Boulet et al.

Critical Care (2024) 28:429 Critical Care

https://doi.org/10.1186/s13054-024-05155-z

RESEARCH Open Access

Impact on fluid balance of an optimized

restrictive strategy targeting non‑resuscitative
fluids in intensive care patients with septic
shock: a single‑blind, multicenter, randomized,
controlled, pilot study
Nicolas Boulet1,2, Jean‑Pierre Quenot3,4,5, Chris Serrand6, Nadiejda Antier2,7, Sylvain Garnier2, Aurèle Buzancais2,
Laurent Muller1,2, Claire Roger1,2, Jean‑Yves Lefrant1 and Saber Davide Barbar1,2*

Abstract
Background In septic shock, the classic fluid resuscitation strategy can lead to a potentially harmful positive
fluid balance. This multicenter, randomized, single-blind, parallel, controlled pilot study assessed the effectiveness
of a restrictive fluid strategy aiming to limit daily volume.
Methods Patients 18–85 years’ old admitted to the ICU department of three French hospitals were eligible for inclu-
sion if they had septic shock and were in the first 24 h of vasopressor infusion. Exclusion criteria were acute kidney
injury requiring renal replacement therapy, end stage chronic kidney disease, and severe malnutrition. Patients
were electronically randomized 1:1 to either an optimized fluid restriction (reducing fluid intake as much as pos-
sible in terms of maintenance fluids and fluids for drug dilution during the first 7 days) or standard fluid strategy.
The primary outcome was cumulative fluid balance (ml/kg) in the first 5 days. Patients and statisticians were blinded
to group arm, but not clinicians.
Results Between September 2021 and February 2023, 1201 patients were screened and 50 included, with two
in the control group withdrawing, thus 48 patients were analyzed (24 in each group). In the first 5 days, the optimized
restrictive strategy and control groups received 89.7 (IQR 35; 128.9) and 114.3 (IQR 78.8; 168.5) ml/kg of fluid, respec-
tively (mean difference: 35.9 ml/kg [0.0; 71.8], p = 0.0506). After 5 days, the median cumulative fluid balance was 6.9
(IQR − 13.7; 52.1) and 35.0 (IQR − 7.9; 40.2) ml/kg in the optimized restrictive strategy and control groups, respectively
(absolute difference 13.2 [95%CI − 15.2; 41.6], p = 0.42). After 28 days, mortality and the numbers of days alive with-
out life support were similar between groups. The main adverse events were severe hypernatremia in 1 and 2 patients
in the fluid restriction strategy and control groups, respectively, and acute kidney injury KDIGO 3 in 4 and 7 patients
in the fluid restriction strategy and control groups, respectively.
Conclusions In ICU patients with septic shock, an optimized restrictive fluid strategy targeting hidden fluid intakes
did not reduce the overall fluid balance at day 5.
Trial registration ClinicalTrials.gov identifier NCT04947904, registered on 1 July 2021.

*Correspondence:
Saber Davide Barbar
[email protected]
Full list of author information is available at the end of the article

© The Author(s) 2024. Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0
International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long
as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if
you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or
parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated
otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To
view a copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by-​nc-​nd/4.​0/.
Boulet et al. Critical Care (2024) 28:429 Page 2 of 13

Keywords Fluid balance, Septic shock, Intensive care unit, Restrictive strategy

Graphical Abstract

Background of mortality [2–5], in particular if sepsis is combined

Sepsis is the cause of nearly 30% of hospitalizations with acute respiratory distress syndrome [3, 4] and
in the intensive care unit (ICU), with a mortality rate acute renal failure [2]. Patients with septic shock can
about 40% 28 days after admission [1]. Fluid resusci- receive large volumes of fluids, of up to 25 L in three
tation therapy is a cornerstone of the initial manage- days [6, 7]. Interestingly, the volume of resuscita-
ment of patients with septic shock but can lead to tion fluids constitutes only one-third of the total fluid
fluid overload. A positive fluid balance in the first few intake received in the acute phase of septic shock [8].
days after admission to an intensive care unit (ICU) is Maintenance/replacement fluids, called “creep fluids”
related with negative outcome and an increased risk or “hidden fluid intakes”, include the fluids to perfuse
Boulet et al. Critical Care (2024) 28:429 Page 3 of 13

electrolytes, the small volumes to keep venous lines therapy). Septic shock was defined as suspected or con-
open (saline or glucose 5%), fluids for drug delivery, firmed infection, a plasma lactate level of ≥ 2 mmol/L and
and parenteral and enteral nutrition. One retrospec- ongoing infusion of a vasopressor agent [1]. Exclusion
tive study including 14,654 ICU patients reported that criteria were acute kidney injury requiring renal replace-
maintenance/replacement fluids accounted for over ment therapy within the next 24 h, end stage chronic
55% of fluid intake during the first five days in the kidney disease, and severe malnutrition with body mass
ICU, with significant sodium and chloride loads [8]. index < 18. Detailed inclusion and exclusion criteria are
Another retrospective study performed in 10 ICUs and presented in Supplementary Table S1.
involving 400 patients, confirmed that maintenance
and dilution fluids accounted for more than 50% of Randomization and blinding
fluids infused during the first three ICU days [9]. Fur- Randomization was performed using a centralized, com-
thermore, the fluid balance in the first three days was puter-generated allocation sequence. Eligible patients
independently associated with 30-day mortality. were randomly assigned in a 1:1 ratio, (in permuted
Several randomized clinical trials (RCT) have inves- blocks of 2), to receive the optimized restrictive strategy
tigated the benefits of different fluid restriction strate- or standard fluid therapy (control group).
gies, targeting uniquely or mostly resuscitative fluids. Treatment group assignments were not masked for cli-
When applied in the acute phase of the management nicians or investigators but were concealed from patients
of patients with septic shock, these strategies resulted and the statisticians.
in small differences in fluid intake and no difference in
survival [10–13]. Interventions
Studies are needed focusing on a restrictive strat- All patients were treated according to the recommenda-
egy targeting non-resuscitative fluid input to opti- tions of the Surviving Sepsis Campaign [15]. During the
mize fluid balance in patients with septic shock. We first 7 days of septic shock, fluid infusion was given fol-
designed a pilot randomized controlled trial (OPTIF- lowing an optimized restrictive strategy or standard fluid
LUID) studying the impact of an optimized restrictive therapy (control group). To calculate the dose of fluids
strategy targeted at reducing all fluid infusions other administered, we always used the patient’s weight on
than resuscitative fluids. The primary objective was to admission to ICU. Trial staff (physicians, clinical phar-
assess the fluid balance after the first 5 days of resus- macists and nurses) were trained before inclusion began.
citation in ICU patients with septic shock. Secondary The coordinating investigator and the clinical pharmacist
outcomes were fluid balance, body weight variation, were available throughout the inclusion period to help
use of diuretics, death, organ failure free days and the investigators comply with the protocol.
length of stay.
Optimized restrictive strategy (experimental group)
A restrictive fluid protocol was applied upon inclusion on
Methods
Day 1 and continued for the first 7 days of hospitaliza-
Trial design and oversight
tion in the ICU. Clinicians were instructed to reduce fluid
OPTIFLUID is a multicenter, single-blind, rand-
intake as much as possible in terms of maintenance fluids,
omized, parallel, controlled, pilot trial. Patients were
drug dilution and artificial nutrition. The management of
recruited between September 20, 2021, and February
resuscitative fluids, defined as fluids prescribed to opti-
25, 2023 in ICUs in the university hospitals of Nîmes
mize blood volume in the treatment of shock, was not
and Dijon, and the general hospital of Alès, France.
modified by the protocol, and was left to the judgement
The study was approved by the Comité de Protection
of clinicians according to their usual monitoring strategy
des Personnes Sud-Méditerranée II, France (num-
(cardiac ultrasound, pulse index continuous cardiac out-
ber 2020-A01952-37). Written informed consent was
put, right heart catheterization, central venous pressure,
obtained from patients or their legal surrogates either
etc.). For drug dilutions, the clinicians consulted charts
before randomization or as soon as possible thereafter
(Supplementary Table S2) showing the maximum pos-
[14]. The trial was funded by Nîmes University Hospi-
sible concentrations for the drugs most frequently used
tal, France.
in ICU, particularly antibiotics, based on literature data
[16]. For artificial nutrition, the intravenous route was
Patients
discouraged before Day 5 [17]. For enteral nutrition, the
Patients between 18 and 85 years’ old were eligible if
protocol provided for the use of a hypercaloric (2 kcal/
they were admitted to the ICU in the early phase of
ml) high protein product, with a total caloric objective
septic shock (within 24 h after the start of vasopressor
Boulet et al. Critical Care (2024) 28:429 Page 4 of 13

of 20 kcal/kg per day in the acute phase and 30 kcal/kg restrictive strategy would lead to a relative 40% reduc-
per day after stabilization [18]. Intravenous fluids other tion in fluid balance, conferring a significant clinical
than those needed to dilute drugs and electrolytes were impact. Assuming a mean fluid balance of 70 mL/kg in
prohibited, and only 2 ml/hour of saline or glucose were the control group [6–8], 40 mL/kg in the experimental
authorized to maintain veins. A clinical pharmacist inter- group and a standard deviation of 35 mL/kg, we esti-
vened daily to check adherence to the fluid restriction mated that 44 patients (22 per group) were needed to
protocol. If the restrictive strategy induced hypovolemia guarantee 80% statistical power with a two-sided alpha
and/or biological abnormalities (sodium, potassium, risk of 5%. This number was increased by 10% to take
chloride, etc.), the treating physician could increase fluid into account any deaths between Day 1 and Day 5. The
therapy in accordance with the protocol’s recommenda- total number of subjects required was therefore 50
tions. The fluid restriction strategy did not involve the patients (25 per group).
systematic administration of diuretics. The primary endpoint, fluid balance at Day 5 (ml/kg),
was compared between groups using a Mann–Whitney
test via intention-to-treat analysis. A second per-pro-
Standard fluid strategy (control group) tocol analysis was performed, where only patients who
The choice of the volume of fluid intake other than resus- had been depleted for at least 5 days were considered,
citative fluids was left to the physician’s discretion, based as well as a sensitivity analysis in which we excluded
on usual practices: usually between 500 and 2000 ml/ patients with bleeding (defined as more than 2 units of
day of maintenance volumes. The dilutions of drugs were blood transfusion in the same day). Categorical variables
those officially planned at the time of marketing approval, are expressed as numbers and percentages; continuous
enteral artificial nutrition (isocaloric, high protein) was variables are expressed as means and standard deviations
initiated as soon as possible, with the same rules for par- or medians and interquartile ranges (IQR) as appropri-
enteral nutrition if the caloric target could not be reached ate. Quantitative variables were compared between two
with enteral nutrition. groups using a Student’s t test if the variable had a nor-
Diuretics were authorized in both groups at the judge- mal distribution, or a Wilcoxon Mann Whitney other-
ment of the clinicians according to the needs of the wise. The mean difference between the two groups was
patients, but were not recommended as long as patients also calculated with the 95% confidence interval. For cat-
had high dose vasopressors. egorical variables, either the ­Chi2 test or Fisher’s exact
test were used and the percentage difference between
Outcomes
the experimental and control group was also calculated,
The primary outcome was fluid balance in the first as was the associated relative risk with the 95% confi-
5 days. Secondary outcomes were fluid balance at Day 3, dence interval. The length of ICU and hospital stay, and
5 and 7; body weight variation at Day 5 and 7; death from the number of days free of mechanical ventilation, vaso-
any cause at Day 7 and 28; the number of days alive with- pressors, and renal-replacement therapy are expressed as
out life support and without organ failure (SOFA = 0) at medians and IQR and were compared using the Mann–
Day 28; the number of days alive without vasopressors, Whitney test. We evaluated safety by calculating the
mechanical ventilation and renal replacement therapy percentage of patients in the two groups presenting met-
at Day 28; the number of days alive outside the ICU at abolic, renal and nutritional adverse events; percentages
Day 28; the ICU and hospital length of stay; and the use were compared with the use of appropriate tests. The sig-
of diuretics (cumulative dose administered over 7 days). nificance level was set at 0.05 for all analyses.
Finally, in addition to unexpected adverse events, pre- Analyses were performed with SAS Enterprise Guide
defined adverse events during the entire ICU stay were software, version 7.1.
collected, specifically metabolic (mild and severe hyper-
natremia, hyponatremia), renal (hyperkalemia, Acute
Kidney Injury KDIGO2 and 3), and nutritional (hypogly- Study protocol
cemia, stage 3 pressure sores acquired in the ICU); cumu- The trial protocol, which includes the statistical analysis
lative insulin dose over 7 days) complications potentially plan, was registered on ClinicalTrials.gov with the identi-
related to the restrictive fluid strategy; norepinephrine fier NCT04947904.
doses and lactate levels over 7 days. For evaluation of patient safety, this trial has been
registered with France National Agency of Drug Safety
Statistical analysis (Agence Nationale de Sécurité du Médicament et des
Considering that "hidden" fluid intake constitutes Produits de Santé, ANSM).
55% of total intake by patients, we estimated that a
Boulet et al. Critical Care (2024) 28:429 Page 5 of 13

Fig. 1 Flow chart

Trial data were monitored at the sites by independent Fluid intervention

monitors and monitored centrally by staff at the coordi- The median cumulative volumes of fluids administered
nating centre. The first and the last author wrote the first in the first 5 days were 89.7 (IQR 35; 128.9) and 114.3
draft of the manuscript, which was reviewed by all the (78.8; 168.5) ml/kg in the optimized restrictive strategy
authors. Statistical analyses were performed by the trial and control group, respectively (mean difference 35.9
statistician in accordance with International Conference [95%CI − 0.0; 71.8], p = 0.0506) (Table 2). The 7-day
on Harmonisation Good Clinical Practice guidelines. The median cumulative volumes of fluids administered were
authors vouch for the accuracy and completeness of the 98.8 (IQR 35.0; 184.6) and 123.8 (IQR 78.8; 217.5) ml/kg
reported analyses and for the fidelity of the trial to the in the optimized restrictive strategy and control groups,
protocol. respectively (mean difference 37.5 [95%CI − 16.5; 91.5],
p = 0.17).
Concerning fluid administration, resuscitative fluid
Results volumes were similar in both groups at Day 5 and Day 7
Population (Table 2). For the non-resuscitative fluids, vein mainte-
Between September 2021 and February 2023, 1201 nance fluids were lower in the optimized restrictive strat-
patients were assessed for eligibility. Fifty patients were egy group at Day 5 (mean difference 20.0 ml/kg [95%CI
randomized, with two in the control group withdraw- 6.4; 33.6], p < 0.01) and Day 7 (mean difference 21.7 ml/
ing consent, leading to 24 analyzable patients per group kg [1.4; 42.1], p < 0.01), whereas other non-resuscitation
(Fig. 1). Patients had an average age of 68.3 ± 10.2 years fluids were similar between groups (Table 2 and Supple-
and an average SAPSII score of 56.5 ± 19.2. Patient char- mentary Table S3, S4 and S5).
acteristics at baseline are shown in Table 1. Patient char-
acteristics at baseline were well balanced between the two Protocol adherence
groups, except for serum creatinine, which was higher in There were no violations of fluid protocol in the opti-
the intervention group (Table 1). During the 28-day trial mized restrictive strategy group, but total fluid volume
period, patients in both intervention groups remained in at Day 5 and Day 7 was lower than expected in the con-
the ICU for a median of 5 days after randomization (IQR, trol group (Table 2).
3 to 10 days).
Boulet et al. Critical Care (2024) 28:429 Page 6 of 13

Table 1 Patient characteristics at baseline

Characteristics Optimized restrictive strategy group Control group
(N = 24) (N = 24)

Mean age (SD)—year 69.5 (± 8.8) 67.0 (± 11.5)

Male sex—no. (%) 12 (50.0) 13 (54.2)
Mean body weight (SD)—kg 67.7 (± 14.8) 73.4 (± 16.6)
Mean body mass index (SD) 24.5 (± 5.1) 26.0 (± 6.7)
Coexisting condition—no. (%)
Hematologic or metastatic cancer 7 (29.2) 8 (33.3)
Ischemic heart disease or heart failure 3 (12.5) 4 (16.7)
Chronic hypertension 14 (58.3) 11 (45.8)
Chronic kidney disease 1 (4.2) 2 (8.3)
Cirrhosis 2 (8.3) 0 (0)
Diabetes mellitus 8 (33.3) 5 (20.8)
Median time from ICU admission to randomization (IQR)—days 1.0 (0.0; 1.0) 1.0 (0.0; 1.0)
Mean SAPSII score (SD) 54.5 (± 21.9) 58.5 (± 16.4)
Referral department—no. (%)
Emergency department or prehospital 12 (50.0) 12 (50.0)
Hospital ward 12 (50.0) 12 (50.0)
Origin of infection—no. (%)
Gastrointestinal 7 (29.2) 8 (33.3)
Pulmonary 12 (50.0) 11 (45.8)
Urinary tract 1 (4.2) 4 (16.7)
Skin or soft tissue 1 (4.2) 1 (4.2)
Other 4 (16.7) 2 (8.3)
Blood values
Median highest plasma lactate (IQR)—mmol per liter 2.1 (1.4; 3.7) 2.3 (1.5; 3.0)
Median highest plasma creatinine (IQR)—µmol per liter 94.5 (67.5; 133.5) 170.5 (75.0; 231.0)
Mean serum sodium level (SD)—µmol per liter 137.0 (± 4.2) 137.6 (± 4.5)
Mean serum potassium level (SD)—µmol per liter 3.9 (± 0.7) 3.9 (± 0.5)
ICU interventions
Median highest dose of norepinephrine (IQR)—μg/kg/min 0.4 (0.2; 1) 0.5 (0.3; 0.8)
Median volume of IV fluid 24 h before randomization (IQR)—ml 30.3 (15.2; 37.5) 21.8 (9.9; 38.7)
Use of respiratory support—no. (%) 16 (66.7) 14 (58.3)
Median lowest PaO2/FiO2 ratio (IQR) 207 (143; 277) 216 (158; 269)

Primary outcome At Day 7, the median cumulative fluid balance was 12.6
At Day 5, the median cumulative fluid balance was 6.9 (IQR − 19.9; 58.7) and 26.4 (IQR 16.3; 39.8) ml/kg in the
(IQR − 13.7; 52.1) versus 35.0 (IQR − 7.9; 40.2) ml/ optimized restrictive strategy and control groups, respec-
kg in the optimized restrictive strategy and control tively (absolute difference 12.5 [95%CI − 24.2; 49.3]
groups, respectively (absolute difference 13.2 [95%CI p = 0.72). There was no significant difference in body
− 15.2; 41.6], p = 0.42) (Table 2, Fig. 2A). Sensitivity weight variation at Day 3, 5 and 7 (Fig. 2B). There was no
analysis excluding patients with hemorrhage requiring statistical difference in survival, organ failure free-days,
blood transfusions gave similar results (Table S6). or ICU and hospital length of stay (Table 3).

Secondary outcomes Adverse events

At Day 3, the median cumulative fluid balance was 9.8 We found no difference in the predefined adverse events,
(IQR − 8.0; 38.8) and 24.7 (IQR 2.2; 37.7) ml/kg in the metabolic disorders, renal function, or nutritional issues
optimized restrictive strategy and control group, respec- (Table 4). Mild hypernatremia (Na 146–150 mmol/L) was
tively (absolute difference 13.8 [95%CI − 6.9; 34.4], observed in 7 patients (29%) in the optimized restrictive
p = 0.19) (Table 3, Fig. 1A). strategy group and 10 (42%) in the control group, relative
Boulet et al. Critical Care (2024) 28:429 Page 7 of 13

Table 2 Median Cumulative Fluid Volumes in ICU during intervention

Fluid Volumes Optimized restrictive Control Group Mean difference p-value
strategy group

Median (IQR)—ml/Kg of admission body weight (N = 24) (N = 24)

Total fluid intake volume after 5 days 89.7 (35; 128.9) 114.3 (78.8; 168.5) 35.9 [0.0; 71.8] 0.0506
Total fluid intake volume after 7 days 98.8 (35.0; 184.6) 123.8 (78.8; 217.5) 37.5 [− 16.5; 91.5] 0.17
Intravenous fluid volume after 5 days
Resuscitative fluids
Crystalloids* 8.3 (0.0; 16.4) 7.5 (0.0; 15.0) 2.8 [− 5.1; 10.7] 0.80
Albumin 0.0 (0.0; 0.0) 0.0 (0.0; 2.7) 1.2 [− 1.4; 3.8] 0.21
Non-Resuscitative fluids
Transfusions 0.0 (0.0; 0.0) 0.0 (0.0; 1.4) 2.5 [− 0.6; 5.5] 0.38
Blood derivatives 0.0 (0.0; 0.0) 0.0 (0.0; 0.0) − 0.1 [− 1.7; 1.6] 0.96
Vein maintenance 7.2 (3.0; 12.5) 24.6 (14.9; 46.2) 20.0 [6.4; 33.6] < 0.001
Ion dilution 1.3 (0.3; 2.9) 1.4 (0.6; 3.2) 0.5 [− 1.1; 2.2] 0.56
Drug dilution 35.3 (12.2; 48.7) 37.0 (20.0; 59.4) 4.2 [− 11.4; 19.8] 0.59
Fluids to treat ­hypernatremia# 0.0 (0.0; 0.0) 0.0 (0.0; 0.0) 0.8 [− 0.6; 2.3] 0.54
Parenteral nutrition 0.0 (0.0; 0.0) 0.0 (0.0; 0.0) 0.8 [− 7.0; 8.7] 0.75
Enteral/oral fluid Volume after 5 days
Enteral nutrition 0.0 (0.0; 19.6) 0.0 (0.0; 21.0) − 0.1 [− 8.2; 7.9] 0.96
Water in nasogastric tube 2.0 (0.0; 11.0) 1.6 (0.0; 12.0) 2.4 [− 3.9; 8.8] 0.96
Oral fluid intake 2.3 (0.0; 3.4) 0.9 (0.0; 3.2) 0.8 [− 2.7; 4.3] 0.49
Intravenous fluid volume after 7 days
Resuscitative fluids
Crystalloids* 9.3 (0.0; 20.3) 8.4 (0.0; 15.0) 1.5 [− 6.7; 9.6] 0.93
Albumin 0.0 (0.0; 0.0) 0.0 (0.0; 2.7) 1.2 [− 1.4; 3.8] 0.21
Non-Resuscitative fluids
Transfusions 0.0 (0.0; 0.0) 0.0 (0.0; 7.6) 3.4 [− 0.3; 7.0] 0.19
Blood derivatives 0.0 (0.0; 0.0) 0.0 (0.0; 0.0) 21.7 [1.4; 42.1] 0.57
Vein maintenance 9.4 (3.0; 17.8) 26.2 (16.9; 49.0) 21.7 [1.4; 42.1] < 0.01
Ion dilution 1.4 (0.3; 3.3) 1.4 (0.6; 3.6) 0.5 [− 1.3; 2.2] 0.66
Drug dilution 39.4 (12.2; 58.5) 38.5 (20.0; 77.3) 4.7 [− 15.6; 25.1] 0.63
Fluids to treat ­hypernatremia# 0.0 (0.0; 0.0) 0.0 (0.0; 0.0) 0.7 [− 0.8; 2.3] 0.54
Parenteral nutrition 0.0 (0.0; 0.0) 0.0 (0.0; 0.0) 0.0 [− 9.3; 9.3] > 0.99
Enteral/oral fluid Volume after 7 days
Enteral nutrition 5.1 (0.0; 32.4) 0.0 (0.0; 22.9) − 1.4 [− 14.3; 11.4] 0.58
Water in nasogastric tube 2.3 (0.0; 13.1) 1.6 (0.0; 19.2) 4.7 [− 3.3; 12.7] 0.87
Oral fluid intake 2.5 (0.0; 5.8) 0.9 (0.0; 4.7) 0.2 [− 3.5; 3.9] 0.44
Total fluid output after 5 days 67.9 (33.0; 118.4) 95.5 (61.5; 131.5) 23.4 [− 6.5; 53.3] 0.12
Total fluid output after 7 days 87.0 (33.0; 167.1) 113.4 (61.5; 184.1) 25.7 [− 19.5; 70.9] 0.26
Fluid output after 5 days
Urine output 63.3 (28.8; 115.1) 92.8 (47.8; 125.2) 18.0 [− 11.4; 47.5] 0.29
Ultrafiltration on RRT​$ 0.0 (0.0; 0.0) 0.0 (0.0; 0.0) 6.2 [− 5.5; 18.0] 0.54
Drains 0.0 (0.0; 0.0) 0.0 (0.0; 0.0) 0.2 [− 2.0; 2.4] 0.93
Diarrhea 0.0 (0.0; 0.0) 0.0 (0.0; 0.0) − 0.3 [− 1.1; 0.6] 1.00
Nasogastric tube 0.0 (0.0; 5.9) 0.0 (0.0; 2.3) − 0.8 [− 5.5; 3.9] 0.76
Fluid output after 7 days
Urine output 64.3 (28.8; 149.7) 95.3 (47.8; 157.9) 17.8 [− 25.0; 60.6] 0.41
Ultrafiltration on RRT​$ 0.0 (0.0; 0.0) 0.0 (0.0; 0.0) 11.0 [− 8.2; 30.1] 0.95
Drains 0.0 (0.0; 0.6) 0.0 (0.0; 0.0) − 1.9 [− 6.6; 2.8] 0.82
Diarrhea 0.0 (0.0; 0.0) 0.0 (0.0; 0.0) 0.1 [− 1.6; 1.7] 0.61
Nasogastric tube 0.0 (0.0; 6.2) 0.0 (0.0; 2.3) − 1.2 [− 6.5; 4.1] 0.76
Boulet et al. Critical Care (2024) 28:429 Page 8 of 13

Table 2 (continued)
*(NaCl 0.9%, Ringer lactate, isotonic balanced solutions)
#
2.5% and 5% glucose solution
$
Renal Replacement Therapy

risk 0.7 [95%CI 0.32; 1.53], p = 0.37. Severe hyperna- (with or without albumin) during the first two weeks of
tremia (Na > 150 mmol/L) was observed in 1 patient (4%) hospitalization in ICU. In this trail, a higher difference in
in the optimized restrictive strategy group and 2 (8%) cumulative fluid balance was obtained with large doses
in the control group, relative risk 0.5 [95%CI 0.05; 5.15], of furosemide and perfusions of 20% albumin. Thus,
p = 0.37. Acute kidney injury KDIGO 2 was observed the findings were probably more dependent on medical
in 9 patients (37%) in the optimized restrictive strategy treatment (and potential side effects) rather than on the
group and 11 patients (46%) in the control group. Median restrictive strategy per se.
cumulative dose of diuretics administered over 7 days Although we were unable to demonstrate a statistically
was similar in both groups: mg furosemide 40 mg (IQR significant reduction in fluid balance, it is noteworthy
20; 16) in the optimized restrictive strategy group vs that the total fluid intakes with our protocol were much
60 mg (IQR 30; 320) in the control group, mean differ- lower than those of the intervention groups in previ-
ence-32 [− 810; 746], p = 0.56. Changes in norepineph- ously cited fluid restriction studies [11, 19]. Notably, the
rine doses and lactate levels during the first 7 days of ICU fluid balance of the control group in our study was also
stay are reported in Supplementary Table S7. lower than, or similar to, that of patients in the interven-
tion group in the previously cited fluid restriction stud-
Discussion ies, without systematic administration of diuretics [11,
In this multicenter, single-blind, randomized, controlled, 19, 21]. This could explain why no difference was seen in
pilot trial, an optimized fluid restriction protocol to clinical outcomes or metabolic parameters such as hyper-
reduce “hidden fluid intakes” did not lead to a relative natremia, acute renal failure and issues related to nutri-
40% reduction in fluid balance at Day 5 when compared tional intake. Promisingly, the optimized restrictive fluid
with standard fluid strategy (control group). The fluid strategy was not associated with notable side-effects.
balances at Day 3, Day 5, changes in weight, number of Our study had several limitations. First, the infused
days free of organ failure and mortality rate at Day 28 fluid volumes in the control group were far below the
were similar in both groups. ranges observed in recent trials on fluid restriction,
We cannot directly compare our results to previ- which prevented us from demonstrating a significant
ous studies on fluid restriction because OPTIFLUID is difference. Possibly the intensivists in the participating
the first study which targeted uniquely non resuscita- centers were convinced of the benefits of the restrictive
tive fluids. Recently, two large RCT [11, 19] assessed the fluid strategy and inadvertently applied a fluid restrictive
impact of a restrictive strategy applied during the ICU strategy also in the control group, explaining the non-
stay focused on all intravenous fluids and not uniquely on representativeness of the control group Moreover, the
resuscitative fluids. In both studies, the restrictive strat- differences in administered volumes were almost entirely
egy reduced cumulative fluid balance with no effect on restricted to maintenance fluids, while dilutional flu-
mortality rate. ids were identical. It is possible that the doctors and the
In the CLASSIC trial, Meyhoff et al. [11] found no nurses at the participating centres, seeing that different
reduction in 90-day mortality or any other secondary dilutions with higher concentrations could be used in the
outcome in septic patients. Their fluid restriction proto- intervention arm, started using the same dilutions in the
col focused mostly on resuscitative fluids (intravenous control arm, which does not constitute a breach of proto-
fluid therapy for hypoperfusion or to correct fluid losses col sensu stricto. An alternative explanation may be that
or dehydration), with a very aggressive protocol forbid- dilutional fluids cannot be adequately minimized as many
ding fluids if the lactate level was lower than 4 mmol/l or drugs require a predefined amount of fluids. Second, the
in absence of oliguria. However, in a secondary post-hoc physicians and nurses were not blinded to group assign-
analysis this strategy did not seem to affect the time to ment due to the individual randomization design. A
resolution of hyperlactatemia [20]. randomized step-wedge cluster design would overcome
In the POINCARE-2 trial [19], Bollaert et al. found this issue, but would require more centers and patients.
no reduction in mortality at day 60 or in any other sec- Thirdly, renal dysfunction was more severe in the control
ondary outcome in a broad range of critically ill patients group, both at baseline and after the intervention phase.
with a fluid restriction protocol based on daily weigh- This difference may be due to the intervention, or to the
ing, salt and fluid restriction and the use of furosemide severity of the patients at the time of randomization,
Boulet et al. Critical Care (2024) 28:429 Page 9 of 13

Fig. 2 A Cumulative fluid balance variation during intervention. B Weight variation from baseline during intervention
Boulet et al. Critical Care (2024) 28:429 Page 10 of 13

Table 3 Primary and secondary outcomes

Optimized restrictive Control group (n = 24) Absolute difference Relative risk p-Value
strategy group
(n = 24)

Primary outcome
Cumulative fluid balance at Day 5—Median 6.9 (− 13.7; 52.1) 35.0 (− 7.9; 40.2) 13.2 [− 15.2; 41.6] 0.42
(IQR) ml/kg of admission body weight
Secondary outcomes
Cumulative fluid balance at Day 3—Median 9.8 (− 8.0; 38.8) 24.7 (2.2; 37.7) 13.8 [− 6.9; 34.4] 0.19
(IQR) ml/kg of admission body weight
Cumulative fluid balance at Day 7—Median 12.6 (− 19.9; 58.7) 26.4 (16.3; 39.8) 12.5 [− 24.2; 49.3] 0.72
(IQR) ml/kg of admission body weight
Body weight variation at Day 3—Median (IQR) 0.0 (0.01; 0.02) 0.01 (0.01; 0.04) 0.01 [− 0.02; 0.05] 0.34
kg (%)
Body weight variation at Day 5—Median (IQR) 0.0 (0.01; 0.03) 0.0 (0.0; 0.0) 0 [− 0.04; 0.03] 0.84
kg (%)
Body weight variation at Day 7—Median (IQR) 0.00 (0.0; 0.0) 0.0 (0.0; 0.0) − 0.01 [− 0.04; 0.03] 0.83
kg (%)
Death by Day 7—N (%) 4 (16.7) 5 (20.8) − 0.04 [− 0.26; 0.18] 0.8 [0.24; 2.62] 1.00
Death by Day 28—N (%) 6 (25.0) 6 (25.0) 0 [− 0.24; 0.24] 1 [0.38; 2.66] 1.00
Days alive without life support and with- 7.0 (5.0; 7.0) 7.0 (5.5;7.0) 0.21 [− 1; 1.4] 0.74
out organ failure [SOFA = 0] at Day 28
Days alive without vasopressors at Day 28— 27.0 (16.0; 27.0) 27.0 (8.0; 27.0) − 0.8 [− 6.7; 5.2] 0.87
Median (IQR)
Days alive without mechanical ventilation 27.0 (18.5; 27.0) 27.0 (15.0; 27.0) − 0.3 [− 6.0; 5.4] 0.93
at Day 28—Median (IQR)
Days alive without renal replacement therapy 27.0 (17.5; 27.0) 27.0 (14.5; 27.0) − 0.4 [− 6.1; 5.3] 0.79
at Day 28—Median (IQR)
Days alive and out of ICU at Day 28—Median 18.0 (2.0; 23.0) 17.5 (0.0; 23.0) − 0.5 [− 6.5; 5.5] 0.93
(IQR)
ICU length of stay—Median (IQR) days 5.0 (3.0; 10.0) 5.0 (3.0; 10.0) 0.6 [− 3.6; 4.7] 0.85
Hospital length of stay—Median (IQR) days 11.0 (8.0; 22.0) 12.0 (8.0; 18.0) 0.1 [− 5.2; 5.4] 0.97

which could justify higher fluid intake in patients in the Nevertheless, the study had several strengths. First,
control group and could also contribute to differences in we targeted hidden fluid intake in contrast to previ-
fluid balance. ous studies. Second, the protocol permitted obtaining
Another limitation of our trial was that the strict the lowest fluid balance on Day 5 ever published in a
eligibility criteria led to excluding some of the most population of septic patients, without restricting the
severe patients. The relatively low level of lactate at volumes of fluids used to manage hypoperfusion in the
baseline suggests that the early acute phase of septic acute phase of septic shock, and without systematically
shock had already resolved. Therefore, our restriction using diuretics. Interestingly, the use of diuretics was
protocol, particularly with regards to safety, should not lower in the intervention group, although this was not
be implemented in the first few hours of management statistically significant. Third, all patients in the inter-
of septic shock. A final point worth emphasizing is that ventional group were able to comply with the protocol,
the artificial nutrition part of the restriction protocol suggesting that the optimized restrictive fluid strategy
had no impact on cumulative fluid volume. This could would be compatible for all patients in septic shock,
be explained by the short duration of mechanical venti- regardless of severity. Finally, we did not observe more
lation, during which patients received enteral nutrition, severe adverse effects in the optimized restrictive strat-
which prevented us from observing the impact of using egy group, especially severe hypernatremia and acute
a hyper-caloric, low-volume formula in the optimized kidney injury KDIGO 3.
restrictive strategy group. Although our pilot study did not achieve signifi-
cance, considering the non-significant trend in fluid
balance reduction in the optimized group (Fig. 2B),
our approach to limiting fluid overload remains worth
Boulet et al. Critical Care (2024) 28:429 Page 11 of 13

Table 4 Adverse events

Optimized restrictive Control group (n = 24) Absolute difference* Relative risk* p-Value
strategy group
(n = 24)

Metabolic disorders
Frequency of mild hypernatremia (between 7 (29.2) 10 (41.7) − 0.13 [− 0.39; 0.14] 0.7 [0.32; 1.53] 0.37
146 and 150 mmol/l)—N (%)
Frequency of severe hypernatremia 1 (4.2) 2 (8.4) − 0.04 [− 0.18; 0.09] 0.5 [0.05; 5.15] 1.00
(> 150 mmol/l)—N (%)
Frequency of hyponatremia 1 (4.2) 1 (4.2) 0 [− 0.11; 0.11] 1 [0.07; 15.08] 1.00
(< 130 mmol/l)—N (%)
Renal function
Frequency of hyperkalemia (> 6 mmol/l)—N 0,0 0,0 0,0 –
(%)
Frequency of Acute Kidney Injury KDIGO2— 9 (37.5) 11 (45.8) − 0.08 [− 0.36; 0.19] 0.82 [0.42; 1.61] 0.56
N (%)
Frequency of Acute Kidney Injury KDIGO3— 3 (12.50) 7 (29.17) − 0.17 [− 0.39; 0.06] 0.43 [0.13; 1.46] 0.16
N (%)
Use of diuretics (cumulative dose adminis- 40 (20; 160) 60 (30; 320) − 31.9 [− 810.7; 746.9] – 0.56
tered over 7 days)—mg furosemide
Nutritional issues
Frequency of hypoglycemia (< 0.6 g/l)—N 2 (8.4) 0 (0.00) 0.08 [− 0.03; 0.19] – 0.49
(%)
Insulin use (cumulative dose administered 99.0 (6.0; 211.5) 84.5 (50.0; 129.00) − 51.5 [− 152.0; 48.9] 0.73
over 7 days)—units
Frequency of stage 3 pressure sores acquired 0 (0.0) 1 (4.2) − 0.04 [− 0.12; 0.04] – 1.00
in the ICU—N (%)
*Adjusted values

exploring in a larger randomized controlled trial, For evaluation of patient safety, this trial was registered
perhaps in addition to other strategies such as early with France National Agency of Drug Safety (Agence
vasopressor use in the first 24 h of shock [10], limit- Nationale de Sécurité du Médicament et des Produits de
ing intravenous fluids in the absence of signs of severe Santé, ANSM).
hypoperfusion [11], and systematic use of diuret- Trial data were monitored at the sites by independent
ics [19]. An environmental and economic assessment monitors and monitored centrally by staff at the coor-
would also be interesting. As we did not observe any dinating center. Statistical analyses were performed by
harm to the patient, the additional savings made on the trial statistician in accordance with International
unnecessary fluids would probably be advantageous for Conference on Harmonisation Good Clinical Practice
ecological and financial reasons. guidelines. The authors vouch for the accuracy and com-
pleteness of the reported analyses and for the fidelity of
the trial to the protocol.
Conclusions
Abbreviations
In our pilot study, an optimized restrictive fluid strategy ICU Intensive Care Unit
targeting hidden fluid intakes did not reduce the overall SAPS II Simplified Acute Physiology Score II
fluid balance at Day 5. However, the very low fluid bal- IQR Interquartile range
KDIGO score Kidney Disease Improving Global Outcomes
ance obtained with the protocol and the unrepresenta- SOFA score Sepsis-related Organ Failure Assessment
tive control group, warrant further exploring this line of
research in a larger randomized study. Supplementary Information
The online version contains supplementary material available at https://​doi.​
Study protocol org/​10.​1186/​s13054-​024-​05155-z.
The trial protocol, which includes the statistical analysis
plan, was registered on ClinicalTrials.gov with the identi- Additional file 1.

fier NCT04947904.
Boulet et al. Critical Care (2024) 28:429 Page 12 of 13

Acknowledgements 4. Sakr Y, Vincent JL, Reinhart K, Groeneveld J, Michalopoulos A, Sprung CL,

We appreciate the interest of family members and patients in research, which Artigas A, Ranieri VM. Sepsis Occurence in Acutely Ill Patients I: high tidal
enabled this trial. We thank research coordinators and physicians from all volume and positive fluid balance are associated with worse outcome in
participating OPTIFLUID centers for their collaboration. We would like to thank acute lung injury. Chest. 2005;128(5):3098–108.
Julie Pisano for help drafting the protocol, Christophe Masseguin for regula- 5. Sakr Y, Rubatto Birri PN, Kotfis K, Nanchal R, Shah B, Kluge S, Schroeder
tory help, Loubna Elotmani, Audrey Ambert, Sabrina Ouberai, Sophie Lloret for ME, Marshall JC, Vincent JL. Intensive care over nations I: higher fluid
data collection, and Sarah Kabani for editorial assistance. OPTIFLUID Investiga- balance increases the risk of death from sepsis: results from a large inter-
tors Dr Charlotte Danière, Dr Aurelien Daurat, Dr Geoffroy Dingemans, Dr national audit. Crit Care Med. 2017;45(3):386–94.
Mikael François, Dr Nicolas Perin, Dr Eve Payan, Dr Remi Trusson, Dr Pauline 6. Perner A, Haase N, Guttormsen AB, Tenhunen J, Klemenzson G, Aneman
Wyllinck; Clinical Pharmacist: Dr Hélène Poujol; Nimes University Hospital. Dr A, Madsen KR, Moller MH, Elkjaer JM, Poulsen LM, et al. Hydroxyethyl
Lauren Beaudeau; Alès General Hospital starch 130/04.2 versus Ringer’s acetate in severe sepsis. N Engl J Med.
2012;367(2):124–34.
Author contributions 7. Russell JA, Walley KR, Singer J, Gordon AC, Hebert PC, Cooper DJ,
NB: Writing—Original Draft. CS: Methodology, Formal analysis. JPQ, NA, SG, AB, Holmes CL, Mehta S, Granton JT, Storms MM, et al. Vasopressin versus
CR: Investigation, Validation. JYL, LM: Writing – review & editing, Supervision. norepinephrine infusion in patients with septic shock. N Engl J Med.
SBD: Conceptualization, Writing – review & editing, Supervision, Funding 2008;358(9):877–87.
acquisition. 8. Van Regenmortel N, Verbrugghe W, Roelant E, Van den Wyngaert T,
Jorens PG. Maintenance fluid therapy and fluid creep impose more
Funding significant fluid, sodium, and chloride burdens than resuscitation fluids in
This study was funded by an internal grant from the CHU de Nîmes. critically ill patients: a retrospective study in a tertiary mixed ICU popula-
tion. Intensive Care Med. 2018;44(4):409–17.
Availability of data and materials 9. Silversides JA, Fitzgerald E, Manickavasagam US, Lapinsky SE, Nisenbaum
Principal data are provided within the manuscript or supplementary informa- R, Hemmings N, Nutt C, Trinder TJ, Pogson DG, Fan E, et al. Deresuscita-
tion files. The other data from this study are available upon reasonable request tion of patients with iatrogenic fluid overload is associated with reduced
to the corresponding author. mortality in critical illness. Crit Care Med. 2018;46(10):1600–7.
10. Heart N. Lung, and blood institute prevention and early treatment
of acute lung injury clinical trials network: early restrictive or liberal
Declarations fluid management for sepsis-induced hypotension. N Engl J Med.
2023;388(6):499–510.
Ethics approval and consent to participate 11. Meyhoff TS, Hjortrup PB, Wetterslev J, Sivapalan P, Laake JH, Cronhjort
The study was approved by the Comité de Protection des Personnes Sud- M, Jakob SM, Cecconi M, Nalos M, Ostermann M, et al. Restriction
Méditerranée II (Number 2020-A01952-37). Written informed consent was of intravenous fluid in ICU patients with septic shock. N Engl J Med.
obtained from patients or their legal surrogates. 2022;386(26):2459–70.
12. Corl KA, Prodromou M, Merchant RC, Gareen I, Marks S, Banerjee D, Amass
Consent for publication T, Abbasi A, Delcompare C, Palmisciano A, et al. The restrictive IV fluid trial
Written informed consent was obtained from patients or their legal surrogates in severe sepsis and septic shock (RIFTS): a randomized pilot study. Crit
either before randomization or as soon as possible thereafter. Care Med. 2019;47(7):951–9.
13. Semler MW, Janz DR, Casey JD, Self WH, Rice TW. Conservative fluid man-
Competing interests agement after sepsis resuscitation: a pilot randomized trial. J Intensive
The authors declare no competing interests. Care Med. 2020;35(12):1374–82.
14. Toulouse E, Granier S, Nicolas-Robin A, Bahans C, Milman A, Andrieu
Author details VR, Gricourt Y. The French clinical research in the European Community
1
Division of Anesthesia, Critical Care, Pain and Emergency Medicine, regulation era. Anaesth Crit Care Pain Med. 2023;42(2): 101192.
UR‑UM103 IMAGINE, University of Montpellier, Nimes University Hospital, 15. Evans L, Rhodes A, Alhazzani W, Antonelli M, Coopersmith CM, French C,
Nîmes, France. 2 Division of Anesthesia, Critical Care, Pain and Emergency Machado FR, McIntyre L, Ostermann M, Prescott HC, et al. Surviving sepsis
Medicine, Nimes University Hospital, University of Montpellier, Nîmes, France. campaign: international guidelines for management of sepsis and septic
3
Department of Intensive Care, Burgundy University Hospital, Dijon, France. shock 2021. Crit Care Med. 2021;49(11):e1063–143.
4
Lipness Team, INSERM Research Center LNC‑UMR1231 and LabEx LipSTIC, 16. Longuet P, Lecapitaine AL, Cassard B, Batista R, Gauzit R, Lesprit P, Haddad
University of Burgundy, Dijon, France. 5 INSERM CIC 1432, Clinical Epidemiol- R, Vanjak D, Diamantis S. Groupe des referents en infectiologie dI-d-F:
ogy, University of Burgundy, Dijon, France. 6 Department of Biostatistics, preparing and administering injectable antibiotics: How to avoid playing
Epidemiology, Public Health and Innovation in Methodology, Nimes University God. Med Mal Infect. 2016;46(5):242–68.
Hospital, University of Montpellier, Nîmes, France. 7 Department of Intensive 17. Casaer MP, Mesotten D, Hermans G, Wouters PJ, Schetz M, Meyfroidt
Care, Alès General Hospital, Alès, France. G, Van Cromphaut S, Ingels C, Meersseman P, Muller J, et al. Early
versus late parenteral nutrition in critically ill adults. N Engl J Med.
Received: 10 September 2024 Accepted: 31 October 2024 2011;365(6):506–17.
18. Reignier J, Plantefeve G, Mira JP, Argaud L, Asfar P, Aissaoui N, Badie J,
Botoc NV, Brisard L, Bui HN, et al. Low versus standard calorie and protein
feeding in ventilated adults with shock: a randomised, controlled, multi-
centre, open-label, parallel-group trial (NUTRIREA-3). Lancet Respir Med.
References 2023;11(7):602–12.
1. Singer M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D, 19. Bollaert PE, Monnier A, Schneider F, Argaud L, Badie J, Charpentier C,
Bauer M, Bellomo R, Bernard GR, Chiche JD, Coopersmith CM, et al. The Meziani F, Bemer M, Quenot JP, Buzzi M, et al. Fluid balance control in
third international consensus definitions for sepsis and septic shock critically ill patients: results from POINCARE-2 stepped wedge cluster-
(sepsis-3). JAMA. 2016;315(8):801–10. randomized trial. Crit Care. 2023;27(1):66.
2. Heung M, Wolfgram DF, Kommareddi M, Hu Y, Song PX, Ojo AO. Fluid 20. Ahlstedt C, Sivapalan P, Kriz M, Jacobson G, Sylvest Meyhoff T, Skov Kaas-
overload at initiation of renal replacement therapy is associated with Hansen B, Holm M, Hollenberg J, Nalos M, Rooijackers O, et al. Effects
lack of renal recovery in patients with acute kidney injury. Nephrol Dial of restrictive fluid therapy on the time to resolution of hyperlactatemia
Transplant. 2012;27(3):956–61. in ICU patients with septic shock. A secondary post hoc analysis of the
3. Heart N. Lung, and blood institute acute respiratory distress syndrome CLASSIC randomized trial. Intensive Care Med. 2024;50(5):678–86.
clinical trials network: comparison of two fluid-management strategies in 21. Cinotti R, Lascarrou JB, Azais MA, Colin G, Quenot JP, Mahe PJ, Roquilly A,
acute lung injury. N Engl J Med. 2006;354(24):2564–75. Gaultier A, Asehnoune K, Reignier J. Diuretics decrease fluid balance in
Boulet et al. Critical Care (2024) 28:429 Page 13 of 13

patients on invasive mechanical ventilation: the randomized-controlled

single blind, IRIHS study. Crit Care. 2021;25(1):98.

Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in pub-
lished maps and institutional affiliations.