E BRIEF REPORT

Personalized Versus Protocolized Fluid Management

Using Noninvasive Hemodynamic Monitoring
(Clearsight System) in Patients Undergoing
Moderate-Risk Abdominal Surgery
Alexandre Joosten, MD,* Shalini Raj Lawrence, MD,* Alexandra Colesnicenco, MD,*
Sean Coeckelenbergh, MD,* Jean Louis Vincent, MD, PhD,†
Philippe Van der Linden, MD, PhD,‡ Maxime Cannesson, MD, PhD,§ and Joseph Rinehart, MD∥

Advances in noninvasive hemodynamic monitoring systems allow delivery of goal-directed fluid

therapy and could therefore be used in less-invasive surgical procedures. In this randomized
controlled trial, we compared closed-loop–assisted goal-directed fluid therapy using a noninva-
sive cardiac output (Clearsight system) monitor (personalized approach) to a protocolized fluid
therapy approach in 40 patients undergoing moderate-risk laparoscopic abdominal surgery.
Cardiac output and stroke volume variations were not significantly different in both groups and
remained within predefined target values >90% of the study time. Personalized fluid therapy
does not seem to offer any hemodynamic advantage over a protocolized approach in this popu-
lation.  (Anesth Analg XXX;XXX:00–00)

C
onsiderable evidence indicates that goal-directed
fluid therapy (GDFT) can improve patient outcomes
after high-risk surgery.1–6 However, there is still debate
regarding the benefit of GDFT strategies in moderate-risk
surgery. Most GDFT protocols require invasive monitoring
devices with insertion of an arterial catheter, and substan-
tial effort is needed to ensure that the protocols are applied
correctly. Recently, we demonstrated that a closed-loop fluid
management system linked to a noninvasive cardiac out-
put (CO) monitor was associated with a high rate of GDFT
protocol compliance in patients undergoing moderate-risk
surgery.7 Among the available noninvasive hemodynamic
monitors, the Clearsight system (Edwards Lifesciences,
Irvine, CA) has the advantage that it directly measures flow
variables, in contrast with the Masimo system (Irvine, CA),
which only provides the pleth variability index.
A study by Stens et al8 recently reported no additional
value of advanced noninvasive CO monitoring to conven-
tional hemodynamic monitoring with regard to postopera-
tive complications in patients undergoing moderate-risk
abdominal surgery. However, from the data presented, it
was not possible to determine whether hemodynamics was
similarly maintained in both groups. Thus, we sought to
further test GDFT strategies in a prospective randomized
controlled trial in moderate-risk surgical patients target-
ing specific predefined hemodynamic end points. We com-
pared a personalized approach (closed-loop–assisted GDFT
approach guided by a noninvasive CO monitor) and a pro-
tocolized fluid therapy approach and hypothesized that
there would be no significant differences between groups.
Consistent with previous work,7,9 our primary outcome was
time spent with either a cardiac index (CI) ≥2.5 L/min/m2
or a stroke volume variation (SVV) <13%.

From the *Department of Anesthesiology, Cliniques Universitaires de
Bruxelles (CUB) Erasme, Université Libre de Bruxelles, Brussels, Belgium;
†Department of Intensive Care, CUB Erasme, Université Libre de Bruxelles,
Brussels, Belgium; ‡Department of Anesthesiology, Centre Hospitalo-
Universitaire (CHU) Brugmann, Université Libre de Bruxelles, Brussels,
Belgium; §Department of Anesthesiology and Perioperative Medicine,
University of California Los Angeles, David Geffen School of Medicine, Los
Angeles, California; and ∥Department of Anesthesiology and Perioperative
Care, University of California Irvine, Orange, California.
Accepted for publication May 8, 2018.
Funding: Departmental.
Conflicts of Interest: See Disclosures at the end of the article.
Supplemental digital content is available for this article. Direct URL citations
appear in the printed text and are provided in the HTML and PDF versions of
this article on the journal’s website (www.anesthesia-analgesia.org).
Trial Registry number: ClinicalTrials.gov (NCT03039946).
Reprints will not be available from the authors.
Address correspondence to Alexandre Joosten, MD, Department of
Anesthesiology, Hospital Erasme, 808 Rt de Lennik, 1070 Brussels, Bel-
­ normal value), preoperative renal insufficiency (serum cre-
gium. Address e-mail to Alexandre.Joosten@erasme.ulb.ac.be or joosten-­
alexandre@hottmail.com.
Copyright © 2018 International Anesthesia Research Society
DOI: 10.1213/ANE.0000000000003553
METHODS
The study was approved by the Ethics Committee of Erasme
Hospital (No: P2016/526), and written informed consent
was obtained from all subjects participating in the trial. The
trial was registered before patient enrollment at clinicaltri-
als.gov (NCT03039946; principal investigator: A.J.; date of
registration: February 1, 2017).
Adult low- to moderate-risk patients who did not
require invasive arterial pressure monitoring and were
listed for moderate-risk surgery (including only elective
laparoscopic–robotic colorectal, gynecological, or urologi-
cal procedures) were included. Exclusion criteria included
<18 years of age, an American Society of Anesthesiologists
score >3, a left ventricular ejection fraction <30%, signifi-
cant cardiac arrhythmias or aortic regurgitation, coagula-
tion disorders (activated partial thromboplastin time >1.5
atinine >2 mg/dL, oliguria, anuria, or hemodialysis), emer-
gency surgery, preoperative infection, and participation in
another trial.

XXX 2018 • Volume XXX • Number XXX www.anesthesia-analgesia.org

1
Copyright © 2018 International Anesthesia Research Society. Unauthorized reproduction of this article is prohibited.
 EE Brief Report

Randomization, Blinding, and Data Collection
Randomization was conducted using Internet-based soft-
ware (http://www.randomization.com), and the random-
ization numbers were placed in sealed envelopes containing
the group assignments. The list of codes was kept in a des-
ignated area by the departmental secretary. Data were col-
lected in the postanesthesia care unit by the nurses in charge
of the patient who were blinded to the group allocation and
on the ward (record of postoperative data and complica-
tions) by the investigators (A.J., S.R.L., A.C., or S.C.).
sufentanil boluses at the discretion of the anesthesia team.
Any necessary adjustments were made at the discretion
of the primary anesthesia providers to keep the bispectral
index level between 40 and 60. Patients were mechanically
ventilated using a volume control mode with a tidal volume
of 8 mL/kg of ideal body weight and a positive end-expi-
ratory pressure of 5 cm H2O using the Zeus Infinity C700
Anesthesia workstation (Dräger Medical GmbH, Lübeck,
Germany). Respiratory rate was adjusted to achieve end-
tidal CO2 between 32 and 36 mm Hg.

Anesthesia Procedures Fluid Therapy

All included patients were allowed to take solid foods until 6
hours before surgery and fluids until 2 hours before surgery.
None of the patients had bowel preparation. In both groups,
premedication consisted of 0.5 mg of alprazolam given on
the morning of surgery. Standard monitoring included
a 5-lead electrocardiogram, pulse oximetry, noninvasive
blood pressure, inspiratory and expiratory gas concentra-
tions, urine output, and processed electroencephalography
monitoring (BIS monitor; Covidien, Dublin, Ireland).
In addition to standard monitoring, all patients had a
noninvasive CO monitor (Clearsight; Edwards Lifesciences)
positioned before anesthesia induction. Anesthesia was
induced with sufentanil (0.2 μg/kg), propofol (2 mg/kg),
and rocuronium (0.6 mg/kg). Anesthesia was maintained
with sevoflurane or desflurane depending on preference and
All patients received 100 mL of isotonic saline during the
induction of anesthesia. In the protocolized group, patients
received a baseline infusion of 4 mL/kg/h of isotonic bal-
anced crystalloid solution (PlasmaLyte; Baxter, Lessines,
Belgium), and the anesthesiologist was blinded to the infor-
mation provided by the Clearsight system. Providers were
permitted to administer additional fluids as deemed neces-
sary. In the personalized group, no baseline fluid therapy
was used. The closed-loop system used in the personal-
ized group has been described extensively in our previous
publications.10–14 The Clearsight monitor was linked to the
closed-loop system (Figure) that delivered 100-mL boluses
of isotonic balanced crystalloid (PlasmaLyte) through a
Sapphire infusion pump (Q-Core, Tel Aviv, Israel) based
on the incoming data. The anesthesiologist in charge of

Figure. Schematic representation of our closed-loop system linked to the noninvasive hemodynamic monitoring device (Clearsight System;
Edwards Lifesciences, Irvine, CA). The closed-loop system was connected to the serial output port of the EV1000 monitor for real-time capture
of data. A Q-Core Sapphire Pump was used by the closed loop to deliver 100-mL boluses of balanced crystalloid solution (PlasmaLyte; Baxter,
Lessines, Belgium). The Sapphire Pump was controlled by the closed-loop system using the software provided by the Q-Core via serial connec-
tion (Commands Server R.00). In brief, the system monitors stroke volume, stroke volume variation, heart rate, and mean arterial pressure,
and it uses this information to optimize stroke volume. The controller uses a model layer to formulate a predicted response to a fluid bolus
and an adaptive layer for bolus-based error correcting for changes induced by surgical and anesthetic conditions. The final action to be taken
by the controller is then determined by a rule-based layer.

2   
www.anesthesia-analgesia.org ANESTHESIA & ANALGESIA
Copyright © 2018 International Anesthesia Research Society. Unauthorized reproduction of this article is prohibited.
 Personalized Versus Protocolized Fluid Management

the patient could interact with the automated system and
deliver or halt a fluid bolus manually if necessary; he/she
could also deliver additional fluid outside of the closed loop
(rescue fluid therapy). In both groups, anesthesiologists
could compensate intraoperative blood loss using a 1-to-1
ratio of colloid (6% hydroxyethyl starch 130/0.4; Fresenius
Kabi GmBH, Bad Homburg, Germany) if deemed necessary.
Boluses of ephedrine were allowed when severe arterial
hypotension (mean arterial pressure, <65 mm Hg) persisted
after appropriate fluid administration.
Outcome Variables
The primary outcome was the percentage of intraopera-
tive time spent within hemodynamic targets, as defined
by a CI ≥2.5 L/min/m2 and/or an SVV <13%. Secondary
outcomes were the times-in-target for the separate CI and
SVV subcomponents of the primary outcome, total volume
of fluid administered, fluid balance, incidence of postopera-
tive complications within 30 days, and lengths of stay in the
postanesthesia care unit and in the hospital.
Statistical Analysis
The assumption of normality for the primary outcome,
assessed using the Shapiro-Wilk test, was not fulfilled, so
continuous variables are shown as medians (25th–75th per-
centiles). Categorical variables are reported as counts and
percentages. Group comparisons for continuous variables
were made using a Mann-Whitney U test, and for categori-
cal variables using a χ2 test, or Fisher exact test when indi-
vidual cell counts were <5. Statistical significance was set
at P < .05. Statistical tests were performed with SPSS (IBM
Corp, Armonk, NY) or R statistics (www.r-project.org).
Previous internal data showed that patients managed
with the closed-loop system spent approximately 90% ±
10% of intraoperative time within the hemodynamic tar-
gets. Assuming similar performance of the closed loop in
this study, we determined a priori that a 10% difference
between groups would be clinically significant, and the
study was powered to this end. Monte Carlo power analy-
sis using a normal sampling distribution showed that to
detect 10% difference in time-in-target over the proto-
colized group with a power of 0.8 and a significance level of
0.05, 20 patients per group would give a power of approxi-
mately 0.85.
RESULTS
Forty consecutive patients were recruited. One patient was
subsequently excluded because of an anaphylactic reac-
tion during induction, which resulted in cancellation of the
surgery. Demographic data for the 2 groups are shown in
Supplemental Digital Content 1, Table 1, http://links.lww.
com/AA/C438.
There was no statistically significant difference in
time-in-target between groups: 97.5% (93.5%–100%) of
intraoperative time for the protocolized group and 99.9%
(97.2%–100%) for the personalized group (P = .10). The
time-in-target measures for the SVV and CI components
were 91% (75%–95%) and 94% (77%–98%) for SVV (P = .19)
and 92% (49%–100%) and 98% (94%–100%) for CI (P = .30)
for the protocolized and personalized groups, respectively
(Table). Intraoperative and postoperative heart rate and
mean arterial pressure were not statistically significantly
different in the 2 groups (Table). The volumes of crystal-
loid administered intraoperatively and use of vasopres-
sors were not significantly different. Intraoperative fluid
balance was significantly lower in the personalized than
in the protocolized group (Supplemental Digital Content
2, Table 2, http://links.lww.com/AA/C439). No patient
in either group required rescue fluid boluses. All patients
were extubated in the operating room at the end of the sur-
gery. The incidence of major and minor complications was
not significantly different between groups. There was no

Table.  Hemodynamic Variables and Patient Outcomes

Protocolized Group Personalized Group
(N = 19) (N = 20) P Value
Hemodynamic variables
  Intraoperative heart rate (beats/min) 73 (67–81) 73 (66–77) .550
  Intraoperative mean arterial pressure (mm Hg) 85 (76–91) 94 (80–97) .134
  Intraoperative stroke volume index (mL/m2) 43 (37–48) 45 (41–55) .127
  Intraoperative cardiac index (L/min/m2) 3.0 (2.5–3.8) 3.4 (2.9–3.9) .296
  Intraoperative stroke volume variation (%) 9 (7–11) 9 (7–10) .461
  % case time with stroke volume variation <13% 91 (75–95) 94 (77–98) .194
  % case time with cardiac index ≥2.5 L/min/m2 92 (49–100) 98 (94–100) .296
  % case time with mean arterial pressure <65 mm Hg 5 (1–22) 3 (1–9) .322
  Preoperative blood lactate (mEq/L) 0.9 (0.6–1.1) 0.9 (0.7–1.4) .170
  Postoperative blood lactate (mEq/L) 1.4 (1.1–1.8) 1.5 (0.8–1.9) .784
  Postoperative heart rate (beats/min) 81 (70–86) 79 (69–88) .835
  Postoperative mean arterial pressure (mm Hg) 91 (80–99) 87 (74–93) .365
Patient outcomes
  Patients with any major complications (%) 3 (16) 3 (15) .946
  Patients with any minor complications (%) 5 (26) 2 (10) .184
  Length of stay postanesthesia care unit (h) 3 (2–4) 3 (2–4) .708
  Length of stay hospital (d) 3 (2–5) 3 (2–6) .478
Data are expressed as median (25th–75th percentiles). P-values are by Mann-Whitney U test for continuous data and by χ2 test for % measures, or by Fisher exact
test when cell counts are <5. Intraoperative variables were recorded by the noninvasive cardiac output monitor at 20-s intervals and averaged. Postoperative
hemodynamic variables are an average of the variables recorded at 4 different time points in the postoperative period (arrival in postanesthesia care unit, +1 h,
+2 h postarrival, and +3 h postarrival).

XXX 2018 • Volume XXX • Number XXX www.anesthesia-analgesia.org

3
Copyright © 2018 International Anesthesia Research Society. Unauthorized reproduction of this article is prohibited.
 EE Brief Report
significant difference between groups in postanesthesia care
unit or hospital lengths of stay (Table; Supplemental Digital
Content 3, Table 3, http://links.lww.com/AA/C440). The
median hospital length of stay was 3 days in both groups.
DISCUSSION
Our results confirmed that a personalized approach, that
is, a closed-loop–assisted GDFT guided by a noninvasive
CO monitor, offered no significant advantage in main-
taining CI or SVV in patients undergoing moderate-risk
abdominal surgery compared to a protocolized approach.
This observation is consistent with the results from the
study by Stens et al8 mentioned earlier. However, adding
to the results from that study, we also recorded advanced
hemodynamic variables in the 2 groups and observed sim-
ilar values, although the anesthesiologists in charge of the
patients in the protocolized group were blinded to these
variables.
A major strength of this study was the use of the closed-
loop–assisted system, which ensured high protocol com-
pliance in the personalized group.7,9 Limitations include
the relatively small number of patients, making the study
underpowered to assess the impact of the fluid therapy
approach on the incidence of postoperative complications.
However, based on the incidence of major complications
in the 2 groups, a Monte Carlo simulation of binomial
proportion tests with an α of .05 suggests that around
35,000 patients per group would be needed to have 80%
power to detect a significant difference between groups.
Baseline maintenance fluid therapy of 4–8 mL/kg/h is usu-
ally recommended for GDFT, so we chose to administer
4 mL/kg/h for our patients undergoing moderate-risk lap-
aroscopic surgery; lower maintenance rates are considered
too restrictive by some experts.15 Finally, the results of our
study should not be extrapolated to surgical procedures
with significant blood loss and/or major fluid shifts.
CONCLUSIONS
Under our study conditions in patients undergoing moder-
ate-risk abdominal surgery, a personalized approach using
closed-loop–assisted GDFT (without baseline crystalloid
infusion) does not seem to have any advantage over a pro-
tocolized approach in maintaining CI or SVV within pre-
defined targets. E
DISCLOSURES
Name: Alexandre Joosten, MD.
Contribution: This author helped design the study, recruit the
patients, collect the data, and draft the final manuscript.
Conflicts of Interest: A. Joosten is a consultant for Edwards
Lifesciences (Irvine, CA).
Name: Shalini Raj Lawrence, MD.
Contribution: This author helped recruit the patients, collect the
data, and draft the final manuscript.
Conflicts of Interest: None.
Name: Alexandra Colesnicenco, MD.
Contribution: This author helped recruit the patients, collect the
data, and draft the final manuscript.
Conflicts of Interest: None.
Name: Sean Coeckelenbergh, MD.
Contribution: This author helped collect and analyze the data, and
draft the final manuscript.
Conflicts of Interest: None.
4   
www.anesthesia-analgesia.org
Copyright © 2018 International Anesthesia Research Society. Unauthorized reproduction of this article is prohibited.
Name: Jean Louis Vincent, MD, PhD.
Contribution: This author helped analyze the data and draft the
final manuscript.
Conflicts of Interest: None.
Name: Philippe Van der Linden, MD, PhD.
Contribution: This author helped analyze the data and draft the
final manuscript.
Conflicts of Interest: P. Van der Linden has received, within the
past 5 years, fees for lectures and consultancies from Fresenius Kabi
GmbH and Janssen-Cilag SA, Belgium.
Name: Maxime Cannesson, MD, PhD.
Contribution: This author helped design the closed-loop system,
analyze the data, and draft the final manuscript.
Conflicts of Interest: M. Cannesson is a consultant for Edwards
Lifesciences (Irvine, CA), Covidien (Boulder, CO), and Masimo
Corp (Irvine, CA). He is a cofounder of Sironis and owns a patent
on closed-loop fluid management and hemodynamic optimization.
Name: Joseph Rinehart, MD.
Contribution: This author helped design the closed-loop system,
analyze the data, and draft the final manuscript.
Conflicts of Interest: J. Rinehart is a consultant for Edwards
Lifesciences (Irvine, CA). He is a cofounder of Sironis and owns
a patent on closed-loop fluid management and hemodynamic
optimization.
This manuscript was handled by: Tong J. Gan, MD.
REFERENCES
1. Michard F, Giglio MT, Brienza N. Perioperative goal-directed
therapy with uncalibrated pulse contour methods: impact on
fluid management and postoperative outcome. Br J Anaesth.
2017;119:22–30.
2. Sun Y, Chai F, Pan C, Romeiser JL, Gan TJ. Effect of periop-
erative goal-directed hemodynamic therapy on postoperative
recovery following major abdominal surgery: a systematic
review and meta-analysis of randomized controlled trials. Crit
Care. 2017;21:141.
3. Cannesson M, Ramsingh D, Rinehart J, et al. Perioperative
goal-directed therapy and postoperative outcomes in
patients undergoing high-risk abdominal surgery: a histor-
ical-prospective, comparative effectiveness study. Crit Care.
2015;19:261.
4. Pearse RM, Harrison DA, MacDonald N, et al; OPTIMISE
Study Group. Effect of a perioperative, cardiac output-guided
hemodynamic therapy algorithm on outcomes following major
gastrointestinal surgery: a randomized clinical trial and sys-
tematic review. JAMA. 2014;311:2181–2190.
5. Kaufmann KB, Stein L, Bogatyreva L, et al. Oesophageal
Doppler guided goal-directed haemodynamic therapy in tho-
racic surgery: a single centre randomized parallel-arm trial. Br J
Anaesth. 2017;118:852–861.
6. Malbouisson LMS, Silva JM Jr, Carmona MJC, et al. A pragmatic
multi-center trial of goal-directed fluid management based on
pulse pressure variation monitoring during high-risk surgery.
BMC Anesthesiol. 2017;17:70.
7. Joosten A, Huynh T, Suehiro K, Canales C, Cannesson M,
Rinehart J. Goal-directed fluid therapy with closed-loop
assistance during moderate risk surgery using noninva-
sive cardiac output monitoring: a pilot study. Br J Anaesth.
2015;114:886–892.
8. Stens J, Hering JP, van der Hoeven CWP, et al. The added value
of cardiac index and pulse pressure variation monitoring to
mean arterial pressure-guided volume therapy in moderate-
risk abdominal surgery (COGUIDE): a pragmatic multicentre
randomised controlled trial. Anaesthesia. 2017;72:1078–1087.
9. Rinehart J, Le Manach Y, Douiri H, et al. First closed-loop goal
directed fluid therapy during surgery: a pilot study. Ann Fr
Anesth Reanim. 2014;33:e35–e41.
10. Rinehart J, Alexander B, Le Manach Y, et al. Evaluation of
a novel closed-loop fluid-administration system based on
dynamic predictors of fluid responsiveness: an in silico simula-
tion study. Crit Care. 2011;15:R278.
11. Rinehart J, Chung E, Canales C, Cannesson M. Intraoperative
stroke volume optimization using stroke volume, arterial
pressure, and heart rate: closed-loop (learning intravenous
ANESTHESIA & ANALGESIA
 Personalized Versus Protocolized Fluid Management
resuscitator) versus anesthesiologists. J Cardiothorac Vasc
Anesth. 2012;26:933–939.
12. Rinehart J, Lee C, Canales C, Kong A, Kain Z, Cannesson M.
Closed-loop fluid administration compared to anesthesiologist
management for hemodynamic optimization and resuscita-
tion during surgery: an in vivo study. Anesth Analg. 2013;117:
1119–1129.
13. Rinehart J, Lee C, Cannesson M, Dumont G. Closed-loop fluid
resuscitation: robustness against weight and cardiac contractil-
ity variations. Anesth Analg. 2013;117:1110–1118.
XXX 2018 • Volume XXX • Number XXX www.anesthesia-analgesia.org
5
Copyright © 2018 International Anesthesia Research Society. Unauthorized reproduction of this article is prohibited.
14. Rinehart J, Liu N, Alexander B, Cannesson M. Review arti-
cle: closed-loop systems in anesthesia: is there a potential for
closed-loop fluid management and hemodynamic optimiza-
tion? Anesth Analg. 2012;114:130–143.
15. Thiele RH, Raghunathan K, Brudney CS, et al; Perioperative
Quality Initiative (POQI) I Workgroup. American Society
for Enhanced Recovery (ASER) and Perioperative Quality
Initiative (POQI) joint consensus statement on perioperative
fluid management within an enhanced recovery pathway for
colorectal surgery. Perioper Med (Lond). 2016;5:24.