Pulse transit time (PTT), usually measured as the time interval from the R-wave peak on an electrocardiogram (ECG) to the arrival of the pulse wave at the periphery, reportedly provides cardiovascular information such as arterial blood vessel stiffness and cardiovascular reactivity.1,2 In addition, many researchers have suggested that PTT may have the potential to be a noninvasive surrogate marker for arterial blood pressure (BP), reflecting transient hemodynamic changes in healthy subjects.1,3
In clinical anesthesia, there has been a long-standing need for a continuous noninvasive measure of BP that can accurately track BP perturbations. Previously, some investigators have found that PTT measurements gave a beat-to-beat indication of changes in BP during obstetric spinal anesthesia and that variations in PTT reflected autonomic responses to tracheal intubation and fluctuations in anesthetic depth.4,5 It is unclear, however, whether these findings are applicable to hypertensive patients with increased arterial stiffness.6,7 We evaluated the correlation between beat-to-beat changes in noninvasive PTT and continuous invasively measured arterial BP during induction of anesthesia in hypertensive patients undergoing kidney transplant surgery.
We retrospectively analyzed electronic medical records stored with data acquisition software (S/5™ Collect, Datex-Ohmeda, Helsinki, Finland) in 23 hypertensive patients with end-stage renal disease who were scheduled for elective kidney transplant surgery under general anesthesia. Radial arterial BP and pulse oximetry monitoring before anesthesia induction of hypertensive kidney transplant recipients had been routine monitoring in our hospital. Patients with cardiac arrhythmias, arteriovenous fistula in both arms, and complications associated with chronic renal failure such as severe anemia (hematocrit < 21%) or hyperkalemia (K+ > 5.0 mEq/L) were excluded. The medical characteristics of these patients are shown in Table 1. Our study protocol was approved by the IRB of Asan Medical Center (Clinical Research Information Service/KCT0000228).
All patients were prepared according to our institution’s standard protocol. No premedication was given, and cardiovascular and antihypertensive medications were continued until the day of surgery, except for the angiotensin-converting enzyme inhibitor or angiotensin II receptor blocker.8 Anesthesia monitoring included measurements of invasive BP, pulse oximetry, ECG, and end-tidal concentration of CO2 using a multiparameter monitor (Datex-Ohmeda S/5™, S/5 Collect, Datex-Ohmeda), all of which were simultaneously recorded throughout the entire procedure. A photoplethysmography (PPG) waveform signal, which measures the change in blood volume,9 was obtained using a fixed-gain pulse oximeter attached to the index finger with a standard finger clamp probe (Ohmeda OxyTip, GE Healthcare, Little Chalfont, United Kingdom). Before the induction of anesthesia, the ipsilateral radial artery was cannulated with a 20-gauge catheter under local anesthesia. The catheter was attached to the stiff fluid-filled pressure tubing and an external pressure transducer (TruWave Disposable Pressure Transducer, Edward Lifescience, Irvine, CA).
Anesthesia was induced with thiopental (4–5 mg/kg), remifentanil (0.5–1 μg/kg), and atracurium (0.6 mg/kg), followed by inhaling of desflurane 6% to 7% in oxygen through a facemask, and then, endotracheal intubation was performed. Beat-to-beat data were continuously recorded with a 300-Hz sampling rate, beginning 1 minute before the induction of anesthesia and continuing until 1 minute after tracheal intubation. The data were converted to an ASCII file for offline analysis using Datex-Ohmeda S/5 Collect and commercially available signal processing software (Windaq, DATAQ Instruments, Akron, OH; and DADiSP, DSP Development, Cambridge, MA). PTT was measured as the time interval between the R-wave peak on the ECG and the maximal upslope of the corresponding PPG waveform (Fig. 1).2 The maximal upslope of the PPG waveform was automatically derived from its first derivatives4,10–13 using the CALC package of the Advanced CODAS analysis software (version 3.25, Windaq, DATAQ Instruments), and the inverse of each PTT value was also calculated to track changes in BP, because they are inversely correlated.4,12,14 Hemodynamic data from each patient were compared at 3 time points: (1) baseline (average of the data recorded 1 minute immediately before anesthesia induction); (2) nadir (lowest BP after the induction of anesthesia, average of 5 values surrounding the nadir); and (3) peak (highest BP after tracheal intubation, average of five values surrounding the peak). Before data analysis, spurious ECG and PPG signals generated by patient movement were concurrently removed. The period between the baseline and the nadir was defined as the “decreasing phase,” and the period from the nadir to the peak was defined as the “increasing phase.”
After analyzing the PTT values from our preliminary study (n = 7), the sample size was determined to detect a projected difference of 10% (23 milliseconds) in PTT before and after pressure decrease with an SD of 30 milliseconds (a type I error of 0.05 and a power of 0.9). It was calculated that 20 patients were required. Expecting a dropout rate, we aimed at enrolling 23 patients. All data are presented as means ± SDs, unless otherwise indicated. A Kolmogorov-Smirnov test with the Lilliefors correction was used to assess normal distribution. Mean differences between the measured variables at 3 time points were analyzed by 1-way repeated measures analysis of variance or multivariate analysis of variance as appropriate. The beat-to-beat correlation between PTT and BP indices throughout the entire induction period, that is, the decreasing phase and the increasing phase, was evaluated in each patient using Pearson correlation coefficients. Fisher Z-transformed Pearson correlation coefficients were compared for between PTT and each BP index using repeated measures analysis of variance.15 To evaluate the difference of correlation between the decreasing and increasing phases, Z-transformed Pearson correlation coefficients of PTT and systolic BP (SBP) were compared using a 2-sided paired t test. The ability of changes in PTT to distinguish a ≥30% BP change at the nadir and the peak was evaluated by a receiver operating characteristic (ROC) curve analysis. SPSS version 12.0 (SPSS Inc, Chicago, IL) and MedCalc version 11.6 (MedCalc Software, Mariakerke, Belgium) were used for statistical analyses. The confidence interval (CI) for area under the curve was calculated using the method described by DeLong et al.16 A P value < 0.05 was considered statistically significant.
After data processing, 14,393 pairs of BP and PTT values were examined. Our data satisfied the assumption of normality using the Kolmogorov-Smirnov test with the Lilliefors correction (P = 0.2), and repeated PTT and the inverse of PTT (1/PTT) data satisfied the Mauchly test of sphericity (P = 0.277 and 0.159, respectively). However, the SBP data did not satisfy the assumption of sphericity (P = 0.046). Representative plots of PTT and SBP from an arbitrarily selected patient are shown in Figure 2. For all patients, the average PTT and SBP at baseline was 248 ± 24 milliseconds (range = 191–287 milliseconds) and 175 ± 27 mm Hg (range = 135–268 mm Hg), respectively, showing poorly controlled hypertension or anxiety without premedication (Table 2). Anesthesia induction resulted in a mean 30.5% decrease in SBP and a mean 17.4% increase in PTT. Conversely, laryngoscopy and tracheal intubation increased mean SBP 24.5% from the nadir, but did not significantly decrease PTT (P = 0.113; Table 2, Fig. 3). During the entire anesthesia induction period showing decreasing and increasing BP, the inverse of PTT demonstrated significantly better correlation with SBP than with the mean arterial BP (MAP; r = 0.81 ± 0.11 vs r = 0.72 ± 0.17; P < 0.001) or diastolic BP (DBP; r = 0.81 ± 0.11 vs r = 0.52 ± 0.24; P < 0.001).
A representative plot that shows the different relationships of PTT with decreasing and increasing BP from a typical patient is illustrated in Figure 4. When we examined the relationship between PTT and SBP during each phase in all patients, we found that the correlation coefficients between 1/PTT and SBP were higher when SBP increased than when SBP decreased (r = 0.83 ± 0.12 vs r = 0.68 ± 0.20; P = 0.001). The difference of the Fisher Z-transformed Pearson correlation coefficients was 0.408 (95% CI, 0.187–0.629).
The ROC curve analysis showed that a 15% change in PTT from the baseline could predict a ≥30% decrease in SBP from the baseline with an area under the ROC curve of 0.85. In contrast, considering a ≥30% increase in SBP, a 15% decrease in PTT predicted a ≥30% increase in SBP with an area under the ROC curve of 0.67. The ROC curves for changes in PTT in relation to ≥30% changes in SBP are shown in Figure 5.
We found that noninvasive beat-to-beat PTT in hypertensive kidney transplantation recipients was better correlated with invasive continuous SBP than with MAP or DBP. Although decreases in SBP during the induction of anesthesia could successfully be detected by tracking simultaneous changes in PTT, increases in SBP after laryngoscopy and tracheal intubation were less closely correlated with PTT. When we examined changes in PTT to indicate the onset of hypotension, a 15% decrease in PTT could predict ≥30% reduction in SBP from baseline.
The induction of anesthesia is a period of great hemodynamic instability in hypertensive patients, and greater intraoperative decreases in BP have been noted in patients with persistent hypertension.17 Therefore, during anesthesia induction in such patients, intermittent noninvasive BP measurements may be too slow and inaccurate to allow early detection and prompt treatment. However, when routine invasive BP monitoring is not clearly indicated, continuous PTT measurement has been demonstrated to be an alternative to provide rapidly available beat-to-beat cardiovascular information.4,5,18 In the present study, we found that beat-to-beat PTT was inversely proportional to continuous SBP: when SBP decreased, PTT lengthened, and vice versa. In addition, the inverse relationship between PTT and SBP was relatively consistent throughout the entire period of anesthesia induction. Specifically, beat-to-beat PTT was better correlated with invasive continuous SBP than with MAP or DBP. This result is similar to the study by Payne et al.,15 in which the PTT (their rPTT) that included the pre-ejection period (PEP) as ours does showed the best linear regression with SBP but was less strongly correlated with MAP and DBP. Likewise, PTT served as a surrogate for beat-to-beat BP in monitoring critically ill infants and children, showing better correlation with SBP (r = 0.73) than with MAP (r = 0.68) and DBP (r = 0.61).19 The possible explanation for this better linear regression with SBP is probably caused by the relationship between SBP, vascular tone, and reflection wave; that is, the decreased vascular tone results in a decrease in the magnitude of the reflection pressure wave.20,21 In this group of hypertensive patients who were studied, we speculate that the reflected pressure wave arriving earlier in the cycle during systole slightly delays ejection, contributing to a longer PEP and PTT (as measured from the ECG). With anesthesia induction and subsequent reduction in vascular tone, the reflection is delayed and arrives later, resulting in earlier ejection and a shorter PEP, so that the actual PTT (excluding PEP) is even longer. Therefore, the good agreement is then because of these offsetting effects that may contribute to the better fit with SBP than with MAP or DBP.
In contrast, it is reported that the relationship between PTT and BP has been inconsistent under certain circumstances such as in patients with chronic heart failure, indicating that disease-induced structural/functional vascular changes influence the transmission of pulse waves,1,18,22 and this relationship became more impaired as the disease progressed.18 Therefore, these studies raised the question as to whether PTT serves as a surrogate for beat-to-beat BP monitoring during the induction of general anesthesia in hypertensive patients with end-stage renal disease, because increased arterial stiffness led to a higher pulse wave velocity in hypertensive renal disease patients.6,7 In the present study, we found that beat-to-beat PTT correlated well with invasive SBP, especially with a reduction in SBP during anesthesia induction, suggesting that the relationship between BP and PTT was not affected by anesthetics administered, although generalized arterial stiffness is the principal feature of the vasculature in hypertensive patients. Our findings are in agreement with the investigation by Sharwood-Smith et al.5 during spinal anesthesia, reporting that the relationship between PTT and BP was similar for the normotensive patients and the patients with pregnancy-induced hypertension and that an increase of 20% in PTT have a sensitivity of 74% and a specificity of 70% in indicating a 10% decrease in MAP.5
When SBP increased after laryngoscopy and tracheal intubation, changes in PTT were not as well correlated with changes in SBP as they were when SBP decreased during induction of anesthesia. Although we do not have a clear explanation of this finding, stressful conditions such as physical exercise may induce alterations in geometric and elastic properties of the vascular tree and systolic and diastolic elastance of the heart, disturbing the relationship between BP and PTT.14 Stimulation from laryngoscopy and tracheal intubation increases sympathetic tone and results in increased myocardial contractility, decreasing PEP. The variation in PTT is postulated to closely follow the variation in PEP.23 Therefore, contractility and vascular tone will be interacting components in terms of their effects on PTT. Furthermore, a change in heart rate (HR) has been shown to be the major confounding factor in the relationship between BP and PTT,15,24 because it affects the timing and aortic pulse wave velocity.25 It has been established that arterial distensibility is a function of HR in anesthetized animals.26Zhang et al.25 demonstrated that HR changes should be considered in the PTT analysis, because HR is a potent predictor of radial pulse wave velocity. Therefore, the discrepancy in PTT-SBP correlation between the decreasing and increasing phases during anesthesia induction may partly be because of significant HR changes after laryngoscopy and tracheal intubation.
There are several methodological issues and limitations of the present study. First, because the QRS is used as the initiation point, altered conduction such as bundle branch blocks or left bundle hemiblocks may also add an additional time component to the PEP and thus alter the relationship between changes in SBP and changes in PTT. In this respect, PTT has an inherent limitation for application to patients with cardiac blocks and arrhythmias. Although Payne et al.15 stated that PTT (their rPTT) should not be used as a marker of purely vascular function because of the significant contribution of PEP to PTT, our results show that PTT does provide some useful information, recognizing the limitation introduced by the inclusion of PEP. However, because the omission of PEP during calculations of PTT results in better correlation with estimated BP,15,27 further studies, using phonocardiography, are warranted to limit the influence of PEP.28 Second, artifacts generated by interference from a PPG signal at a peripheral site or disturbances caused by chest wall movements of ECG recordings are major drawbacks in measuring PTT.1 Motion-related artifacts may partly be prevented by securing the finger and transducer to the supporting cast; however, PTT measurements per se may be influenced by ambient temperature,29 limb position,30 and applied force of the contact sensor.31 Although these factors were strictly controlled in our study, it is unclear whether they would affect the applicability of PTT in various clinical settings. Third, in the ROC curve analysis, further studies with a larger sample size will be needed to determine the best cutoff values with high sensitivity, specificity, and CI.
In summary, we found that beat-to-beat changes in PTT can successfully track changes in SBP and can predict a reduction in SBP during anesthesia induction. Changes in PTT were better correlated with decreasing than with increasing changes in SBP. Despite these limitations, our findings show that PTT could detect instantaneous hemodynamic instability during anesthesia induction, even if the magnitude of the changes in SBP cannot be derived from PTT. These findings suggest that PTT may have potential as a noninvasive index when invasive BP monitoring is not used in hypertensive patients during anesthesia induction.
Name: Sung-Hoon Kim, MD.
Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.
Attestation: Sung-Hoon Kim has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Jun-Gol Song, MD, PhD.
Contribution: This author helped analyze the data and write the manuscript.
Attestation: Jun-Gol Song has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Ji-Hyun Park, MD.
Contribution: This author helped write the manuscript.
Attestation: Ji-Hyun Park has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Jung-Won Kim, MD.
Contribution: This author helped conduct the study.
Attestation: Jung-Won Kim has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Yong-Seok Park, MD.
Contribution: This author helped analyze the data.
Attestation: Yong-Seok Park has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Gyu-Sam Hwang, MD, PhD.
Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript
Attestation: Gyu-Sam Hwang has seen the original study data, reviewed the analysis of the data, approved the final manuscript, and is the author responsible for archiving the study files.
This manuscript was handled by: Dwayne R. Westenskow, PhD.
The authors thank Seon-Ok Kim (Department of Clinical Epidemiology and Biostatistics, Asan Medical Center) for her assistance with the statistical methods.
1. Naschitz JE, Bezobchuk S, Mussafia-Priselac R, Sundick S, Dreyfuss D, Khorshidi I, Karidis A, Manor H, Nagar M, Peck ER, Peck S, Storch S, Rosner I, Gaitini L. Pulse transit time by R-wave-gated infrared photoplethysmography: review of the literature and personal experience. J Clin Monit Comput. 2004;18:333–42
2. Smith RP, Argod J, Pépin JL, Lévy PA. Pulse transit time: an appraisal of potential clinical applications. Thorax. 1999;54:452–7
3. Geddes LA, Voelz MH, Babbs CF, Bourland JD, Tacker WA. Pulse transit time as an indicator of arterial blood pressure. Psychophysiology. 1981;18:71–4
4. Singham S, Voss L, Barnard J, Sleigh J. Nociceptive and anaesthetic-induced changes in pulse transit time during general anaesthesia. Br J Anaesth. 2003;91:662–6
5. 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
6. Wang MC, Tsai WC, Chen JY, Huang JJ. Stepwise increase in arterial stiffness corresponding with the stages of chronic kidney disease. Am J Kidney Dis. 2005;45:494–501
7. Barenbrock M, Spieker C, Laske V, Heidenreich S, Hohage H, Bachmann J, Hoeks AP, Rahn KH. Studies of the vessel wall properties in hemodialysis patients. Kidney Int. 1994;45:1397–400
8. Colson P, Saussine M, Séguin JR, Cuchet D, Chaptal PA, Roquefeuil B. Hemodynamic effects of anesthesia in patients chronically treated with angiotensin-converting enzyme inhibitors. Anesth Analg. 1992;74:805–8
9. Shelley KH. Photoplethysmography: beyond the calculation of arterial oxygen saturation and heart rate. Anesth Analg. 2007;105:31–6
10. Jaryal AK, Selvaraj N, Santhosh J, Anand S, Deepak KK. Monitoring of cardiovascular reactivity to cold stress using digital volume pulse characteristics in health and diabetes. J Clin Monit Comput. 2009;23:123–30
11. Yoon Y, Cho JH, Yoon G. Non-constrained blood pressure monitoring using ECG and PPG for personal healthcare. J Med Syst. 2009;33:261–6
12. Wong MY, Poon CC, Zhang YT. An evaluation of the cuffless blood pressure estimation based on pulse transit time technique: a half year study on normotensive subjects. Cardiovasc Eng. 2009;9:32–8
13. Foo JY, Lim CS. Pulse transit time as an indirect marker for variations in cardiovascular related reactivity. Technol Health Care. 2006;14:97–108
14. Wong MY, Pickwell-MacPherson E, Zhang YT. The acute effects of running on blood pressure estimation using pulse transit time in normotensive subjects. Eur J Appl Physiol. 2009;107:169–75
15. 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
16. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics. 1988;44:837–45
17. Prys-Roberts C, Greene LT, Meloche R, Foëx P. Studies of anaesthesia in relation to hypertension. II. Haemodynamic consequences of induction and endotracheal intubation. Br J Anaesth. 1971;43:531–47
18. Wagner DR, Roesch N, Harpes P, Körtke H, Plumer P, Saberin A, Chakoutio V, Oundjede D, Delagardelle C, Beissel J, Gilson G, Kindermann I, Böhm M. Relationship between pulse transit time and blood pressure is impaired in patients with chronic heart failure. Clin Res Cardiol. 2010;99:657–64
19. Wippermann CF, Schranz D, Huth RG. Evaluation of the pulse wave arrival time as a marker for blood pressure changes in critically ill infants and children. J Clin Monit. 1995;11:324–8
20. Westerhof BE, Guelen I, Westerhof N, Karemaker JM, Avolio A. Quantification of wave reflection in the human aorta from pressure alone: a proof of principle. Hypertension. 2006;48:595–601
21. Kips JG, Rietzschel ER, De Buyzere ML, Westerhof BE, Gillebert TC, Van Bortel LM, Segers P. Evaluation of noninvasive methods to assess wave reflection and pulse transit time from the pressure waveform alone. Hypertension. 2009;53:142–9
22. Masè M, Mattei W, Cucino R, Faes L, Nollo G. Feasibility of cuff-free measurement of systolic and diastolic arterial blood pressure. J Electrocardiol. 2011;44:201–7
23. Middleton PM, Chan GS, O’Lone E, Steel E, Carroll R, Celler BG, Lovell NH. Changes in left ventricular ejection time and pulse transit time derived from finger photoplethysmogram and electrocardiogram during moderate haemorrhage. Clin Physiol Funct Imaging. 2009;29:163–9
24. Lantelme P, Mestre C, Lievre M, Gressard A, Milon H. Heart rate: an important confounder of pulse wave velocity assessment. Hypertension. 2002;39:1083–7
25. Zhang YL, Zheng YY, Ma ZC, Sun YN. Radial pulse transit time is an index of arterial stiffness. Hypertens Res. 2011;34:884–7
26. Mangoni AA, Mircoli L, Giannattasio C, Ferrari AU, Mancia G. Heart rate-dependence of arterial distensibility in vivo. J Hypertens. 1996;14:897–901
27. Allen RA, Schneider JA, Davidson DM, Winchester MA, Taylor CB. The covariation of blood pressure and pulse transit time in hypertensive patients. Psychophysiology. 1981;18:301–6
28. Newlin DB. Relationships of pulse transmission times to pre-ejection period and blood pressure. Psychophysiology. 1981;18:316–21
29. Zhang XY, Zhang YT. The effect of local mild cold exposure on pulse transit time. Physiol Meas. 2006;27:649–60
30. Foo JY, Wilson SJ, Williams GR, Harris MA, Cooper DM. Pulse transit time changes observed with different limb positions. Physiol Meas. 2005;26:1093–102
31. Teng XF, Zhang YT. The effect of applied sensor contact force on pulse transit time. Physiol Meas. 2006;27:675–84