J Clin Monit Comput
DOI 10.1007/s10877-015-9744-1
ORIGINAL RESEARCH
Continuous noninvasive cardiac output determination using
the CNAP system: evaluation of a cardiac output algorithm
for the analysis of volume clamp method-derived pulse contour
Julia Y. Wagner4 • Julian Grond2 • Jürgen Fortin2 • Ileana Negulescu3
Miriam Schöfthaler3 • Bernd Saugel3,1
Received: 22 March 2015 / Accepted: 24 July 2015
Ó Springer Science+Business Media New York 2015
Abstract The CNAP system (CNSystems Medizintechnik
AG, Graz, Austria) provides noninvasive continuous arterial
pressure measurements by using the volume clamp method.
Recently, an algorithm for the determination of cardiac output
by pulse contour analysis of the arterial waveform recorded
with the CNAP system became available. We evaluated the
agreement of the continuous noninvasive cardiac output
(CNCO) measurements by CNAP in comparison with cardiac
output measurements invasively obtained using transpul-
monary thermodilution (TDCO). In this proof-of-concept
analysis we studied 38 intensive care unit patients from a
previously set up database containing CNAP-derived arterial
pressure data and TDCO values obtained with the PiCCO
system (Pulsion Medical Systems SE, Feldkirchen, Ger-
many). We applied the new CNCO algorithm retrospectively
to the arterial pressure waveforms recorded with CNAP and
compared CNCO with the corresponding TDCO values
(criterion standard). Analyses were performed separately for
(1) CNCO calibrated to the first TDCO (CNCO-cal) and (2)
& Bernd Saugel
bernd.saugel@gmx.de
1
Present Address: Department of Anesthesiology, Center of
Anesthesiology and Intensive Care Medicine, University
Medical Center Hamburg-Eppendorf, Martinistrasse 52,
20246 Hamburg, Germany
2
CNSystems Medizintechnik AG, Reininghausstrasse 13,
8020 Graz, Austria
3
II. Medizinische Klinik und Poliklinik, Klinikum rechts der
Isar der, Technischen Universität München, Ismaninger
Strasse 22, 81675 Munich, Germany
4
Department of Anesthesiology, Center of Anesthesiology and
Intensive Care Medicine, University Medical Center
Hamburg-Eppendorf, Martinistrasse 52, 20246 Hamburg,
Germany
•
CNCO autocalibrated to biometric patient data (CNCO-auto).
We did not perform an analysis of trending capabilities
because the patients were hemodynamically stable. The
median age and APACHE II score of the 22 male and 16
female patients was 63 years and 18 points, respectively.
18 % were mechanically ventilated and in 29 % vasopressors
were administered. Mean ± standard deviation for CNCO-
cal, CNCO-auto, and TDCO was 8.1 ± 2.7, 6.4 ± 1.9, and
7.8 ± 2.4 L/min, respectively. For CNCO-cal versus TDCO,
Bland–Altman analysis demonstrated a mean difference of
?0.2 L/min (standard deviation 1.0 L/min; 95 % limits of
agreement -1.7 to ?2.2 L/min, percentage error 25 %). For
CNCO-auto versus TDCO, the mean difference was -1.4 L/
min (standard deviation 1.8 L/min; 95 % limits of agreement
-4.9 to ?2.1 L/min, percentage error 45 %). This pilot
analysis shows that CNCO determination is feasible in criti-
cally ill patients. A percentage error of 25 % indicates
acceptable agreement between CNCO-cal and TDCO. The
mean difference, the standard deviation, and the percentage
error between CNCO-auto and TDCO were higher than
between CNCO-cal and TDCO. A hyperdynamic cardiocir-
culatory state in a substantial number of patients and the
hemodynamic stability making trending analysis impossible
are main limitations of our study.
Keywords Cardiac output
Hemodynamic monitoring

Vascular unloading technology
Pulse contour analysis

Intensive care unit
1 Introduction
The CNAP system (CNSystems Medizintechnik AG, Graz,
Austria) allows continuous noninvasive beat-by-beat
recording of the arterial pressure waveform [1, 2].
123

The CNAP device is easy to apply and uses an enhancement
of the Peňáz principle, a method first described in 1973, to
generate an arterial waveform [2, 3]. For arterial waveform
recording, an inflatable finger cuff is applied to the patient’s
finger [2]. The finger artery’s diameter is assessed by an inte-
grated photo-plethysmograph that controls that the blood vol-
ume in the finger artery is kept constant. A controller adjusts the
finger cuff pressure in order to keep the blood volume constant
throughout the cardiac cycle. The pressure that is required to
keep the volume constant corresponds to the true arterial
pressure waveform [2]. But the original Peňáz principle is not
working if arterial diameter and wall tension are altered due to
vasoconstriction and vasodilation. The CNAP device elimi-
nates such vasomotoric influences by the use of concentrically
interlocking loops and a so-called VERIFI-algorithm (‘‘Vaso-
motoric Elimination and Reconstructed IdentiFication of the
Initial setpoint’’) [4]. This algorithm continuously analyzes the
shape of the waveforms and thereby allows distinguishing
between blood volume shifts due to changes in arterial pressure
and those due to changes of the arterial diameter. This dis-
tinction is crucial for stable long-term tracking of arterial
pressure. The resulting finger cuff derived arterial pressure
signal is calibrated by a transfer function to oscillometrically
obtained arterial pressure values using an upper arm cuff.
Studies have shown that the technology provides arterial
pressure values with reasonable agreement in comparison
with invasive arterial pressure measurements in anesthetized
patients during surgical procedures [5–7], intensive care unit
(ICU) patients [8, 9], and emergency department patients [10].
Recently, the manufacturer released a novel cardiac
output (CO) algorithm for continuous noninvasive CO
(CNCO) monitoring based on the pulse contour analysis of
the noninvasively obtained arterial pressure waveform. The
proprietary CNCO algorithm incorporates physiological
markers such as the areas under the pulse waveform during
systole and during diastole, accounting for the physiolog-
ical factors of preload, contractility, afterload and com-
pliance of the vessels. The CNCO algorithm was finetuned
using machine learning tools on comprehensive data sets.
So far, no validation studies evaluating the CO mea-
surements by CNAP–CNCO exist.
The aim of our proof-of-concept analysis was to assess
the agreement of CNCO measurements in comparison with
CO measurements invasively obtained using transpul-
monary thermodilution (TDCO) in ICU patients.
2 Materials and methods
2.1 Study design and patients
The study was performed in a medical ICU of a German
university hospital (Klinikum rechts der Isar der
123
J Clin Monit Comput
Technischen Universität München, Munich, Germany) and
was approved by the institutional ethics committee
(Ethikkommission der Fakultät für Medizin der Technis-
chen Universität München). All patients or their legal
representatives gave written informed consent.
In order to evaluate the novel CNCO algorithm, we
retrospectively applied the new CNCO algorithm to ICU
patients from a previously set up database containing
(a) CNAP-derived arterial pressure data and waveforms
and (b) TDCO values obtained with the PiCCO system
(Pulsion Medical Systems SE, Feldkirchen, Germany).
When collecting data for this hemodynamic database, we
simultaneously recorded invasive and noninvasive arterial
pressure waveforms over a total time period of 15 min. The
total time period was split into 3 intervals of 5 min with
varying time between the 3 time points depending on the
clinical routine. The TDCO measurements were performed
at the beginning of each 5-min interval. This structured
hemodynamic database including TDCO values and arte-
rial pressure waveforms obtained with the CNAP system
now allowed us to apply the newly released CNCO algo-
rithm to the arterial pressure waveforms recorded with
CNAP and to compare the resulting CNCO values with the
corresponding TDCO values (criterion standard).
Data on arterial pressure measurements from the data-
base used for the present study have been previously
reported in a method comparison analysis [8].
In the database, arterial pressure data from a total of 57
patients were available. Arterial pressure waveform
recordings of 6 patients could not be used for the evalua-
tion of the CNCO algorithm because in 4 patients no
CNAP arterial waveforms had been recorded and in 2
patients the data set was incomplete and hence data syn-
chronization impossible. We additionally excluded 13
patients from data analysis because of an untypically deep
dicrotic notch, making beat detection and hence the cal-
culation of heart rate as well as pulse contour analysis
unreliable. Therefore, we included 38 patients in the final
statistical analysis.
2.2 Cardiac output determination
by transpulmonary thermodilution
Transpulmonary thermodilution was used as criterion
standard method for CO determination in this study. As
described above, when collecting arterial waveform data
for the database on CNAP arterial pressure measurements
[8], we simultaneously systematically recorded TDCO
data. We recorded 3 TDCO values in each patient. TDCO
measurements were performed as described previously
using a PiCCO-Plus or PiCCO-2 monitor (Pulsion Medical
Systems SE) [11, 12]. For a thermodilution measurement,
we injected 15 mL of iced 0.9 % saline via a central
J Clin Monit Comput
venous catheter in the central venous circulation. Using a
thermistor at the tip of the dedicated arterial catheter placed
in the abdominal aorta through the femoral artery (Pul-
siocath PV2015L20, Pulsion Medical Systems SE) we
recorded the thermodilution curves used to calculate
TDCO. We determined a TDCO value by averaging the
results of 3 consecutive thermodilution measurements.
Thus, 9 (3 9 3) indicator injections per patient were per-
formed resulting in 3 TDCO values per patient.
2.3 Cardiac output algorithm applied to the CNAP-
derived arterial pressure waveforms
For the comparison of the 3 TDCO values (criterion stan-
dard) with the corresponding, i.e. simultaneously mea-
sured, CNCO values, we applied the new CNCO algorithm
retrospectively to the arterial pressure waveforms recorded
with CNAP.
For the comparison between CNCO and TDCO, the
CNCO values were averaged over 30 s before the times the
TDCO values were read out.
The algorithm used for the determination of CNCO is
based on the pulse contour analysis of the arterial wave-
form noninvasively recorded with CNAP.
Analyses were performed separately for (1) CNCO
calibrated to the first TDCO (CNCO-cal) and (2) CNCO
autocalibrated to biometric patient data (CNCO-auto).
For CNCO-cal, the first TDCO value was used to
externally calibrate the CNCO value.
Autocalibration of CNCO resulting in CNCO-auto is
obtained by a proprietary nomogram using biometric data.
A calibration factor is obtained from a polynomial function
of the subject’s gender, age, height and weight. This cali-
bration factor is used to obtain absolute CO values and
accounts for the typical physiological conditions of pri-
marily healthy subjects. However, it cannot account for all
pathological conditions because it is solely based on bio-
metric data.
2.4 Data
Because we had recorded three TDCO values in each of the
38 patients in the database, 114 TDCO were available for
comparison with the corresponding CNCO values. In 3 of
the 114 data pairs, no CNAP arterial waveform was
available at the time of TDCO reading since the CNAP
monitor performed a arterial pressure setpoint search or
NBP calibration leaving us with 111 pairs of CO mea-
surements for comparative analyses.
As described above, for CNCO-cal, the first TDCO
value was used to externally calibrate the CNCO value.
Therefore, the first of the 3 pairs of CO measurements in
each patient could not be included in the Bland–Altman
analysis for the comparison CNCO-cal versus TDCO.
Therefore, this analysis was performed with the remaining
73 data pairs. For the analysis of CNCO-auto versus
TDCO, we used all 111 pairs of CO measurements for
comparative analyses.
2.5 Statistical analyses
For statistical analyses we used Excel 2011 (Microsoft Corp.,
Redmond, Washington, USA), MATLAB version 7.8.0
(R2009a) (The MathWorks Inc., Natick, Massachusetts,
USA) and Python version 2.7 (Python Software Foundation,
Beaverton, Oregon, USA). Patient characteristics are pre-
sented for all patients as absolute numbers with percentages or
median with interquartile range (25–75 % percentile range)
and for each individual patient. For CNCO-cal, CNCO-auto,
and TDCO, we separately calculated mean ± standard
deviation (SD) for the observed CO values. In addition, we
computed Bland–Altman plots considering that repeated
measurements in each individual were analyzed [13] and the
percentage error [14] for the comparison of TDCO versus
CNCO-cal and TDCO versus CNCO-auto.
3 Results
The characteristics of the 38 patients are presented in
Table 1.
Mean ± SD for CNCO-cal, CNCO-auto, and TDCO was
8.1 ± 2.7, 6.4 ± 1.9, and 7.8 ± 2.4 L/min, respectively.
For CNCO-cal versus TDCO, Bland–Altman analysis
demonstrated a mean difference of ?0.2 L/min with a SD
of 1.0 L/min and 95 % limits of agreement of -1.7 to
?2.2 L/min (Fig. 1a). The percentage error was 25 %.
For CNCO-auto versus TDCO, the mean difference was
-1.4 L/min with a SD of 1.8 L/min and 95 % limits of
agreement of -4.9 to ?2.1 L/min (Fig. 1b). The percent-
age error was 45 %.
Table 2 provides a direct comparison of the results for
the mean of differences, limits of agreement, and per-
centage error between CNCO-cal and CNCO-auto.
4 Discussion
In this proof-of-concept analysis, we retrospectively
applied the new CNCO algorithm to arterial pressure
waveforms previously recorded with CNAP and compared
CNCO-cal and CNCO-auto values with simultaneously
measured TDCO values (criterion standard) in 38 critically
ill patients in order to evaluate the measurement perfor-
mance of CNCO for the first time. According to the find-
ings of this pilot analysis, CNCO determination is feasible
123
 J Clin Monit Comput

Table 1 Patient and clinical

Patients
characteristics
Sex, male, n (%) 22 (56)
Weight (kg) 79 (69–85)
Height (m) 1.75 (1.66–1.79)
Body mass index (kg m-2) 25.9 (23.2–27.7)
Age (years) 63 (55–71)
Reason for intensive care unit admission
Acute liver failure, n (%) 9 (24)
Gastrointestinal bleeding, n (%) 3 (8)
Sepsis, n (%) 6 (16)
Respiratory insufficiency, n (%) 9 (24)
Other, n (%) 11 (29)
Clinical characteristics at day of study measurement
Mechanical ventilation, n (%) 7 (18)
Norepinephrine therapy, n (%) 11 (29)
Norepinephrine dose (lg min-1) 7 (2–10)
Midazolam therapy, n (%) 9 (24)
Advanced hemodynamic monitoring data
Cardiac index (L/min/m2) 4.1 (3.4–4.7)
Cardiac power index (W/m2) 0.6 (0.7–0.9)
Central venous pressure (mm Hg) 9 (15–19)
Global end-diastolic volume index (mL/m2) 686 (767–910)
Extravascular lung water index (mL/kg) 7 (9–13)
Pulmonary vascular permeability index 1.2 (1.4–2.1)
Systemic vascular resistance index (dyn s cm-5 m2) 1399 (1099–1718)
Intensive care unit scores
Acute Physiology and Chronic Health Evaluation II Score, points 18 (22–26)
Sequential Organ Failure Assessment score, points 5 (8–11)
Data are presented as the median and interquartile ranges (25–75 % percentile) or as absolute frequencies
with percentages

in critically ill patients. The percentage error of 25 %
between CNCO-cal and TDCO indicates acceptable
agreement, while the percentage error was higher for
CNCO-auto versus TDCO. Furthermore, the absolute mean
of differences was higher for CNCO-auto versus TDCO
than for CNCO-cal versus TDCO (-1.4 versus ?0.2 L/
min) and the limits of agreement were higher (-4.9 to
?2.1 versus -1.7 to ?2.2 L/min). These results suggest
that further refinements are needed to improve CNCO-auto
measurements that are autocalibrated based on the patient’s
biometrics.
In the recent years, a variety of new CO monitors were
developed, including minimally invasive and completely
noninvasive technologies [15–17]. As CO is a crucial
hemodynamic parameter in critical care and anesthesiology
[17] the concept of continuous noninvasive CO assessment is
intriguing. If the measurement performance of innovative,
noninvasive means of CO estimation was proven to be
comparable to established invasive methods, complications
associated with invasive CO assessment could be avoided.
Sequentially combining invasive, less invasive, and nonin-
vasive CO monitoring technologies might in the future allow
providing advanced hemodynamic monitoring individually
adapted for the patient’s specific clinical situation [18, 19].
In the context of our study findings, it is important that
the 30 % percentage error threshold suggested by Critchley
and Critchley [14] to define acceptable agreement between
different technologies for CO assessment has repeatedly
been questioned [17, 20, 21]. Peyton and Chong [20]
suggested that a clinically more realistic percentage error
cut-off value for the agreement of minimally invasive CO
monitors with thermodilution-derived CO would be 45 %.
Although we performed our method comparison study
by retrospectively applying the now available CNCO
algorithm to previously recorded arterial waveforms, our
results represent realistic clinical data because we delib-
erately did not screen the CNCO and TDCO measurements
for artifacts.
Our study allows conclusions about the measurement
performance of the new CNCO algorithm. However,

123
J Clin Monit Comput

Fig. 1 a Comparison of cardiac

output measurements obtained
with the CNAP system
(calibrated to the first
transpulmonary thermodilution
cardiac output; CNCO-cal) and
transpulmonary thermodilution-
derived cardiac output (TDCO);
b Comparison of cardiac output
measurements obtained with the
CNAP system (autocalibrated to
biometric data; CNCO-auto)
and transpulmonary
thermodilution-derived cardiac
output (TDCO). A Bland–
Altman plot with the mean of
differences (continuous
horizontal line) and 95 % limits
of agreement (1.96 9 standard
deviation; dashed horizontal
lines) is shown

because patients in this study—although critically ill— comparison with TDCO is not possible based on our data
were hemodynamically stable and we did not perform CO (96 % of consecutive CO measurements showed absolute
modifying maneuvers during study measurements, an differences \15 %). To evaluate whether CNCO can fol-
evaluation of the trending capabilities of CNCO in low changes in CO determined with a criterion standard

123

Table 2 Comparison between CNCO-cal and CNCO-auto
Mean of the differences (standard deviation) (L/min)
Upper and lower 95 % limit of agreement (L/min)
Percentage error (%)
CNCO continuous noninvasive cardiac output, TDCO transpulmonary thermodilution derived cardiac output
technique, prospective validation studies should include
interventions inducing rapid hemodynamic changes such as
a passive leg raising maneuver or a fluid challenge test.
While the present study is the first validation study for
CNCO, another technology for the noninvasive assessment
of CO based on the volume clamp method [i.e., ClearSight
system (Edwards Lifesciences, Irvine, CA, USA); formerly
known as Nexfin system (BMEye, Amsterdam, The
Netherlands)] has been evaluated in clinical studies
before—with inconsistent results. While Broch et al. [22]
demonstrated that the Nexfin system allows CO estimation
in cardiac surgery patients with a percentage error of
\30 % when compared with transpulmonary thermodilu-
tion, higher percentage errors between 38 % and 58 %
were reported in other clinical studies evaluating the sys-
tem in cardiac surgery patients in comparison with pul-
monary artery or transpulmonary thermodilution [23–26].
In 45 mixed ICU patients, Ameloot and colleagues [27]
revealed a percentage error of 36 % between Nexfin-CO
and CO determined by transpulmonary thermodilution. In
another study in medical ICU patients, Monnet et al. [28]
however observed a percentage error of 57 % when com-
paring Nexfin-derived cardiac index with transpulmonary
thermodilution measurements.
In general, when comparing an innovative hemody-
namic monitoring technology in a clinical method com-
parison study, a crucial question with regard to the study
design is which criterion standard method should be used
[19, 21]. Because thermodilution methods are considered
to be the clinical criterion standard for CO determination
[21], we chose to compare CNCO values with TDCO
measurements assessed by single-indicator transpulmonary
thermodilution. However, this approach made it necessary
to perform this validation analysis in a patient group treated
based on advanced hemodynamic monitoring using
transpulmonary thermodilution unrelated to the present
study. Therefore, we performed this study in critically ill
patients treated in the ICU despite the fact that this group
of complex patients is probably not the target population
for innovative technologies for continuous noninvasive CO
measurement in the future [19]. There are limitations of our
study that need to be mentioned. We retrospectively
applied the new CNCO algorithm to ICU patients from a
previously set up database. A relatively large number of
123
J Clin Monit Comput
CNCO calibrated to the CNCO autocalibrated to biometric
first TDCO (CNCO-cal) patient data (CNCO-cal)
0.2 (1.0) -1.4 (1.8)
-1.7 to ?2.2 -4.9 to 2.1
25 45
arterial waveforms proved to be unsuitable for the retro-
spective analysis in this proof-of-concept study.
Especially the fact, that patients treated for disease states
altering vasomotor tone such as sepsis and liver failure
were included in our study might be a limitation of our
study design with regard to generalizability to other patient
collectives such as elective surgical patients. In this con-
text, it is noteworthy that the mean TDCO in our study
collective was high (7.8 L/min). A hyperdynamic cardio-
circulatory state is characteristic for patients with low
systemic vascular resistance such as patients with sepsis or
liver failure. These special hemodynamic characteristics of
our study collective must be borne in mind when inter-
preting the results regarding CNCO-auto that is derived
based on biometric data.
5 Conclusions
This pilot analysis shows that CNCO determination is
feasible in critically ill patients. A percentage error of 25 %
indicates acceptable agreement between CNCO-cal and
TDCO.
Conflict of interest J.Y.W. and B.S. received refunds of travel
expenses from CNSystems Medizintechnik AG (Graz, Austria). B.S.
collaborates with Pulsion Medical Systems SE (Feldkirchen, Ger-
many) as member of the Medical Advisory Board. J.G. is employee of
CNSystems Medizintechnik AG (Graz, Austria), which has developed
and markets the CNAP system. J.F. is CEO and founder of CNSys-
tems, receives salary, has equity interests, and is named as inventor on
several CNAP-patents. For I.N. and M.S. there is no conflict of
interest to disclose.
References
1. Fortin J, Marte W, Grullenberger R, Hacker A, Habenbacher W,
Heller A, Wagner C, Wach P, Skrabal F. Continuous non-inva-
sive blood pressure monitoring using concentrically interlocking
control loops. Comput Biol Med. 2006;36:941–57.
2. Saugel B, Dueck R, Wagner JY. Measurement of blood pressure.
Best Pract Res Clin Anaesthesiol. 2014;28:309–22.
3. Penaz J. Current photoelectric recording of blood flow through
the finger. Cesk Fysiol. 1975;24:349–52.
4. Fortin J, Wellisch A, Maier K. CNAP—evolution of continuous
non-invasive arterial blood pressure monitoring. Biomed Tech.
2013;. doi:10.1515/bmt-2013-4179 (Berl).
J Clin Monit Comput
5. Jeleazcov C, Krajinovic L, Munster T, Birkholz T, Fried R,
Schuttler J, Fechner J. Precision and accuracy of a new device
(CNAPTM) for continuous non-invasive arterial pressure moni-
toring: assessment during general anaesthesia. Br J Anaesth.
2010;105:264–72.
6. Ilies C, Bauer M, Berg P, Rosenberg J, Hedderich J, Bein B, Hinz
J, Hanss R. Investigation of the agreement of a continuous non-
invasive arterial pressure device in comparison with invasive
radial artery measurement. Br J Anaesth. 2012;108:202–10.
7. Kumar GA, Jagadeesh AM, Singh NG, Prasad SR. Evaluation of
continuous non-invasive arterial pressure monitoring during
induction of general anaesthesia in patients undergoing cardiac
surgery. Indian J Anaesth. 2015;59:21–5.
8. Wagner JY, Negulescu I, Schofthaler M, Hapfelmeier A, Meidert
AS, Huber W, Schmid RM, Saugel B. Continuous noninvasive
arterial pressure measurement using the volume clamp method:
an evaluation of the CNAP device in intensive care unit patients.
J Clin Monit Comput. 2015;. doi:10.1007/s10877-015-9670-2.
9. Jagadeesh AM, Singh NG, Mahankali S. A comparison of a
continuous noninvasive arterial pressure (CNAP) monitor with an
invasive arterial blood pressure monitor in the cardiac surgical
ICU. Ann Card Anaesth. 2012;15:180–4.
10. Wagner JY, Prantner JS, Meidert AS, Hapfelmeier A, Schmid
RM, Saugel B. Noninvasive continuous versus intermittent arte-
rial pressure monitoring: evaluation of the vascular unloading
technique (CNAP device) in the emergency department. Scand J
Trauma Resusc Emerg Med. 2014;22:8.
11. Saugel B, Kirsche SV, Hapfelmeier A, Phillip V, Schultheiss C,
Schmid RM, Huber W. Prediction of fluid responsiveness in
patients admitted to the medical intensive care unit. J Crit Care.
2013;28(537):e531–9.
12. Saugel B, Phillip V, Gaa J, Berger H, Lersch C, Schultheiss C, Thies
P, Schneider H, Hollthaler J, Herrmann A, Schmid RM, Huber W.
Advanced hemodynamic monitoring before and after transjugular
intrahepatic portosystemic shunt: implications for selection of
patients—a prospective study. Radiology. 2012;262:343–52.
13. Bland JM, Altman DG. Agreement between methods of mea-
surement with multiple observations per individual. J Biopharm
Stat. 2007;17:571–82.
14. Critchley LA, Critchley JA. A meta-analysis of studies using bias
and precision statistics to compare cardiac output measurement
techniques. J Clin Monit Comput. 1999;15:85–91.
15. Marik PE. Noninvasive cardiac output monitors: a state-of the-art
review. J Cardiothorac Vasc Anesth. 2013;27:121–34.
16. Monnet X, Teboul JL. Minimally invasive monitoring. Crit Care
Clin. 2015;31:25–42.
17. Saugel B, Cecconi M, Wagner JY, Reuter DA. Noninvasive
continuous cardiac output monitoring in perioperative and
intensive care medicine. Br J Anaesth. 2014;114:562–75.
18. Alhashemi JA, Cecconi M, Hofer CK. Cardiac output monitoring:
an integrative perspective. Crit Care. 2011;15:214.
19. Wagner JY, Saugel B. When should we adopt continuous non-
invasive hemodynamic monitoring technologies into clinical
routine? J Clin Monit Comput. 2014;29:1–3.
20. Peyton PJ, Chong SW. Minimally invasive measurement of car-
diac output during surgery and critical care: a meta-analysis of
accuracy and precision. Anesthesiology. 2010;113:1220–35.
21. Saugel B, Reuter DA. Are we ready for the age of non-invasive
haemodynamic monitoring? Br J Anaesth. 2014;113:340–3.
22. Broch O, Renner J, Gruenewald M, Meybohm P, Schottler J,
Caliebe A, Steinfath M, Malbrain M, Bein B. A comparison of
the Nexfin(R) and transcardiopulmonary thermodilution to esti-
mate cardiac output during coronary artery surgery. Anaesthesia.
2012;67:377–83.
23. Fischer MO, Avram R, Carjaliu I, Massetti M, Gerard JL, Hanouz
JL, Fellahi JL. Non-invasive continuous arterial pressure and
cardiac index monitoring with Nexfin after cardiac surgery. Br J
Anaesth. 2012;109:514–21.
24. Fischer MO, Coucoravas J, Truong J, Zhu L, Gerard JL, Hanouz
JL, Fellahi JL. Assessment of changes in cardiac index and fluid
responsiveness: a comparison of Nexfin and transpulmonary
thermodilution. Acta Anaesthesiol Scand. 2013;57:704–12.
25. Hofhuizen C, Lansdorp B, van der Hoeven JG, Scheffer GJ,
Lemson J. Validation of noninvasive pulse contour cardiac output
using finger arterial pressure in cardiac surgery patients requiring
fluid therapy. J Crit Care. 2014;29:161–5.
26. Bubenek-Turconi SI, Craciun M, Miclea I, Perel A. Noninvasive
continuous cardiac output by the Nexfin before and after preload-
modifying maneuvers: a comparison with intermittent thermodi-
lution cardiac output. Anesth Analg. 2013;117:366–72.
27. Ameloot K, Van De Vijver K, Broch O, Van Regenmortel N, De
Laet I, Schoonheydt K, Dits H, Bein B, Malbrain ML. Nexfin
noninvasive continuous hemodynamic monitoring: validation
against continuous pulse contour and intermittent transpulmonary
thermodilution derived cardiac output in critically ill patients.
Scientific World J. 2013;2013:519080.
28. Monnet X, Picard F, Lidzborski E, Mesnil M, Duranteau J,
Richard C, Teboul JL. The estimation of cardiac output by the
Nexfin device is of poor reliability for tracking the effects of a
fluid challenge. Crit Care. 2012;16:R212.
123