In recent years, continuous flow left ventricular assist devices (cf-LVAD) have been applied as a bridge to heart transplantation (BTT) in end-stage heart failure patients. However, as these devices generate continuous flow, arterial pulsations are considerably reduced or completely absent depending on the remaining cardiac function.1,2 Thus, even though invasive measurement of arterial blood pressure is considered standard procedure in the early postoperative stage, noninvasive measurement for long-term evaluation in these patients has been challenging.3
For subsequent evaluation of blood pressure after the early postoperative stage, noninvasive means of assessment are essential for proper patient care. In clinical practice, the classical Riva-Rocci/Korotkoff method, used most frequently to obtain arterial blood pressure values, is based on the detection of audible sounds and oscillometric methods that analyze pressure variations in an upper arm cuff during deflation. However, patients supported by cf-LVAD have reduced pulsatility, which may fall below the sensitivity range of these blood pressure measurement devices. As a result, noninvasive blood pressure measurements become unreliable or even impossible to obtain.
Recently, a new blood pressure measurement device was introduced that is based on the volume-clamp method of Peňáz using a finger cuff.4 This device, the Nexfin monitor (BMEYE B.V., Amsterdam, The Netherlands, Figure 1), allows noninvasive and continuous beat-to-beat brachial arterial pressure measurements. It was recently validated against the Riva-Rocci/Korotkoff method and is considered to be a reliable method for noninvasive continuous blood pressure measurement.5,6 In this study, we prospectively investigated the clinical feasibility of measuring noninvasive blood pressure during conditions of reduced pulsatility with the Nexfin monitor.
The study consisted of two parts. First, noninvasive arterial blood pressure measured by the Nexfin monitor was compared with invasive arterial pressure in conditions of reduced arterial pressure pulsatility during cardiac surgery. Second, the performance of the Nexfin monitor during the measurement of noninvasive blood pressure in patients with a cf-LVAD was evaluated during regular echocardiographic patient examination.
Noninvasive Arterial Blood Pressure by the Nexfin Monitor
The Nexfin monitor applies a fast feedback system to control pressure in a finger cuff, such that finger artery diameter is kept at a constant level (clamped) as measured by an optical plethysmograph in the finger cuff.7 The cuff has to equal real-time finger arterial blood pressure at any instant to achieve this clamped state. When the artery is in its unloaded state, cuff pressure equals continuous finger arterial pressure. The pressure is automatically calibrated using an algorithm that determines the unloaded state of the artery. To that end, the plethysmogram is analyzed during short periods of steady cuff pressure levels. The calibration procedure is referred to as “Physiocal” and is performed at regular intervals. By using software, a transfer function model is applied to reconstruct the brachial arterial blood pressure waveform from the finger arterial pressure and to correct for pressure differences because of resistance to flow.8,9 Together these techniques provide a noninvasive arterial blood pressure waveform reconstruction of the finger arterial blood pressure waveform.10,11 Additionally, a heart reference system is used to measure and correct for hydrostatic differences between the finger and the heart.
Measurements During CPB
To investigate the noninvasive arterial blood pressure (NAP) measured by the Nexfin monitor in conditions of reduced arterial pressure pulsatility, the values were compared with invasive arterial pressure during cardiopulmonary bypass (CPB). This study was approved by the Medical Ethics Committee of the Academic Medical Center of the University of Amsterdam. Written informed consent was obtained from 21 patients scheduled for coronary artery bypass grafting or valve replacement or reconstruction. Figure 2 shows a flow diagram of patient inclusion.
All patients received standard monitoring (ECG, pulse oximetry, temperature, end-tidal carbon dioxide partial pressure). Anesthesia was induced with propofol and continued with morphine and sufentanil. Muscular blockade was achieved with pancuronium and sedation was achieved with midazolam. Blood pressure was regulated by pharmacological vasodilatation (nitroglycerine) or vasoconstriction (ephedrine and metaraminol). Heparin was given before CPB. The investigator was blinded for the number and quantity of pharmacological agents used per patient.
Mechanical circulation was achieved by means of a roller pump based or a centrifugal heart-lung machine (Stockert, Irvine, CA, USA). At the initiation of CPB, paralysis of the heart was achieved by administration of cardioplegia in the ascending aorta after which the heart-lung machine provided extracorporeal circulatory support. Pump flow data from the heart lung machines and blood gas levels during surgery were not considered in this study.
Intraarterial pressure (IAP) was measured in the radial artery (IAP) with a fluid filled 20-G radial artery catheter (Ref RA-04020, Arrow International Inc, Reading, PA) connected to a standard transducer (Edwards Lifescience, Irvine, CA). After zeroing the manometer system, its dynamic response was checked by the fast flush technique.12 The arterial line was flushed when considered necessary by the anesthesiologist. The IAP was displayed by a monitor (Philips Medical Systems Andover, MA) and acquired through an analog output (Hewlett Packard HPM1006A). Subsequently, an appropriate size Nexfin monitor finger cuff was applied to the mid-phalanx of the middle finger13 ipsilateral to the invasive blood pressure measurement device. The heart reference system was set at the level of the pressure transducer of the arterial line.13 NAP and IAP were measured simultaneously during the CPB phase. Blood pressure signals were sampled at 200 Hz and the digital data were stored for analysis.
During the CPB phase, 10-second averages were calculated over a period of 30 minutes. The starting point of the data analysis was the moment when arterial pulsations disappeared from the IAP signal after administration of cardioplegia. The NAP periods during a Physiocal (calibration of NAP) were excluded from analysis. Pressure oscillations were computed as the difference between maximum and minimum pressure within a 1-second interval. Individual and group averaged NAP–IAP differences were determined over the 30-minute time range. Bland–Altman plots were made of all data, and XY-plots (NAP vs IAP) were made for each patient.
Measurements During cf-LVAD Support
Blood pressure was measured with the Nexfin monitor during regular echocardiographic analyses of 10 patients on a cf-LVAD. All patients were supported by a Heartmate II LVADs (Thoratec Corp, Pleasanton, CA). During these evaluations, a Pump Speed Change Procedure (PSCP) was performed evaluating ventricular properties at different pump speed settings. After blood anticoagulation with heparin was administered to the patient, the pump speed setting of the Heartmate II LVAD was reduced from the baseline setting to 6,000 RPM in steps of 1,000 RPM for 1-minute per step. At each step, anatomical and functional properties of the heart were visualized with echocardiography. Thereafter, the pump speed was set back to the baseline settings of the LVAD. Systolic, diastolic, and mean arterial pressures (MAP) were determined from the pressure waveform as 10 second averages for each pump speed setting during the procedure.
NAP vs IAP During CPB
Twenty-one patients gave informed consent (Table 1). Eighteen of these patients underwent coronary artery bypass graft surgery and three patients received an aortic valve replacement. Three patients were excluded from the study because of operational or technical problems involving a dislocated heart reference system and a malfunctioning cuff (Figure 2). In a third patient, the Nexfin failed to register blood pressure despite continuous startup attempts after initiation of CPB. Ultimately, 18 patients were considered for further analysis. A frequency plot of all IAP and NAP data (3240 comparisons) showed normal distribution (data not shown). MAP expressed as mean ± SD was 56.4 ± 10.5 mm Hg for IAP and 55.2 ± 11.4 mm Hg for NAP. One1 patient was supported by a centrifugal pump based heart-lung machine and 17 patients were supported by a roller pump based heart-lung machine. The roller pump based heart-lung machine generated small pressure oscillations in arterial pressure. The centrifugal pump based heart-lung machine generated pressure with less pulsatility which can be regarded as constant pressure. Mean amplitude of the pressure oscillations was 4.3 ± 3.8 mm Hg for IAP and 3.8 ± 3.5 for NAP. Mean pressures and pressure oscillations measured by the two techniques were not significantly different (p = 0.53 and p = 0.61, Table 2). NAP-IAP differences were displayed in Bland-Altman fashion (Figure 3). The correlation between NAP and IAP for each individual patient was represented in XY-plots (Figure 4) with R2 values between 0.65 and 0.99 (Table 3). NAP − IAP average difference of all patients together was −1.3 ± 6.5 mm Hg (Table 2). The average difference in pulse oscillations was −0.5 ± 1.0 (Table 2).
NAP in Patients With CF-LVAD
The Nexfin monitor was applied to measure noninvasive arterial blood pressure in 10 outpatients (Table 1) during a PSCP. All measurements were successful. NAP was assessed at all pump speed settings without any measurement difficulties throughout the procedures.
Averages of the pressures recorded during the PSCP are shown in Figure 5. The pressure recordings displayed characteristic physiological changes in arterial pressure waveform induced by changes in pump speed settings of the cf-LVAD. Systolic pressures remained relatively constant, whereas diastolic and MAPs increased with increase in pump speed settings. As a result, the pulse pressures decreased in response to an increase in pump speed. Furthermore, the pressure waveform acquired a sinusoidal shape at high pump speed settings (Figure 6). At these pump speed settings, simultaneous echocardiography visualization of the heart revealed that the aortic valve was permanently closed. When the pump speed was reduced, the dictrotic notch reappeared on the blood pressure waveform whereas the aortic valve started opening. Hence, the notch appeared at the upper part of the waveform and the waveform acquired a letter M-shape. Further decrease in the pump speed caused the dicrotic notch on the pressure waveform to shift towards the diastolic pressure.
This study showed that noninvasively assessed absolute arterial pressure values with the Nexfin monitor under hemodynamic conditions of reduced pulsatility were similar to those measured invasively using a pressure catheter. Furthermore, the Nexfin monitor appropriately reproduced the arterial pressure waveform in patients supported by a cf-LVAD. Studies performed during the early development of the Finger pressure methodology already suggested that noninvasively measured continuous arterial pressure might become possible in humans by applying this technology.14 Even though no requirements exist regarding the accuracy of noninvasive continuous blood pressure measurements under reduced pulsatility, NAP-IAP differences fell within the general guidelines of the Association for the Advancement of Medical Instrumentation15 with an average bias of ±5 mm Hg and a standard deviation of 8 mm Hg. However, these guidelines are based on comparisons between the Riva-Rocci/Korotkoff and invasive arterial blood pressure measurements.5 Therefore, such guidelines may be less appropriate in determining the efficacy of devices that measure continuous blood pressure noninvasively such as the Nexfin monitor.5 Furthermore, the T-Line Tensymeter (Tensys Medical, Inc., San Diego, CA) is also used to measure continuous noninvasive arterial blood pressure in regular patients. However, the capability of this device to assess blood pressure measurements under reduced arterial pulsations has not yet been investigated. Another approach based on a modified Riva-Rocci method combined with a wrist watch ultrasound sensor has previously been suggested in the past to measure blood pressure during reduced arterial pulsatility in patients with cf-LVADs.3 However, such approach, even if feasible, would only provide pressure approximations based on intermittent measurements of the blood pressure.
Regular and reliable monitoring of arterial blood pressure is considered important during the postoperative care of continuous-flow pump recipients and further management after discharge from the hospital. For example, monitoring the ventricular afterload is important as it affects the flow through a cf-LVAD during support. A high afterload reduces pump flow and may compromise unloading of the ventricle. Furthermore, the noninvasive nature of measurement with the Nexfin monitor is not associated with complications related to invasive cannulation. The device is easy to apply and makes it especially appropriate to measure blood pressure during clinical evaluations.
The arterial pressure waveform may be applied to assess both ventricular and LVAD function during mechanical circulatory support. Pressure pulsatility in these patients is determined by the relative contribution of the remaining ventricular function and the flow through the cf-LVAD. During the PSCP, we observed distinct changes in the arterial pressure waveform. At high pump speed settings, the pressure waveform became a sinusoid wave without a dicrotic notch. As the pump speed setting was decreased, the dicrotic notch reappeared. Initially, the pressure waveform acquired a letter M-shape whereas further decrease in pump speed caused the notch to shift toward the diastolic pressure. This observation may be associated with increased duration of aortic valve opening with decrease in cf-LVAD speed.2 Thus, opening or permanent closure of the aortic valve can by identified by assessing the dicrotic notch on the arterial pressure waveform. Furthermore, as the arterial pressure waveform can now be obtained noninvasively, NAP may be applied to assess the maximum positive time derivative of the arterial blood pressure (arterial dP/dt max). As LVADs are also applied as bridge to recovery,16,17 the arterial dP/dtmax may become a potential measure to assess possible recovery of the ventricle. The possibility of applying the arterial dP/dtmax to quantify left ventricular function during cf-LVAD support is an interesting subject for future investigation.
In this study, we used simultaneous ipsilateral measurements to compare IAP and NAP. If the finger cuff is placed ipsilateral to the pressure catheter, interference with NAP measurements may occur as the finger cuff is located distal to the pressure catheter. Contralateral measurement of IAP and NAP would have avoided this potential problem. On the other hand, when using this method, measurements may be affected by physiological pressure differences between the left and right arm. The approximation of IAP by NAP suggests that the influence of the catheter is probably small.
During blood pressure measurement with Nexfin, the arterial diameter of the finger is clamped at a constant value. If pulsations in the blood pressure disappear after the measurement has started, Nexfin will show a constant pressure waveform. Such a measurement can be continued for hours. For the proper determination of the diameter at which the finger arteries should be clamped, some pulsation is needed to assess the pressure—diameter relationship of the arteries (Wesseling method) and start the measurement.7 If blood pressure changes are very small at measurement startup, this volume clamp diameter cannot be determined and measurement cannot start. The same occurs when the device suspends a measurement after an error and tries to resume after a few seconds during a period without pulsations or with very small pulsations. This was observed in only one patient during CPB but in none of the patients on cf-LVAD support. In the aggregate, pulse pressures higher than 5 mm Hg seem to be well within the range of measurement of the Nexfin monitor.
Low arterial pulse pressures, as observed in the cf-LVAD patients, did not seem to be a limitation in acquiring noninvasive arterial blood pressure measurements. However, vasoconstriction of the finger arteries manifested by cold hands may limit measurement of finger arterial pressure. Nevertheless, the blood pressure waveform was assessed in patients with a cf-LVAD during PSCP without difficulties. During measurement, the arteries are never occluded so that arterial blood flow to the finger is always maintained. Some discoloring of the fingertip may occur, because the venous flow is impeded. This discoloring disappears as soon as the cuff is released. Most subjects soon get accustomed to the cuff pressure on the finger, however, some might experience discomfort. Finally, we chose not to compare noninvasively- with invasively measured blood pressures during the PSCP because these patients were anticoagulated with heparin to avoid possible thrombotic complications resulting from a decrease in the number of revolutions of the cf-LVAD.
This study shows that noninvasive continuous arterial blood pressure measurement is possible in patients with reduced pulsatility because of continuous flow mechanical circulatory support. During CPB, NAP was similar to IAP. The Nexfin monitor also successfully measured blood pressure in patients supported by a cf-LVAD. Therefore, we expect that the Nexfin monitor will enable clinicians to measure the arterial blood pressure waveform noninvasively in patients supported by cf-LVADs without the risks related to invasive measurements.
1. Potapov EV, Loebe M, Nasseri BA, et al
: Pulsatile flow in patients with a novel nonpulsatile implantable ventricular assist device. Circulation
102: III183–III187, 2000.
2. Stainback RF, Croitoru M, Hernandez A, et al
: Echocardiographic evaluation of the Jarvik 2000 axial-flow LVAD. Tex Heart Inst J
32: 263–270, 2005.
3. Schima H, Boehm H, Huber L, et al
: Automatic system for noninvasive blood pressure determination in rotary pump recipients. Artif Organs
28: 451–457, 2004.
4. Wesseling KH: A century of noninvasive arterial pressure measurement: from Marey to Penaz and Finapres. Homeostasis
36: 2–3, 1995.
5. Eeftinck Schattenkerk DW, Van Lieshout JJ, Van den Meiracker AH, et al
: Nexfin noninvasive continuous blood pressure validated against Riva-Rocci/Korotkoff. Am J Hypertens
22: 378–383, 2009.
6. Akkermans J, Diepeveen M, Ganzevoort W, et al
: Continuous noninvasive blood pressure monitoring, a validation study of Nexfin in a pregnant population. Hypertens Pregnancy
28: 230–242, 2009.
7. Wesseling KH, De Wit B, Van der Hoeven GMA, et al
: Physiocal, calibrating finger vascular physiology for Finapres. Homeostasis
36: 67–82, 1995.
8. Gizdulich P, Imholz BPM, van den Meiracker AH, et al
: Finapres tracking of systolic pressure and baroreflex sensitivity improved by waveform filtering. J Hypertens
14: 243–250, 1996.
9. Gizdulich P, Prentza A, Wesseling KH: Models of brachial to finger pulse wave distortion and pressure decrement. Cardiovasc Res
33: 698–705, 1997.
10. Westerhof BE, Guelen I, Parati G, et al
: Variable day/night bias in 24-h noninvasive finger pressure against intrabrachial artery pressure is removed by waveform filtering and level correction. J Hypertens
20: 1981–1986, 2002.
11. Bogert LWJ, Harms MP, Pott F, et al
: Reconstruction of brachial pressure from finger arterial pressure during orthostasis. J Hypertens
22: 1873–1880, 2004.
12. Gardner RM: Direct blood pressure measurement—dynamic response requirements. Anesthesiology
54: 227–236, 1981.
13. Nexfin HD: Operator's manual. 0086-06
. Amsterdam, The Netherlands, BMEYE B.V., 2008.
14. Smit NT, Wesseling KH, Wit B: Evaluation of two prototype devices producing noninvasive, pulsatile, calibrated blood pressure measurement from a finger. J Clin Monit
1: 17–29, 1985.
15. AAMI: American National Standard for Manual, Electronic, or Automated Sphygmomanometers
. Arlington, VA, Association for the Advancement of Medical Instrumentation, 2002.
16. Frazier OH, Myers TJ: Left ventricular assist system as a bridge to myocardial recovery. Ann Thorac Surg
68: 734–741, 1999.
Copyright © 2010 by the American Society for Artificial Internal Organs
17. Mancini D, Beniaminovitz A, Levin H, et al
: Low incidence of myocardial recovery after left ventricular assist device implantation in patients with chronic heart failure. Circulation
98: 2383–2389, 1998.