Geoff Sharwood-Smith,
Send letter to journal:
|
Payne, Webb and Maxwell write to draw attention to their own findings in a carefully performed laboratory study of young men with radial arterial cannulae placed to measure arterial pressure. They caution against the “use of PTT to estimate systolic pressure” This is something of a misrepresentation of the main thrust of our article, as we carefully worded the title ‘an assessment of pulse transit time to indicate cardiovascular changes during obstetric spinal anaesthesia’. Despite this, we agree with several of the points they make, although we feel they exaggerate some others. In their study they infused norepinephrine, angiotensin II, glyceryl trinitrate, and salbutamol intravenously and measured the time interval between the R wave and the onset of the finger plethysmograph wave. They also used thoracic impedance measurements to estimate the time of cardiac ejection, which allowed an estimate of the time interval for pressure wave transmission without the delay between cardiac electrical activation (indicated by the R wave of the ECG) and the end of isovolumic cardiac contraction, which precedes cardiac ejection. In most of the drug treatments that they used, this estimate of the pre-ejection period time varied very little. However when salbutamol was given, marked tachycardia resulted, (from 66 to 125 beat/min), and the estimate of pre-ejection time was reduced greatly from 91 to 30 mS. Such changes are consistent with previous reports.[1;2] However they are not frequent in the context we present. We also suspect that pre-ejection periods of 130 mS are most unusual in the patients we studied, and thus the contribution of this time to the transit time we measured is likely to be substantially less than the values they estimate. Many of the measurements in this field have to be inferential, using non-invasive measures such as impedance. Although impedance changes during the cardiac cycle have been studied extensively, and often found to relate well to other measures of cardiac function, the exact contribution of blood distribution in lungs, major vessels, and heart to the global impedance changes remain to be directly demonstrated.[3] Indeed with the use of different agents to change cardiac output, the inherent weakness of impedance measurements to measure cardiac output has become apparent.[4] This weakness may also extend to measures of ejection: there can be differences between pressure (responsible for the PTT) and flow or volume change which are measured, by inference, from impedance changes.[5;6] Such difficulties underline the value of testing these measures in a variety of circumstances. Payne, Webb and Maxwell found that salbutamol, with the associated tachycardia, generated a completely different relationship between pulse transit time and blood pressure and concluded that blood pressure could only be inexactly estimated from the transit times. We would certainly agree that pooling data obtained using different agents will weaken the ability to predict blood pressure but reiterate and emphasise that we did not suggest that blood pressure be actually measured in this way. We showed that a change in PTT was a robust indication that changes in blood pressure were occurring, and suggested that this would be a much more prompt indication than non-invasive sphygmomanometry. Since the changes in heart rate in our study were much less than those caused by salbutamol infusion, we also contend that our guarded inferences on the elastic properties of the conducting vessels, made in the discussion, are justified, particularly since we were careful to conclude “we found no evidence that the properties of large arteries are affected by pregnancy- induced hypertension”. Finally we should remind readers that the pressure we measured, indirectly, was brachial artery pressure. There are differences between brachial and radial artery pressure, and these are also affected by vasoactive drugs. Indeed, it’s likely that the pressure in the brachial artery may be more representative of the “pressure” that affects pulse transit time than does radial artery pressure. After all, it is measured at the middle of the vessel, and not at one end! It would be of great interest to us to know the time relationship between the arterial pressure and the photoplethysmograph signal. We have found that commonly used clinical monitors introduce a time delay in the latter measurement. In addition, the effect of the vasoactive agents on this relationship would be of considerable scientific interest, assuming that the fidelity of the plethysmograph signal was sufficient. Payne and co-workers [7] give no details of the methods they used to measure and record the plethysmograph. We agree that the determination of the foot of the plethysmograph wave is a more consistent measure of the timing of this wave, but this is not a measure that can be made easily in real time, whereas the position on the wave that we measured could be easily computed with the analogue device we used. Reference List (1) Talley RC, Meyer JF, Mcnay JL. Evaluation of Pre-Ejection Period As An Estimate of Myocardial Contractility in Dogs. American Journal of Cardiology 1971; 27: 384-&. (2) Harris WS, Schoenfe CD, Weissler AM. Effects of Adrenergic Receptor Activation and Blockade on Systolic Preejection Period Heart Rate and Arterial Pressure in Man. J Clin Invest 1967; 46: 1704-&. (3) Wang YQ, Haynor DR, Kim Y. A finite-element study of the effects of electrode position on the measured impedance change in impedance cardiography. Ieee Transactions on Biomedical Engineering 2001; 48: 1390- 1401. (4) Critchley LAH, Peng ZY, Fok BS, James AE. The effect of peripheral resistance on impedance cardiography measurements in the anesthetized dog. Anesthesia and Analgesia 2005; 100: 1708-1712. (5) Peterson LH. The dynamics of pulsatile blood flow. Circ Res 1954; 2: 127-139. (6) Zamir M. Wave reflections. The Physics of Pulsatile Flow. New York: Springer-Verlag; 2000. 147-185. (7) Payne RA, Symeonides CN, Webb DJ, Maxwell SRJ. Pulse transit time measured from the ECG: an unreliable marker of beat-to-beat blood pressure. J Appl Physiol 2006; 100: 136-141. Conflict of Interest:We are the authors of the paper |
|||
|
|
|||
|
Rupert A Payne, Clinical Lecturer Clinical Pharmacology Unit, Centre for Cardiovascular Science, The University of Edinburgh, David J. Webb, Simon R. Maxwell.
Send letter to journal:
|
In their paper, Sharwood-Smith and co-workers(1) describe the attraction of using pulse transit time (PTT) as an indicator of changes in beat-to-beat blood pressure during anaesthesia. This is a simple, non- invasive technique, using ECG and photoplethysmograph signals which are both monitored routinely in clinical practice. There are important aspects to this technique, however, which may be of relevance when applying it more widely in the clinical setting. The authors comment on the intra-cardiac time delay (pre-ejection period, PEP) between ECG R-wave and mechanical ventricular ejection. PEP can vary markedly with vasoactive drugs and heart rate(2,3). In young healthy individuals it may be as long as 130ms, and can contribute anywhere between 12% and 35% of the PTT measured to the finger. Although the vascular path length and thus transit time to the toe is greater, the PEP may still account for over 30% of the overall toe PTT. Indeed, the PTT values reported by the authors are probably increased by a duration of around 50ms relatively independently of PEP, due to measurement at the point of photoplethysmogram maximal slope as opposed to the wave nadir, underplaying the contribution of PEP still further. This cardiac delay makes it difficult to draw firm conclusions about large or conduit artery properties from this study. The authors found a significant correlation between MAP and toe PTT in the population as a whole, but it is the strength of the correlation in a single individual that determines the technique’s ability to predict changes in blood pressure. Although the authors reported the sensitivity and specificity of the technique to be around 70%, the correlation slope between directly measured systolic arterial pressure and the PTT measured from R-wave to finger can be affected by vasoactive drugs(3,4). Furthermore, mean arterial pressure has a far weaker correlation than systolic pressure with R-wave to finger PTT, although the correlation between mean pressure and true vascular transit time (PTT minus PEP) is better. This is of particular importance in an anaesthetic setting where rapid, pharmacologically induced changes in blood pressure may occur, and the wide range of drugs employed by the anaesthetist may have differing effects on the PTT-pressure relationship. Use of PTT to estimate systolic pressure may thus be unreliable, and estimation of mean arterial pressure even more so without prior knowledge of PEP. We agree with the authors that this method may have a role in predicting rapid shifts in blood pressure, which are not detectable by sphygmomanometry, but emphasise that the relationship between PTT and arterial pressure is not fixed, and use of this method as a reliable alternative to direct arterial pressure monitoring is inadvisable. It is also inappropriate to use PTT measured from the ECG as a measure of arterial stiffness. 1. Sharwood-Smith G, Bruce J, Drummond G. Assessment of pulse transit time to indicate cardiovascular changes during obstetric spinal anaesthesia. Br J Anaesth 2006; 96: 100-5 2. Belz GG. Systolic time intervals: a method to assess cardiovascular drug effects in humans. Eur J Clin Invest 1995; 25 (S1): 35 -41. 3. Payne RA, Symeonides CN, Webb DJ, Maxwell SR. Pulse transit time measured from the ECG: an unreliable marker of beat-to-beat blood pressure. J Appl Physiol 2006; 100: 136-41 4. Steptoe A, Smulyan H, Gribbin B. Pulse wave velocity and blood pressure change: calibration and applications. Psychophysiology 1976; 13: 488-93 Conflict of Interest:None declared |
|||
|
|
|||
|
Gordon B Drummond, Senior Lecturer Department of Anaesthesia, Critical Care and Pain Medicine, Geoffrey Sharwood-Smith and Jean Bruce
Send letter to journal:
|
We did not set out to estimate blood pressure using pulse transit time. As stated clearly in the abstract and the introduction, we studied the relationship between pulse transit time and non invasive blood pressure, to obtain information that might be used clinically as an early warning of possible blood pressure changes. Our prime aim was to observe PTT and assess its value to indicate changes in blood pressure. Our hypothesis is that the information would be available before the non- invasive device gave a reading. This hypothesis remains to be tested: what we did find was that the measurements correlated, and that the PTT was reasonably good at predicting a decrease in non-invasive blood pressure. We made no claims to be able to estimate the blood pressure on the basis of the pulse transit time. What is a “gold standard” when it comes to blood pressure, anyway? Pressures vary within the vascular system, with time and site. Pressure waveforms in an artery at the wrist are not the same as in the brachial artery or the aorta. As we discussed, the pulse transit time is a measure, inter alia, of a compound of vascular elastance, from heart to finger. Physiologically it makes little sense to use one of these measures to actually estimate the other: but a lot of sense to correlate them, as we did. Indeed, one could argue that clinicians, used to using NIBP regularly to manage routine anaesthesia, would rather have a device that told them what the next NIBP value would be, than what the infrequently used radial arterial pressure might be. Bland and Altman’s papers are some of the most highly cited in the medical literature, and sadly their method is one of the most mis-used. Very frequently, a cloud of spots representing several measures taken from a number of subjects is presented, confusing and obscuring inter- and intra- individual variation. These authors have stated very cogently the reasons for not using a “gold standard”.(ref 1) We were naturally aware of their method but baulked at the comparison of time with blood pressure. Our use of correlation was essentially illustrative. We do have data from a similar individual (spinal anaesthesia for caesarian section) which show simultaneous measures of pulse transit time and continuously measured radial arterial pressure (by applanation tonometry) which also illustrate our contention that the relationship(correlation) between these measures is excellent. However no measure, in this field, is necessarily “the best”. Horses for courses! Reference List (1) Bland JM, Altman DG. Comparing Methods of Measurement - Why Plotting Difference Against Standard Method Is Misleading. Lancet 1995; 346(8982):1085-1087. Conflict of Interest:None declared |
|||
|
|
|||
|
Andrew RM Conway Morris, SHO Anaesthesia and Intensive Care Royal Infirmary of Edinburgh
Send letter to journal:
|
Editor The quest for a non-invasive measure of beat-by-beat blood pressure continues. Sharwood-Smith et al’s description of the measurement pulse transit time (PTT)(1) is an interesting contribution to this area. However their paper has two key flaws which reduce its usefulness. The first was to compare PTT to blood pressure estimated by non-invasive, oscillometric devices. Although this is the method in commonest clinical use it is not the gold standard and has been shown to be inaccurate, particularly at low blood pressures (2,3). For a monitoring device to be accepted it should be shown to compare favourably with the existing gold standard, i.e. invasive blood pressure measurement by arterial cannulation. The second issue I have is with the statistical tests used to compare the measures. Bland and Altman have clearly demonstrated that simply showing correlation is not sufficient to demonstrate equivalence between two tests(4). The correct method of analysis would have been to produce a Bland-Altman plot. Although the authors have shown a degree of utility in predicting a reduction in blood pressure, the statistical analysis presented does not allow us to draw wider implications of PTT as a non- invasive, beat by beat measure of blood pressure. 1 Sharwood-Smith G, Bruce J and Drummond G. Assessment of pulse transit time to indicate cardiovascular changes during obstetric spinal anaesthesia. Br J Anaesth 2006; 96: 100-5 2 Gourdeau M, Martin R, Lamarche Y and Tetreault L. Oscillometry and direct blood pressure: a comparative clinical study during deliberate hypotension. Canadian Anaesthetists' Society Journal 1986 33:300-7 3 Bur A, Hirschl MM, Herkner H et al. Accuracy of oscillometric blood pressure measurement according to the relation between cuff size and upper -arm circumference in critically ill patients Critical Care Medicine. 2000 28:371-6 4 Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical assessment. Lancet 1986; 1 (8476):307-310. Conflict of Interest:None declared |
|||