Blood pressure (BP) is a key intraoperative hemodynamic parameter that guides the anesthesiologists in maintaining an optimal balance between the patients’ stress condition and the depth of anesthesia. Basically, traditional noninvasive BP measurements by an upper arm cuff device at every 3 to 5 minutes intervals is a standard practice in all patients undergoing anesthesia, and is endorsed by the American Society of Anesthesiologists (ASA).1 However, intermittent noninvasive BP measurement may not be ideal in patients with high ASA status or in those undergoing high-risk surgeries. In such patients, continuous real-time BP monitoring using an intra-arterial catheter accrues a distinct advantage.2,3 However, due to its invasive nature, use of an arterial catheter may be accompanied by known complications, such as trauma, bleeding, hematoma, infection, thrombosis, embolism, distal ischemia, and the formation of pseudoaneurysms.4,5
Considering the advantages and disadvantages of these 2 methods, a continuous and beat-to-beat noninvasive BP monitoring might be of benefit in patients undergoing routine surgeries. TL-300 (Tensys Medical Inc., San Diego, CA) is a new generation device that displays real-time, continuous BP waveforms in a totally noninvasive manner. It measures BP based on radial artery applanation tonometry6 and monitors heart rate with real-time beat-to-beat systolic (SBP), diastolic (DBP), and mean blood pressure (MBP), which is comparable with the parameters measured by invasive BP determined by arterial catheter. Besides, the radial artery applanation tonometry allows for estimation of the cardiac output based on pulse contour analysis using the recorded BP curve.7,8
However, the performance of TL-300 technique has not been validated against the gold standard invasive BP technique. The objective of this retrospective study was to evaluate the accuracy and precision of the TL-300 against the standard invasive BP by radial artery catheter in patients undergoing elective neurosurgery.
PATIENTS AND METHODS
This study was approved by the research ethics committee of Fuzhou General Hospital of Nanjing Military Region (2014119Y). As the study involved retrospective analysis of patient data, the requirement for written informed consent from individual patients was waived off.
A comprehensive retrospective review was conducted in patients undergoing neurosurgery whose intraoperative BP was monitored by both invasive and noninvasive methods in Fuzhou General Hospital between December 2014 and March 2015. Owing to the lack of prior experience with TL-300 technology, the preliminary observation involved only in small number of patients who met the following criteria: (1) age more than 18 years; (2) receiving elective neurosurgery under general anesthesia; (3) radial pulse palpable in both wrists; (4) ASA status I to III; (5) in an intraoperative supine position. Patients would not be considered to receive TL-300 when they met the following criteria: (1) presence of arteriovenous shunts in the study limbs; (2) difference of >10 mm Hg between upper limbs as measured by brachial cuff BP technique; (3) congenital or acquired anatomic differences between the 2 wrists; (4) patients with systemic vasculopathies, cardiovascular disease, severe hypertension (baseline BP >180/110 mm Hg) or diabetes mellitus (preoperative fasting blood glucose >10 mmol/L); (5) history of vascular surgery of the study limbs; and (6) expected duration of operation <3 hours.
BP values of these patients were included in the current retrospective analysis only if the noninvasive BP monitoring was performed using TL-300 method, and invasive BP monitoring by radial arterial catheter method. Patients in whom the dorsalis pedis artery catheter was used for invasive BP measurement, as well as the cases where missing pairs of BP values exceeded >30%, were excluded from the analysis.
Monitoring and General Anesthesia
All patients were transferred to operation room without any premedication and received standard monitoring (ECG, SpO2, RR, PETCO2, and Narcotrend index; MonitorTechnik, Bad Bramstedt, Germany) intraoperatively. Invasive BP was measured by an arterial catheter placed in one of the radial arteries and connected to a transducer that was calibrated at the level of the patient’s right atrium. To evaluate the damping coefficient of an arterial line tubing system, a fast flush test was performed, and the artery line and transducer were inspected for any technical snag or for the presence of air bubbles that could cause erroneous readings. Noninvasive BP was measured by TL-300 device on the contralateral wrist. A special splint was used to immobilize the patients’ wrist in an extended position. After palpation and labeling of the patient’s radial artery, the sensor was affixed by using a bracelet placed on the ulnar side of the wrist guided by the alignment marks. The electromechanical system integrated in the bracelet positioned the sensor over the radial artery to obtain the maximum arterial pulse and optimal pressure signal. The bracelet was placed at the same level as the arterial catheter transducer. Patient’s weight and height was entered in the device, after which, the monitor displayed the heart rate, BP waveform, and numeric values (SBP, DBP, and MBP) at 1-second intervals.
General anesthesia was induced with propofol (2 to 2.5 mg/kg) or etomidate (0.2 to 0.6 mg/kg) followed by sufentanil (0.3 to 0.5 μg/kg) and rocuronium (0.6 to 0.8 mg/kg) or cisatracurium (0.2 to 0.3 mg/kg). After intubation, anesthesia was maintained with sevoflurane (2 to 3 vol/vol%) and remifentanil (0.3 to 0.5 mg/h). The depth of anesthesia was regulated to maintain Narcotrend index in D2 to E1 (46 to 20). Vascular active drugs were administered when the BP fluctuated higher or lower than ±20% to 30% of baseline values.
Baseline demographics and procedural details were sourced from the patients’ medical charts and the electronic database of our hospital. The original numeric values of BP acquired from both arterial catheter and TL-300 methods during the entire surgical procedure were obtained from the Anesthesia Information Management System (Medicalsystem Co., Ltd., Suzhou, China). All data were extracted independently in a predesigned data collection form by 2 authors (W.Q.L. and H.H.W.). Any discrepancies were crosschecked and a consensus reached by involving a third author (C.S.S.).
All data were expressed as mean±SD. Statistical analyses were conducted by using the IBM SPSS Statistics (version 22; SPSS Inc., Chicago, IL) and OriginPro (version 9.2.257; OriginLab, Northampton, MA) software according to previous publications.9–12 Normal distribution of data was assessed by P-P plots. To compare the relationship between invasive arterial catheter and TL-300, the data measured by 2 methods were analyzed by using linear regression and coefficient of determination (r2) computed. The bias for paired BP values was evaluated by using Bland-Altman plots for repeated measurements.13 Bias was calculated by the mean of individual differences between TL-300 and invasive arterial catheter method. On the basis of Bland-Altman analysis, the mean bias with SD and the corresponding 95% limits of agreement were presented. P<0.05 was considered as statistical significance.
Baseline Characteristics and Clinical Data
The data search yielded 28 patients in whom BP was monitored using both TL-300 and invasive BP monitoring concomitantly. Five patients were excluded because of >30% missing pairs of BP values and/or invasive BP monitoring by dorsalis pedis artery catheter. Additional 558 BP data were excluded because of missing pairs of BP values. A total of 4381 pairs of BP values (88.7%) pertaining to 23 patients (14 males and 9 females) at a total of 4939 given time were included in the final analysis. The data selection is shown in Figure 1. The baseline demographics and information pertaining to surgery and anesthesia are shown in Table 1.
Comparison Between Noninvasive and Invasive BP
Table 2 presented the comparison of BP values measured by TL-300 and arterial catheter. The range of noninvasive SBP, DBP, and MBP values was between 61 to 199, 31 to 132, and 40 to 144 mm Hg, respectively. And their average±SD values were 113.9±19.4, 70.7±14.4, and 84.7±14.9 mm Hg, respectively. The range of invasive SBP, DBP, and MBP values was between 65 to 192, 31 to 122, and 46 to 142 mm Hg, respectively. And their average±SD values were 112.6±18.9, 67.9±12.7, and 82.9±13.9 mm Hg, respectively. The overall distribution and P-P plots of noninvasive and invasive BP are presented in Figure 2 and Supplementary Figure 1, Supplemental Digital Content 1, https://links.lww.com/JNA/A28. The mean skewness of invasive SBP, DBP, and MBP was 0.037, 0.745, and 0.713, respectively, whereas it was 0.617, 0.485, and 0.626, respectively, with noninvasive measurement. The mean kurtosis of invasive SBP, DBP, and MBP was 0.917, 1.117, and 1.042, respectively, whereas it was 0.603, 0.580, and 0.695, respectively, with noninvasive measurement. Interestingly, invasive DBP values showed a significantly higher “skewness and kurtosis” than invasive SBP values. All distributions are leptokurtic with slight central tendencies and more extreme values than expected for a normal distribution. The summary of kurtosis and skewness with their SDs in invasive and noninvasive method are shown in Supplementary Table 1, Supplemental Digital Content 2, https://links.lww.com/JNA/A29.
Comparison Between Noninvasive and Invasive BP Difference
On comparing the 4381 SBP pairs, TL-300 showed a high coefficient of determination with the invasive SBP (y=0.978x+7.758, r2=0.908; Fig. 3A). We did not observe any apparent change in the regression relationship between either low or high BP. The noninvasive SBP values were higher than invasive values with a bias of 1.3±5.87 mm Hg (P<0.001). The 95% limits of agreement of bias for SBP were from −10.2 to 12.8 (Fig. 3D). The percentage of bias of SBP pairs between 2 methods within 5, 10, and 15 mm Hg was 64.2%, 92.9%, and 99.0%, respectively (Table 3). Compared with other nonlinear regression of SBP pairs, a linear model presented a higher coefficient of determination (Supplementary Table 2, Supplemental Digital Content 2, https://links.lww.com/JNA/A29).
For DBP pairs, TL-300 showed lower coefficient of determination in comparison with the invasive DBP (y=1.018x+1.561, r2=0.803; Fig. 3B); however, there was no apparent change in the regression relationship between either low or high BP. The noninvasive DBP values were higher than invasive values with a bias of 2.8±6.40 mm Hg (P<0.001). The 95% limits of agreement of bias for SBP were from −9.8 to 15.3 mm Hg (Fig. 3E). The percentage of bias of DBP pairs between 2 methods within 5, 10, and 15 mm Hg was 59.3%, 87.6%, and 96.6%, respectively (Table 3). Compared with other nonlinear regression of DBP pairs, a linear model also presented a higher coefficient of determination (Supplementary Table 3, Supplemental Digital Content 2, https://links.lww.com/JNA/A29).
For MBP pairs, TL-300 also showed a high coefficient of determination with the invasive MBP (y=1.033x−0.875, r2=0.922; Fig. 3C). Similarly, there was no apparent change in the regression relationship between either low or high BP. The noninvasive MBP values were higher than invasive values with a bias of 1.8±4.20 mm Hg (P<0.001). The 95% limits of agreement of bias for MBP were from −6.4 to 10.1 mm Hg (Fig. 3F). The percentage of bias of MBP pairs between 2 methods within 5, 10, and 15 mm Hg was 76.0%, 97.6%, and 99.7%, respectively (Table 3). Similarly, compared with other nonlinear regression of MBP pairs, a linear model presented a higher coefficient of determination (Supplementary Table 4, Supplemental Digital Content 2, https://links.lww.com/JNA/A29).
In the current retrospective study, we compared TL-300 noninvasive BP measurements with invasive radial artery beat-to-beat BP monitoring. The TL-300 determined SBP, DBP, and MBP accurately and precisely with association bias and precision for SBP, DBP, and MBP being 1.3±5.87, 2.8±6.40, and 1.8±4.20 mm Hg, respectively. Between 2 methods, the percentage of bias of BP pairs within 5, 10, and 15 mm Hg was acceptable (SBP: 64.2%, 92.9%, and 99.0%; DBP: 59.3%, 87.6%, and 96.6%; MBP: 76.0%, 97.6%, and 99.7%).
Noninvasive BP measured by oscillometric upper arm cuff is one of the classic methods for monitoring of hemodynamics. However, the inability to track rapid fluctuations in BP or short episodes of hemodynamic instability in real time is a limitation.14 According to the advanced perioperative management consensus statement, the criteria for an ideal hemodynamic monitoring device includes accuracy, reproducibility, ease of use, ready available, patient safety, besides the ability to capture as much information as possible to guide hemodynamic therapy. It also recommends that a continuous beat-to-beat hemodynamic parameter is more reliable than intermittent BP values.15 As invasive BP by arterial catheter provides accurate continuous BP values, it is considered the gold standard for BP monitoring and widely used intraoperatively,3 especially in critical patients, those with high ASA status or for high-risk surgeries, that may require fine tuning of the depth of anesthesia and timely management of hemodynamic instability. However, due to its invasive nature the arterial catheter may be difficult to place and lead to arterial complications.4,5 Thus, a reliable and continuous beat-to-beat noninvasive BP monitor would benefit elective high ASA and high-risk surgery patients. T-Line Tensymeter (Tensys Medical Inc.) is an alternative method with good accuracy and precision, especially for those patients in whom monitoring for rapidly changing individual BP readings are important intraoperatively without the need for arterial cannulation and where intermittent automatic oscillometric monitors may not be able to deliver a high-degree precision and accuracy.16,17 Ron Dueck et al18 revealed that the bias in MBP, SBP, and DBP for TL-200 was 2.3±5.9 Hg, 2.3±7.8, and 1.7±6.2 mm Hg, respectively, whereas for the TL-200pro the values were 0.72±5.15, −1.39±8.85, and 4.36±6.64 mm Hg, respectively.19 TL-300 is a new generation completely noninvasive device with a new, modified disposable locator in sensor to track patient’s radial artery7; however, the evidence base backing its accuracy and precision in clinical settings is yet to be established.
The Association for the Advancement of Medical Instrumentation ANSI/AAMI SP10 recommends that the test device would be clinically acceptable if the mean difference is <5 mm Hg with its SD <8 mm Hg compared with the reference method.20 On the basis of our results, the TL-300 met this recommendation. Moreover, it also met the standard of British Hypertension Society grade “A” accuracy which requires the percentage of bias within 5, 10, 15 mm Hg >60%, 85%, 95%, respectively.21 Besides, TL-300 had a relatively high precision with the narrow BP bias 95% limits of agreements (SBP: −10.2 to 12.8, DBP −9.8 to 15.3, and MBP: −6.4 to 10.2 mm Hg). Compared with previous T-Line devices series, TL-300 also showed good precision with more narrow SD.18,19,22–24 Besides, the current study also confirmed the ability of TL-300 to track BP changes reliably. Its coefficient of determination (r2) with invasive BP for SBP, DBP, and MBP was 0.908, 0.803, and 0.922, respectively. The r2 of TL-300 and invasive BP values were all high, although the DBP correlation was characteristically lower. No apparent change in the regression relationship was found at either low or high BP values. These findings suggest a good correlation of TL-300 measurements with the invasive method.
Although every effort was made to minimize bias during data collection and analyses, several limitations exist in the present study. First, as the current study was a retrospective analysis, a prospective study protocol is needed. Second, the sample size of 23 patients for this analysis was relatively small. Third, no very high or low body mass index as well as critically ill patients were included, although previous study demonstrated that TL-200 was competent for the obese18 and critically ill patients.7 The question on how to evaluate and validate new technologies for hemodynamic monitoring is of crucial importance and complex. Thus, what we need are methodologically sound and correct method comparison studies including a variety of different patients without selecting/excluding patients. Fourth, the evidence of kurtosis, that is presence of very high values in the distribution of BP values, was noted, which might be due to limited number of subjects included in the study. In addition, the invasive DBP values also showed a significantly higher skew and kurtosis than the invasive SBP values. Technically, the T-Line device measures MBP, then computes a derived SBP and DBP value. It is typical for the DBP values to be less reliable than the MBP and SBP values. This is noteworthy insofar as it affects the precision of the invasive DBP values. Some invasive BP values may have been affected by “damping,” due to inadequate flushing of the invasive BP line. This concern seems valid given that the data collection and analysis occurred in a retrospective manner. It is also noteworthy that kurtosis was lower for the TL-300 values than the invasive BP values for all 3 measurements. This information could imply an advantage for the TL-300 BP values. A prospective study should make it feasible to specifically look at details that affect invasive versus noninvasive beat-to-beat BP precision. Finally, the slope changes were noticed in the Bland-Altman plot for the lower or higher range of all BP values. These slope changes may indicate an element of subpopulation differences, for example, study subjects with undiagnosed hypertension, or alternatively, differences in anesthetic agent dose(s) and/or perhaps use of vasodilator medication to limit intraoperative hypertension.
To summarize, TL-300 is a promising new noninvasive beat-to-beat device with acceptable accuracy and precision in measurement of BP values in patients undergoing routine surgeries. It can, to a large extent, bridge the gap between noninvasive but intermittent and continuous but invasive BP monitoring. However, further prospective method comparison studies are needed.
1. Kim SH, Lilot M, Sidhu KS, et al. Accuracy and precision of continuous noninvasive arterial pressure monitoring compared with invasive arterial pressure: a systematic review and meta-analysis. Anesthesiology. 2014;120:1080–1097.
2. Cannesson M, Pestel G, Ricks C, et al. Hemodynamic monitoring and management in patients undergoing high risk surgery: a survey among North American and European anesthesiologists. Crit Care. 2011;15:R197.
3. Chen G, Zuo Y, Yang L, et al. Hemodynamic monitoring and management of patients undergoing high-risk surgery: a survey among Chinese anesthesiologists. J Biomed Res. 2014;28:376–382.
4. Brzezinski M, Luisetti T, London MJ. Radial artery cannulation: a comprehensive review of recent anatomic and physiologic investigations. Anesth Analg. 2009;109:1763–1781.
5. Haddad F, Zeeni C, El Rassi I, et al. Can femoral artery pressure monitoring be used routinely in cardiac surgery? J Cardiothorac Vasc Anesth. 2008;22:418–422.
6. Saugel B, Dueck R, Wagner JY. Measurement of blood pressure. Best Pract Res Clin Anaesthesiol. 2014;28:309–322.
7. Wagner JY, Sarwari H, Schon G, et al. Radial artery applanation tonometry for continuous noninvasive cardiac output measurement: a comparison with intermittent pulmonary artery thermodilution in patients after cardiothoracic surgery. Crit Care Med. 2015;43:1423–1428.
8. Saugel B, Meidert AS, Langwieser N, et al. An autocalibrating algorithm for non-invasive cardiac output determination based on the analysis of an arterial pressure waveform recorded with radial artery applanation tonometry: a proof of concept pilot analysis. J Clin Monit Comput. 2014;28:357–362.
9. Wagner JY, Saugel B. When should we adopt continuous noninvasive hemodynamic monitoring technologies into clinical routine? J Clin Monit Comput. 2015;29:1–3.
10. Vos JJ, Poterman M, Mooyaart EA, et al. Comparison of continuous non-invasive finger arterial pressure monitoring with conventional intermittent automated arm arterial pressure measurement in patients under general anaesthesia. Br J Anaesth. 2014;113:67–74.
11. Saugel B, Grothe O, Wagner JY. Tracking changes in cardiac output: statistical considerations on the 4-quadrant plot and the polar plot methodology. Anesth Analg. 2015;121:514–524.
12. Hapfelmeier A, Cecconi M, Saugel B. Cardiac output method comparison studies: the relation of the precision of agreement and the precision of method. J Clin Monit Comput. 2015. [Epub ahead of print].
13. Bland JM, Altman DG. Agreement between methods of measurement with multiple observations per individual. J Biopharm Stat. 2007;17:571–582.
14. Pytte M, Dybwik K, Sexton J, et al. Oscillometric brachial mean artery pressures are higher than intraradial mean artery pressures in intensive care unit patients receiving norepinephrine. Acta Anaesthesiol Scand. 2006;50:718–721.
15. Vincent JL, Rhodes A, Perel A, et al. Clinical review: update on hemodynamic monitoring—a consensus of 16. Crit Care. 2011;15:229.
16. Chung E, Chen G, Alexander B, et al. Non-invasive continuous blood pressure monitoring: a review of current applications. Front Med. 2013;7:91–101.
17. Rutten AJ, Ilsley AH, Skowronski GA, et al. A comparative study of the measurement of mean arterial blood pressure using automatic oscillometers, arterial cannulation and auscultation. Anaesth Intensive Care. 1986;14:58–65.
18. Dueck R, Goedje O, Clopton P. Noninvasive continuous beat-to-beat radial artery pressure via TL-200 applanation tonometry. J Clin Monit Comput. 2012;26:75–83.
19. Saugel B, Meidert AS, Hapfelmeier A, et al. Non-invasive continuous arterial pressure measurement based on radial artery tonometry in the intensive care unit: a method comparison study using the T-Line TL-200pro device. Br J Anaesth. 2013;111:185–190.
20. Bucci FA Jr, Waterbury LD. Prostaglandin E2 inhibition of ketorolac 0.45%, bromfenac 0.09%, and nepafenac 0.1% in patients undergoing phacoemulsification. Adv Ther. 2011;28:1089–1095.
21. O’Brien E, Petrie J, Littler W, et al. An outline of the revised British Hypertension Society protocol for the evaluation of blood pressure measuring devices. J Hypertens. 1993;11:677–679.
22. Janelle GM, Gravenstein N. An accuracy evaluation of the T-Line Tensymeter (continuous noninvasive blood pressure management device) versus conventional invasive radial artery monitoring in surgical patients. Anesth Analg. 2006;102:484–490.
23. Szmuk P, Pivalizza E, Warters RD, et al. An evaluation of the T-Line Tensymeter continuous noninvasive blood pressure device during induced hypotension. Anaesthesia. 2008;63:307–312.
24. Saugel B, Fassio F, Hapfelmeier A, et al. The T-Line TL-200 system for continuous non-invasive blood pressure measurement in medical intensive care unit patients. Intensive Care Med. 2012;38:1471–1477.