Introduction: Pulse transit time (PuTT) is the interval between R-wave on electrocardiogram (ECG) and upstroke on peripheral arterial blood pressure (BP) waveform or pulse plethysmographic waveform. It has been suggested that changes in PuTT reflect changes in arterial BP and cardiac output (CO). The study tested the hypothesis that changes in PuTT reflect CO changes in anesthetized ventilated surgical patients. Materials and Methods: Surgical patients monitored with direct BP and non-invasive cardiac outputs were included in study. Patients with hypotension and normal CO were treated by phenylephrine (phenylephrine group, n= 41). Patients who required volume loading received bolus of colloid solution (volume loading group, n=22). Arterial and plethysmographic PuTT changes were compared with stroke volume index (SVI) and BP changes. Results: Although systolic BP increased in both groups, it increased to a greater extent following phenylephrine therapy. SVI increased following volume loading but did not change after phenylephrine administration. Arterial and plethysmographic PuTTs shortened in both groups following treatment. There was a weak but significant correlation between arterial PuTT changes and systolic BP changes (β = – 0.314; P = 0.006) and SVI (β = – 0.289; P = 0.01) as well as between plethysmographic PuTT changes and systolic BP changes (β = -0.374; P = 0.001). Correlation between changes of plethysmographic PuTT and SVI was non-significant. Conclusion: Low correlation coefficients demonstrate that PuTT changes do not correlate closely with changes either in BP or in CO. As such, we therefore conclude that PuTT changes do not reflect changes in CO or BP, and PuTT is not applicable for estimation of CO or BP changes in anesthetized ventilated patients.
Cardiac Output (CO) is an important hemodynamic variable that is rarely measured despite the fact it has been suggested that guiding perioperative therapy according to CO may improve outcome. Despite its perceived importance, CO is not measured routinely because of the complexity and the invasiveness it entails. Less invasive as well as noninvasive methods to measure cardiac output are available but they are less reliable and robust than invasive measurements. Among these techniques the thoracic bioreactance was evaluated intensively. In a multicentre study the non-invasive cardiac output monitor (NICOM) was assessed in mixed population of patients in cardiac care units, ICUs, and cardiac catheterization laboratories in comparison with pulmonary artery catheter derived cardiac output (either continuous cardiac output measurements or intermittent bolus pulmonary artery thermodilution measurements), a low bias has been observed. Comparable results were obtained in studies in post cardiac surgery patients when using cardiac output measurements obtained with transpulmonary thermodilution and calibrated pulse contour analysis or with pulmonary artery thermodilution as the criterion standard.
Pulse transit time (PuTT) refers to the time it takes a pulse wave to travel between two arterial sites. It is measured as the interval between the R-wave on an electrocardiogram (ECG) and the upstroke on the peripheral arterial blood pressure (BP) curve or on the plethysmographic curve (Figure 1). Several studies have suggested that changes in PuTT reflect changes in CO and systolic arterial BP [8-12]. However, CO was not measured in any of these studies, and changes in CO were assumed to result from medication administration or body position changes. Moreover, correlation of PuTT and CO was evaluated only in spontaneously breathing conscious volunteers and was not evaluated in the perioperative setting with anesthetized mechanically ventilated patients. This study tested the hypothesis that in anesthetized and ventilated surgical patients, changes in PuTT reflect changes in CO, rather than changes in BP. We therefore designed this study to create two exclusive clinical conditions associated either with increases in CO without changes in BP by volume loading, or alternatively increased BP without changes in CO by the administration of a vasoconstrictor (phenylephrine).
Materials and Methods
The study was approved by institutional review board (Helsinki Committee of The Lady Davis Carmel Medical Center, Haifa, Israel; protocol number CMC-09-0124; ref: 920040086) and registered in ClinicalTrial.gov (registration number: NCT 01259687; ID: CMC-09-0124-CTIL).
Patients eligible for this study were those scheduled for elective surgery under general anesthesia who required continuous direct BP monitoring based on the type of surgery and/or their medical condition. After giving signed informed consent, all study participants received standard anesthesia care.
On the morning of surgery, patients received 10 mg diazepam orally. Those who were being treated chronically with β-adrenergic antagonists received regular doses of this medication. Angiotensin converting enzyme inhibitors were not given on the day of surgery. All patients were anesthetized intravenously with fentanyl 2-5 mcg/kg, propofol 1-2 mg/kg and rocuronium 0.6-1.0 mg/kg followed by endotracheal intubation. After intubation, positive pressure ventilation was applied with a tidal volume 8 mL/kg of ideal body weight and a respiratory rate of 8-12 bpm to achieve an end-tidal carbon dioxide level of 33-37 mmHg. Anesthesia was maintained with inhalation of isoflurane (end-expiratory concentration of at least 1.2%) in air/O2 (FiO2 =0.4) and additional intermittent boluses of opiates and neuromuscular blocking agents. Intraoperative fluid management included an initial 7 mL/kg bolus of crystalloid solution (lactated Ringer’s solution) followed by continuous infusion as needed. Arterial BP was measured continuously through an intraarterial 20 G catheter in a radial artery and monitored on a Datex-Ohmeda AS-3 monitor (Datex, Helsinki, Finland). Pulse oximetry photoplethysmography was monitored from an index finger probe on the same side. CO was measured by a non-invasive cardiac output monitor (NICOM, Cheetah Reliant, Cheetah Medical (UK) Limited, Maidenhead, Berkshire, UK) which in addition to providing CO/cardiac index and stroke volume/ stroke volume index (SVI) also calculates stroke volume variation (SVV).
Patients who developed hypotension with systolic BP <80 mmHg (<90 mmHg in those with documented hypertensive disease)  and had normal CO (cardiac index ≥2.5 L/ min*m2) were treated by boluses of phenylephrine 50-100 mcg intravenously until the systolic BP increased by at least 20% from baseline (the phenylephrine group). The assumption was that the systolic BP of these patients would increase without significant changes in cardiac index/stroke volume index (SVI) . Patients who required volume replacement based on clinical judgment received a bolus of colloid solution (6% hydroxyethyl starch 200/0.5; Fresenius Kabi, Deutschland GmbH, Germany) at a dose of 7 ml/kg over 15 minutes (the volume loading group).ECG, arterial pressure and pulse oximetric plethysmographic waveforms were simultaneously recorded for each study event before and after the intervention using a recording module of the Datex-Ohmeda AS3 monitor. Values of CO/ cardiac index, SVI and SVV as measured by the NICOM were recorded before and after volume loading, or after systolic BP had reached the desired level.
Systolic and diastolic BP, heart rate, and arterial and plethysmographic PuTT were measured offline from the records. PuTT was measured as the interval between the R wave on the ECG and the beginning of the upstroke on the arterial pressure or the plethysmographic waveform (Figure 1).
Figure 1. Pulse transit time (PuTT) measurements.
A representative sample of the record was scanned and pasted onto a Microsoft Office Word document. Parallel lines have been drawn through the peak of R-wave of ECG and the PuTT points on the arterial pressure and the plethysmographic waveforms. The connecting lines designate the PuTT interval. Microsoft Office Word measures the length of the line when it chosen. The length of the line designated 200 milliseconds (according to the grid of the monitor record) is measured by the same manner. Each record was scanned at a resolution of 300 dpi to the JPEG format picture and then pasted onto a Microsoft Office Word document (Office 2007; Microsoft Co., Redmond, WA). The lines designating the PuTT intervals as well as the 200 msec lengths (according to the grid of the monitor record) were drawn over the curves and their lengths were measured in millimeters using a format shape function of Microsoft Office Word (Figure 1).
PuTT values were calculated according to the following equation: PuTT (msec) = “PuTT” (mm) x 200 / “200 msec” (mm)
The continuous variables are presented as means, standard deviations and medians. The categorical variables are presented as percentages. Comparisons between the two groups (phenylephrine vs. volume loading) were analyzed using the Chi-square test for the categorical variable, and the independent t-test or the Mann-Whitney test as appropriate. Comparisons within the same group were analyzed using the paired t-test or the Wilcoxon ranked test as appropriate for the continuous variables. Correlations between the percent changes of variables were analyzed using the Pearson correlation. A multivariate linear regression was used to check which clinical variables were independently correlated to PuTT. Differences in PuTT values related to medical history and chronic medications were assessed using independent t-test. Coefficient of variation was calculated for the main variables. When assuming a power of 80%, an alpha of 0.05 and a Pearson correlation of 0.3, 60 independent measurements are needed to achieve a statistically significant correlation. P values < 0.05 were considered significant. Statistical analysis was performed using PASW 18 software.
Forty one measurements in 27 patients were included in the phenylephrine group and 22 measurements in 22 patients were included in the volume loading group; patient demographics are listed in Table 1. There were no differences in demographic data and medical history between the two groups (table 1). The analysis of the effects of chronic medications (β-adrenergic antagonists, ACE-inhibitors and Ca++-channel blockers) and medical history (ischemic heart disease, hypertension disease, chronic obstructive pulmonary disease and diabetes mellitus) on arterial and plethysmographic PuTT changes showed low correlation, and a lack of statistical significance except in the correlation between plethysmographic PuTT changes and diabetes mellitus (P=0.03) and chronic treatment by ACE-inhibitors (P=0.04) (table 1). Hemodynamic variables in the patients of each group are presented in Table 2. Although systolic BP increased significantly in both groups, the increase was less after volume loading than after phenylephrine therapy. The SVI increased significantly after volume loading but did not change after treatment with phenylephrine.
Both arterial and plethysmographic PuTT measurements were shorter in both groups following treatment. PuTT was corrected for the RR interval. The relationship between PuTT changes and changes in systolic BP and SVI was analyzed for all patients in both groups. There was a weak but nevertheless significant correlation between arterial PuTT changes and systolic BP changes (β = -0.314; P = 0.006; 95% CI = 0.304 to -0.054) (Figure 2) and between arterial PuTT and SVI (β = -0.289; P = 0.01; 95% CI = -0.335 to -0.045) (Figure 3). There was also a weak but significant correlation between plethysmographic PuTT changes and systolic BP changes (β = -0.374; P = 0.001; 95% CI = -0.327 to -0.092) (Figure 4). No significant correlation was found between changes in plethysmographic PuTT and SVI (Figure 5). To estimate the consistency of the arterial and the plethysmographic PuTTs the coefficients of variation were calculated.
Table 1. Demographics of the phenylephrine and the volume loading groups
Table 2. Hemodynamic variables in patients treated with phenylephrine and patients treated with volume loading
Figure 2. Correlation between changes in arterial pulse transit time (Art.PuTT) and systolic blood pressure (SBP).
Figure 3. Correlation between changes in arterial pulse transit time (Art.PuTT) and stroke volume index (SVI).
Figure 4. Correlation between changes in plethysmographic pulse transit time (Pleth.PuTT) and systolic blood pressure (SBP).
Figure 5. Correlation between changes in plethysmographic pulse transit time (Pleth.PuTT) and stroke volume index (SVI).
These coefficients of variation were low and similar in range to the coefficient of variation of systolic BP and of HR (table 2).
Cardiac output is one of important variables that reflect the patient’s condition during the perioperative period and its optimization has been shown to improve outcome [1,2]. Most clinical methods of CO measurement are invasive or not accurate enough or user dependent. It possible, however, that directions and degrees of change in CO (rather than absolute values) can provide us with clinically useful information . There is therefore a need to develop new methods of CO measurement or variables that reflect changes in CO accurately. Our hypothesis was that changes in PuTT reflect changes in CO in anesthetized and mechanically ventilated patients. The results of this study led us to reject this hypothesis. The study was adequately powered and its findings revealed a weak correlation between changes in PuTT (regardless of whether analyzed from arterial or plethysmographic waveforms) and changes in SVI or BP.
Taking into consideration earlier studies which found BP to have an impact of on PuTT [11,12,15], we studied PuTT changes occurring in clinically relevant intraoperative events with the aim of differentiating, as much as possible, between influences of systolic BP on PuTT and those of CO. After phenylephrine therapy an increase in SBP occurred without any concomitant changes in SVI (table 2). By contrast, volume loading resulted in a mild increase in systolic BP and a significant increase in SVI (table 2). The two interventions thus resulted in a significantly large and distinctive enough increment in the two physiological parameters, systolic BP and SVI, to allow an analysis of the effect of each of these on PuTT. The PuTT values during the study were consistent and the coefficients of variation were low (table 2). Whereas previous studies used absolute values of PuTT, we calculated PuTT as fraction (%) of RR interval to minimize the effect of changes in HR. For similar reasons, we analyzed SVI rather than cardiac index. Previous studies were performed on non-anaesthetized and spontaneously breathing subjects or nonsurgical critically ill patients, while our study was performed on patients who were anesthetized and mechanically ventilated. Our groups of patients represent clinically relevant situations in which systolic BP increases without changes in CO, or CO increases with only very mild changes in systolic BP.
Associations between PuTT and systolic blood pressure
Earlier studies [11,12] have found that PuTT does not reflect absolute values of BP but rather only directional changes: shorter PuTT was associated with an increase in systolic BP while longer PuTT was associated with a decrease of systolic BP. In the present study, overall we observed inconsistent shortening of PuTT during increases in systolic BP; however, there were several patients in whom PuTT increased (Figure 2). The effect of halogenated inhaled agents on the left ventricular – arterial coupling (all patients in our study received isoflurane) is a possible explanation for the inconsistency observed in our anesthetized patients as compared with intact subjects of previous studies [16,17].
Associations between PuTT and cardiac output
Previous studies have described a relationship of PuTT and CO based on the assumption of CO response, even though CO was not actually measured. Obrist and colleagues  reported attenuation of the magnitude of correlation between PuTT and SBP after the administration of a β-adrenergic antagonist. Chan and colleagues  found that during headup tilt, presumably causing central hypovolemia, PuTT increased, but this was accompanied by a significant decrease in the RR interval. In the study of Payen and colleagues , PuTT decreased after the use of both norepinephrine (which increased systolic BP) and salbutamol (which left systolic BP unchanged, but HR increased from 66 to 125 bpm). The design of this study is very different from those of previous ones in terms of patient population, anesthesia and, more importantly, CO was measured. Coefficients of correlation between changes in PuTT and SVI were low (β = -0.289 and -0.188) for both arterial and plethysmographic measurements, although they were statistically significant only for arterial PuTT (P=0.01). Low correlation between changes in SVI and changes in PuTT (Figure 3, Figure 5) makes PuTT a weak indicator of changes in CO.
There are several limitations to our study. As we said above all new minimal or non-invasive methods of measuring CO are less reliable in measuring absolute values of CO, and this includes the method based on bioreactance that we used. For this reason we didn’t analyze absolute values of CO, but only the percentage of changes. Furthermore, the study design evaluated only increases (not decreases) in both CO and systolic BP.
In conclusion, our findings demonstrate that PuTT changes do not correlate closely neither with changes in BP nor with changes in CO. The study was sufficiently powered and type 2 error was ruled out. Our conclusion is that since cardiac performance or changes in BP in anesthetized ventilated patients are not indicated by PuTT, PuTT measurement is currently of limited value in regular clinical practice.
1. Giglio M, Dalfino L, Puntillo F, Rubino G, Marucci M et al. Haemodynamic goaldirected therapy in cardiac and vascular surgery. A systematic review and meta-analysis. Interact. Cardiovasc Thorac Surg. 2012, 15(5): 878-887.
2. Giglio MT, Marucci M, Testini M, Brienza N. Goal-directed haemodynamic therapy and gastrointestinal complications in major surgery: a meta-analysis of randomized controlled trials. Br J Anaesth. 2009, 103(5): 637-646.
3. Cannesson M, Pestel G, Ricks C, Hoeft A, Perel A. Hemodynamic monitoring and management in patients undergoing high risk surgery: a survey among North American and European anesthesiologists. Crit Care. 2011, 15(4): R197.
4. Vincent JL, Rhodes A, Perel A, Martin GS, Della Rocca G et al. Clinical review: update on hemodynamic monitoring - a consensus of 16. Crit Care. 2011, 15(4): 229.
5. Raval NY, Squara P, Cleman M, Yalamanchili K, Winklmaier M et al. Multicenter evaluation of noninvasive cardiac output measurement by bioreactance technique. J Clin Monit Comput. 2008, 22(2): 113–119.
6. Pierre Squara, Dominique Rotcajg, Dominique Denjean, Philippe Estagnasie et al. Comparison of monitoring performance of bioreactance vs. pulse contour during lung recruitment maneuvers. Crit Care 2009, 13(4): R125.
7. Marque S, Cariou A, Chiche J-D, Squara P. Comparison between Flotrac-Vigileo and bioreactance, a totally noninvasive method for cardiac output monitoring. Crit Care. 2009, 13(3): R73.
8. Smith RP, Argod J, Pepin J-L, Levy PA. Pulse transit time: an appraisal of potential clinical applications. Thorax. 1999, 54(5): 452-458.
9. Naschitz JE, Bezobchuk S, Mussafia-Priselac R, Sundick S, Dreyfuss D et al. Pulse transit time by R-wave-gated infrared photoplethysmography: review of the literature and personal experience. J Clin Monit Comput. 2004, 18(5-6): 333-342.
10. Fooa JYA, Lim CS. Pulse transit time as an indirect marker for variations in cardiovascular related reactivity. Technol Health Care. 2006, 14(2): 97-108.
11. Obrist PA, Light KC, McCubbin JA, Hutcheson JS, Hoffer JL et al. Pulse transit time: relationship to blood pressure and myocardial performance. Psychophysiology. 1979, 16(3): 292-301.
12. Pollak MH, Obrist PA. Aortic-radial pulse transit time and ECG Q-wave to radial pulse wave interval as indices of beatby-beat blood pressure change. Psychophysiology. 1983. 20(1): 21-28.
13. Bijker JB, van Klei WA, Kappen TH, van Wolfswinkel L, Moons KG,et al. Incidence of intraoperative hypotension as a function of the chosen definition. Anesthesiology. 2007, 107(2): 213–220.
14. Thiele RH, MD, Nemergut EC, Lynch III C. The clinical implications of isolated alpha1 adrenergic stimulation. Anesth Analg. 2011, 113(2): 297–304.
15. 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(1): 136-141.
16. Hettrick DA, Pagel PS, Warltier DC. Desflurane, sevoflurane and isoflurane impair canine left ventricular-arterial coupling and mechanical efficiency. Anesthesiology. 1996, 85(2): 403-413.
17. Deryck YL, Fonck K, Baerdemaeker DE, Naeije R, Brimioulle S et al. Differential effects of sevoflurane and propofol anesthesia on left ventricular–arterial coupling in dogs. Acta Anaesthesiol Scand. 2010, 54(8): 979-986.
18. Chan GS, Middleton PM, Celler BG, Wang L, Lovell NH et al. Change in pulse transit time and pre-ejection period during head-up tilt-induced progressive central hypovolaemia. J Clin Monit Comput. 2007, 21(5): 283–293.