Fluid Bolus Over 15–20 Versus 5–10 Minutes Each
in the First Hour of Resuscitation in Children With
Septic Shock: A Randomized Controlled Trial
Jhuma Sankar, MD1; Javed Ismail, MD1; M. Jeeva Sankar, MD, DM1; Suresh C.P., MD2;
Rameshwar S. Meena, MD2
Objectives: To compare the effect of administration of 40–60 mL/
kg of fluids as fluid boluses in aliquots of 20 mL/kg each over
15–20 minutes with that over 5–10 minutes each on the compos-
ite outcome of need for mechanical ventilation and/or impaired
oxygenation—increase in oxygenation index by 5 from baseline in
the initial 6 and 24 hours in children with septic shock.
Design: Randomized controlled trial.
Setting: Pediatric emergency and ICU of a tertiary care institute.
Patients: Children (< 18 yr old) with septic shock.
Interventions: We randomly assigned participants to 15–20 minutes
bolus (study group) or 5–10 minutes bolus groups (control group).
Measurements and Main Results: We assessed the composite
outcomes in the initial 6 and 24 hours after fluid resuscitation in
both groups. We performed logistic regression to evaluate factors
associated with need for ventilation in the first hour. Data were
analyzed using Stata 11.5. Of the 96 children, 45 were randomly
assigned to “15–20 minutes group” and 51 to “5–10 minutes
group.” Key baseline characteristics were not different between
the groups. When compared with 5–10 minutes group, fewer chil-
dren in 15–20 minutes group needed mechanical ventilation or
had an increase in oxygenation index in the first 6 hours (36% vs
57%; relative risk, 0.62; 95% CI, 0.39–0.99) and 24 hours (43%
vs 68%; relative risk, 0.63; 95% CI, 0.42–0.93) after fluid resusci-
tation. We did not find any difference in secondary outcomes such
as death (1.2; 0.70–2.03), length of stay (mean difference: 0.52;
–1.72 to 2.7), or resolution of shock (0.98; 0.63–1.53).
1
Department of Pediatrics, All India Institute of Medical Sciences, New
Delhi, India.
2
Department of Pediatrics, PGIMER, Dr RML Hospital, New Delhi, India.
Clinical trial registration number: CTRI/2014/09/005050.
Supplemental digital content is available for this article. Direct URL cita-
tions appear in the printed text and are provided in the HTML and PDF
versions of this article on the journal’s website (http://journals.lww.com/
pccmjournal).
The authors have disclosed that they do not have any potential conflicts
of interest.
For information regarding this article, E-mail: jhumaji@gmail.com
Copyright © 2017 by the Society of Critical Care Medicine and the World
Federation of Pediatric Intensive and Critical Care Societies
DOI: 10.1097/PCC.0000000000001269
Pediatric Critical Care Medicine www.pccmjournal.org 1
Copyright © 2017 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies.
Unauthorized reproduction of this article is prohibited
Conclusion: Children receiving fluid boluses over 5–10 minutes
each had a higher risk of intubation than those receiving boluses
over 15–20 minutes each. Notwithstanding the lack of difference
in risk of mortality and the possibility that a lower threshold of intu-
bation and mechanical ventilation was used in the presence of fluid
overload, our results raise concerns on the current recommenda-
tion of administering boluses over 5–10 minutes each in children
with septic shock. (Pediatr Crit Care Med 2017; XX:00–00)
Key Words: duration of bolus; fluid bolus; fluid overload; septic
shock
C
urrent guidelines for septic shock in children advocate
aggressive fluid resuscitation of up to 60 mL/kg—as
boluses of 20 mL/kg each over 5–10 minutes to achieve
the desired heart rate and blood pressure (1). The boluses
are repeated until perfusion improves, unless hepatomegaly,
increased work of breathing, rales, or a gallop rhythm develops.
The goals of such early and rapid fluid resuscitation are to lower
the heart rate and capillary refill times, and to increase blood
pressure and central venous oxygen saturation (2). Studies have
shown that this aggressive approach to fluid resuscitation leads
to a nine-fold reduction in mortality (3, 4). Children respond
well to such volume resuscitation, as they tend to tolerate larger
fluid volumes per kilogram of body weight than adults (3).
Administration of bolus fluid over 5–10 minutes each, how-
ever, may not be followed in many units in developing countries
due to two major reasons: first, difficulty in administration of
large volumes in such short duration due to shortage of human
resources. Administration of large volumes over 5–10 minutes
more often requires manual pull-push technique to administer
the boluses (5). In resource-limited countries with overcrowd-
ing and lack of space and human resources, it becomes rather
impossible to administer the boluses in such a short duration.
Second, fear of fluid overload and need for ventilation with
boluses administered so quickly. For example, in a previous
study from India (6), the authors reported an increase in the
occurrence rate of hepatomegaly in children receiving boluses
 Sankar et al
over 5–10 minutes duration. Few units are fully equipped in
resource-restricted settings to initiate and sustain mechanical
ventilation in those getting ventilated due to fluid overload/
pulmonary edema; therefore, this is an important safety con-
cern in all other units not so well-equipped to administer the
boluses so quickly.
Recent evidence from resource-restricted settings (Fluid
Expansion as Supportive Therapy [FEAST] trial) showed worse
outcomes with bolus fluid administration (7). Following the pub-
lication of the FEAST trial, researchers have begun questioning the
basics of fluid resuscitation including the need for and duration of
administration of boluses. In the FEAST trial, however, children who
are more vulnerable to the detrimental effects of volume loading
such as those with malaria, severe anemia, and severe malnutrition
were common. The trial results, therefore, need to be interpreted
in the appropriate clinical context. Different studies have reported
varying duration of administration of the three boluses ranging
from 60 to 90 minutes (3, 4, 6) and the only one randomized con-
trolled trial (RCT) published (7) showed increased rates of hepa-
tomegaly but not ventilation with boluses administered so quickly.
Thus, there is paucity of evidence on the optimal duration
of administration of fluid boluses in the first hour of fluid
resuscitation. Given the background, we undertook this study
to address the question as to what is the optimal duration/
speed of bolus fluid administration during initial resuscitation
that would be safe and better tolerated in children with septic
shock in the initial hour of resuscitation.
MATERIALS AND METHODS
Setting and Participants
This was an RCT conducted from September 2013 to August
2014 in a tertiary care emergency and PICU in North India. We
screened all children with features of septic shock—defined
as children with suspected infection and having two or more
clinical signs of decreased perfusion (2, 8) for eligibility (Elec-
tronic supplement 1, Supplemental Digital Content 1, http://
links.lww.com/PCC/A482).
Children with contraindications to central venous catheter
insertion, those with dengue, malaria, severe anemia, severe
malnutrition, primary cardiac illness, those on noninvasive
ventilation before developing shock, and those with prior
fluid and inotrope therapy were excluded. Eligible children
were enrolled after obtaining written informed consent from
one of the parents. The procedure of obtaining consent was a
staged one beginning with verbal assent followed by delayed
written consent. This is described in detail in Electronic sup-
plement 2 (Supplemental Digital Content 2, http://links.lww.
com/PCC/A483). The study was approved by the Institutional
Ethics Committee (IEC), and the trial was registered retrospec-
tively—because of procedural delays in registration system—
in the Clinical Trials Registry of India (CTRI/2014/09/005050).
Randomization and Blinding
Children fulfilling eligibility criteria and not having any of the
exclusion criteria were enrolled. They were then stratified into
2 www.pccmjournal.org XXX 2017 • Volume XX • Number XXX
Copyright © 2017 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies.
Unauthorized reproduction of this article is prohibited
two strata depending on their mechanical ventilation status—
those already on ventilator at the time of recognition of shock
to “stratum A” and those not on ventilators to “stratum B.”
Children in both strata were randomized to one of the fluid
bolus groups—“15–20 minutes group” and “5–10 minutes
group” by the resident on duty. The resident would open the
serially numbered, sealed, opaque envelopes after obtaining
verbal consent. We used variable block randomization—using
random number table generated from computer—in varying
block sizes of 2–8. All the parents who had given verbal con-
sent initially gave written consent later—there were no post-
randomization exclusions.
Because of the nature of the intervention, the clinical team
could not be masked to the treatment allocation in the two
groups. However, the clinical team treating the patient was
not involved in recording outcomes and the outcome recorder
(R.S.M.) was blinded to the group allocation.
Objectives and Outcome Measures
Our primary objective was to compare the effect of adminis-
tration of fluid boluses in volumes of 20 mL/kg each (total vol-
ume of 40–60 mL/kg in the first hour) over 15–20 minutes with
that over 5–10 minutes (control group) in the first hour of
resuscitation on the composite outcome of need for mechani-
cal ventilation (i.e., “newly ventilated”) and/or impaired oxy-
genation (i.e., increase in peak oxygenation index [OI] by 5 in
those already ventilated) in the first 6 and 24 hours following
fluid resuscitation, in children with septic shock. We defined
newly ventilated as the need for ventilation for respiratory
failure and/or persistent “hemodynamic instability” (i.e., not
attaining the therapeutic endpoints even after receiving up to
60 mL/kg of fluids and epinephrine of up to 0.3 μg/kg/min) (9)
with or without features of fluid overload (Electronic supple-
ment 1, Supplemental Digital Content 1, http://links.lww.com/
PCC/A482).
Our secondary objectives were to compare the difference
between the two groups with regard to proportion developing
clinical fluid overload (hepatomegaly and/or rales), propor-
tion of children requiring ventilation for fluid overload alone,
changes in OI, respiratory rates, Fio2 requirement and satura-
tion from baseline to up to 24 hours, total fluids received in mL/
kg at 6 and 24 hours, and fluid overload % (FO%) at 24 hours
(10). Other clinically important outcomes such as proportion
achieving therapeutic endpoints and time to achievement, ino-
trope score, duration of ventilation (Electronic supplement 1,
Supplemental Digital Content 1, http://links.lww.com/PCC/
A482), duration of inotrope therapy, Pediatric Logistic Organ
Dysfunction (PELOD) score (11), ICU stay, and mortality
were also compared between the two groups. Definitions are
provided in Electronic supplement 1 (Supplemental Digital
Content 1, http://links.lww.com/PCC/A482).
Methodology
Once enrolled, children in both groups were managed as per
standard pediatric guidelines for management of septic shock
(1, 2, 12). A year before this study, we had concluded a similar

study using this protocol in our unit (13), and thereafter, this
was the standard of care in patients with septic shock. After
this study, we ensured compliance with the protocol every time
a child was admitted with septic shock. We took special care
to ensure that the above protocol was the standard of care like
other established protocols for common ICU conditions. The
only change made in the protocol for this study was the dura-
tion of administration of the fluid boluses. As per the protocol,
at the onset, all children received boluses of normal saline up
to 60 mL/kg in the first hour with careful monitoring for fea-
tures of fluid overload (hepatomegaly, basal rales, or gallop)
to titrate fluid therapy in the first hour or until central venous
pressure (CVP) line was established (whichever was earlier).
The CVP line was to be established within the first hour in all
patients (it was established within first 30 min in most cases
during or after the first bolus). Further fluid therapy (beyond
the first hour) was essentially guided by CVP and clinical fea-
tures of hypoperfusion—a cut-off of less than 8 with one or
more clinical criteria for shock was used for administering
further fluid boluses. No more fluids were administered if the
CVP reached the respective cut-offs. Instead, we titrated the
vasoactive/inotropic support to optimize perfusion in these
children as per protocol (1, 2, 12). The therapeutic endpoints
targeted after each bolus are described in Electronic supple-
ment 1 (Supplemental Digital Content 1, http://links.lww.
com/PCC/A482). In the event of a child requiring intubation
and mechanical ventilation during or after fluid resuscitation,
the same was initiated. In addition, those who had persistent
hemodynamic instability, defined as not attaining the thera-
peutic endpoints even after receiving up to 60 mL/kg of flu-
ids and epinephrine of up to 0.3 μg/kg/min, or those who had
respiratory failure without any other features of fluid overload
were also intubated (9). After this, we followed a uniform pro-
tocol for optimizing further ventilation (14).
The method of bolus administration was infusion bag
method in the 15–20 minutes group and manual push–pull
method in the 5–10 minutes group. Given that the major
intervention was timing, we took the following steps to ensure
compliance with the timing. In both groups, a stop clock was
used for keeping the time. The timing was maintained by the
team administering the bolus (the senior resident in all cases).
The residents were trained before the onset of the trial on the
method of administering the bolus including keeping the tim-
ing through simulated case scenarios. After the training period,
in 10 patients each, the boluses were administered over 5–10
and 15–20 minutes by the residents under supervision of
the principal investigator. The compliance to the timing was
ensured, and any difficulties in achieving this were addressed
and corrected during this period. The formal trial was started
only after this.
Statistical Analysis
Data were entered into Microsoft Excel 2007 (Microsoft,
München, Germany) and analyzed using Stata 11.5 (Stata Corp,
College Station, TX). Categorical data are presented as number
(%), whereas continuous variables are presented as mean (sd),
Pediatric Critical Care Medicine www.pccmjournal.org 3
Copyright © 2017 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies.
Unauthorized reproduction of this article is prohibited
Online Clinical Investigation
if normally distributed and median (interquartile range [IQR]),
if skewed. Statistical analysis was performed using Student t
test/Wilcoxon rank sum test and chi-square test for continuous
and categorical variables, respectively. We calculated the relative
risk (RR)/mean difference with 95% CI for all the outcomes.
Sample Size Calculation
We calculated that 105 children per group were required to show
a relative reduction of 33%—from our unit data of 60–40%—in
the composite outcome, assuming an alpha error of 5% and power
of 80%. Our IEC mandates (we did not have a formal Data Safety
Monitoring Board [DSMB] for logistic reasons) that interim
analysis be performed at 40–50% of enrollment for clinical tri-
als with safety concerns. Accordingly, we performed the interim
analysis after enrolling 96 children (after about 50% enrollment).
The IEC did not use any formal stopping rules but had decided a
priori that a significant increase in ventilatory requirements and/
or deaths in any of the groups as the potential trigger to terminate
the trial. Because the interim analysis revealed significantly high
risk of composite outcome in the 5–10 minutes bolus group, the
IEC recommended stopping the trial.
To further examine the association between duration/speed
of bolus administration and increased need for mechanical
ventilation in the first hour, we performed multivariate logis-
tic regression using proportion newly ventilated at first hour
as the dependent variable and factors likely to influence the
outcome—decided a priori—as the independent variables.
We took only nonventilated children as the comparison group
for this purpose. The results of the multivariate analysis are
reported as odds ratio/mean difference with 95% CIs.
RESULTS
A total of 104 children were eligible during the study period, of
which eight were excluded as per prespecified exclusion criteria
(Fig. 1). There were no cases of protocol violations in any of the
groups and none had to be excluded therefore for this reason.
Of the 96 children, 45 were randomly assigned to 15–20 min-
utes group and 51 to 5–10 minutes group. Key baseline charac-
teristics including age, Pediatric Index of Mortality 2, PELOD
score, diagnosis, and hemodynamic and respiratory variables
were not different between the groups (Table 1). The mean (sd)
duration of administration of first bolus was 17 (1.5) versus 6
(0.8) minutes; for second bolus was 17 (1.2) versus 6 (1.1) min-
utes and 17 (1.4) versus 7.5 (1) minutes in the 15–20 minutes
group and 5–10 minutes group, respectively (Table 2).
Primary Outcomes
Fewer children in the 15–20 minutes group experienced the com-
posite outcome of newly ventilated or “increase in OI” immedi-
ately after fluid resuscitation in the first 6 and 24 hours—36%
versus 57% (RR, 0.62; 95% CI, 0.39–0.99) and 43% versus 68%
(RR, 0.63; 95% CI, 0.42–0.93), respectively (Table 2). The num-
ber needed to treat (NNT) was “5” and “4” at 6 and 24 hours,
respectively. The risk of composite outcome was significantly
lower in the 15–20 minutes group even in the first hour (18% vs
43%; RR, 0.41; 95% CI, 0.20–0.83). The NNT was 4.
 Sankar et al
Figure 1. Study flow of patients enrolled into the study.
The proportion of newly ventilated patients immediately
after fluid resuscitation (first hour) was even lower in the
15–20 minutes group (11% vs 35%; RR, 0.31; 95% CI, 0.13–
0.78). This continued to remain lower even by 6 and 24 hours
(Table 2). The median (IQR) time to initiating mechanical ven-
tilation was 215 minutes (52–340) and 25 minutes (15–180) in
the 15–20 minutes group and 5–10 minutes group, respectively
(p < 0.0001).
The OI increased from 8 and 7.2 to 9.5 and 13 at “6 hours”
and to 10.3 and 16 by “24 hours” in the 15–20 minutes group
and 5–10 minutes group, respectively. Even at “1 hour,” the
increase was lower in the 15–20 minutes group with the OI
increasing from 8 and 7.2 to 8.3 and 12 from baseline to 1 hour
(Table 3).
Secondary Outcomes
At the end of first hour, all those who were newly ventilated in
both groups were due to fluid overload. However, in the next 5
hours (i.e., in the first 6 hr of fluid resuscitation), five children
each in the 15–20 minutes group and the 5–10 minutes group
had to be ventilated for other reasons—epinephrine refractory
shock (Table 2). The numbers ventilated for other reasons in
the first 24 hours increased to 8 and 10 in the 15–20 minutes
group and 5–10 minutes group, respectively. If these children
were excluded from the 6- and 24-hour analyses, the propor-
tion of children “newly ventilated for fluid overload” alone was
also significantly lower in the 15–20 minutes group in the first
6 hours (8/40 vs 20/46; p = 0.02) and first 24 hours (8/37 vs
20/41; p = 0.01).
There was no difference between the two groups with regard
to other key clinical outcomes such as proportion achieving
therapeutic endpoints in first 24 hours or time to achieving
therapeutic endpoints, mortality, or others (Table 2). Although
there were no differences in baseline RR, Spo2, Pao2, and OI
between the two groups, there were differences in maximum
4 www.pccmjournal.org XXX 2017 • Volume XX • Number XXX
Copyright © 2017 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies.
Unauthorized reproduction of this article is prohibited
RR, lowest Pao2, lowest Spo2,
peak OI, and maximum Fio2
recorded in first 6 hours and
from 6 to 24 hours (Table 3).
The proportion developing
hepatomegaly was also higher
(41% vs 18%; RR, 0.4 [0.18–
0.85]) in the 5–10 minutes
group.
Additional Analyses
On multivariate logistic
regression to evaluate associa-
tion of factors associated with
“need for ventilation in the
first hour,” only the duration
of bolus administration over
5–10 minutes remained sig-
nificant after adjusting for all
other factors (p = 0.006). Elec-
tronic supplement 3 (Supple-
mental Digital Content 3, http://links.lww.com/PCC/A484)
shows the univariate and multivariate analyses of factors asso-
ciated with need for mechanical ventilation in the first hour.
DISCUSSION
Our findings are important as they raise important concerns
about the current recommendation of administering each
fluid bolus over 5–10 minutes in children with septic shock (1,
12). We observed that administration of fluid bolus over 15–
20 minutes each reduced the need for mechanical ventilation
(newly ventilated) in the first hour when compared with the
current recommendations of administering each bolus over
5–10 minutes (18% vs 43%; RR, 0.41; 95% CI, 0.20–0.83). The
NNT was 4. After 6 and 24 hours also, the need for mechani-
cal ventilation continued to be lower in the 15–20 minutes
group. The effect of duration or speed of fluid administration
on need for ventilation in first hour remained significant even
after adjusting for baseline demographic, physiologic, and ill-
ness severity variables further confirming the association.
At the end of the first hour, the proportion of children who
were newly ventilated for fluid overload was significantly lower
in the 15–20 minutes group as compared to the 5–10 minutes
group. At the end of 6 and 24 hours also, the proportion of
children newly ventilated for fluid overload continued to be
significantly lower in the 15–20 minutes group as compared to
the 5–10 minutes group. The probable association between the
speed of bolus administration and development of fluid over-
load necessitating mechanical ventilation is probably explained
by the changes in the respiratory parameters observed in the
two groups with time after the fluid resuscitation. For exam-
ple, the saturations and Pao2 dropped over time (lowest values
being 82% and 91% and 64 mm Hg and 76 mm Hg, respec-
tively, at 1 hr and in first 6 hr) and the Fio2 requirements and
peak OI increased (from 7.2 to 16) in the 5–10 minutes group
 Online Clinical Investigation

TABLE 1. Baseline Characteristics of the Two Study Groups

Variable 15–20 Min Group (n = 45) 5–10 Min Group (n = 51) p

Age (yr), median (IQR) 6.5 (1.5–11) 6.6 (0.9–12) 0.80

Male gender 31 (69) 30 (59) 0.30
Weight for age (z score) –1.79 (–3.01 to –1.21) –1.60 (–3.01 to –0.83) 0.53
Body mass index for age (z score) –2.64 (–4.7 to –1.31) –2.15 (–4.12 to –0.87) 0.33
Pediatric Index of Mortality 2 score (predicted probability of 74 (49–79) 71 (43–77) 0.18
death %), median (IQR)
Pediatric Logistic Organ Dysfunction score at admission, 12 (3–33) 12 (2–21) 0.35
median (IQR)
Shifted from emergency department 24 (53) 31 (61) 0.47
Prior antibiotic therapy 19 (42) 24 (47) 0.78
Days of antibiotic therapy, mean (sd) 3.8 (1.9) 3.4 (1.8) 0.08
Source of infection 0.15
 Pneumonia 13 (29) 7 (14)
 Meningitis 9 (20) 14 (27)
  Abdominal infection 6 (13) 11 (22)
  Skin and soft tissue infection 3 (7) 0
 Tuberculosis 5 (11) 3 (6)
 Malaria 0 2 (4)
  Septicemia without focus 5 (11) 7 (14)
Clinical parameters, mean (sd)
  Heart rate (beats/min) 135 (33) 131 (27) 0.51
  Respiratory rate 38 (13) 39 (14) 0.82
  Mean arterial pressure (mm Hg) 62 (13) 59 (12) 0.20
  Central venous pressure (cm H2O), median (IQR) 6 (4–9) 5 (3–6) 0.93
  Capillary refill time (s) 3 (0.7) 3 (0.9) 0.64
  Glasgow Coma Scale 10.5 (3.2) 10.3 (3.7) 0.4
Comorbidities 12 (27) 13 (25.5) 0.89
  Nephrotic syndrome 4 (8) 6 (12)
  Hematologic malignancies 2 (4) 3 (5)
  Chronic liver disease 1 (2) 0 (0)
  Underlying immunodeficiency 3 (7) 3 (5)
  Autoimmune disorders 2 (4) 1 (1.5)
Laboratory parameters
  Lactate in mmol/L, median (IQR) 2.9 (1.1–3.6) 3.0 (0.9–3.8) 0.47
  pH, mean (sd) 7.39 (0.15) 7.41 (0.10) 0.58
  Bicarbonate in mmol/L, median (IQR) 18 (16–25) 20 (15–22) 0.99
  Base excess, median (IQR) 8 (4.2–12.8) 7.8 (3.9–11.8) 0.90
  Superior vena cava saturation (%), mean (sd) 72 (11) 70 (12) 0.53

(Continued)

Pediatric Critical Care Medicine www.pccmjournal.org 5

Copyright © 2017 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies.
Unauthorized reproduction of this article is prohibited
 Sankar et al

TABLE 1. (Continued). Baseline Characteristics of the Two Study Groups

Variable 15–20 Min Group (n = 45) 5–10 Min Group (n = 51) p

  Proportion with superior vena cava O2 saturation ≤ 70% 27 (61) 34 (66) 0.8
  Oxygen saturation (%), mean (sd) 92 (15) 93 (14) 0.85
  Hemoglobin (g/dL), mean (sd) 9.7 (2.2) 10 (2.4) 0.52
  Total leukocyte count (/mm3), median (IQR) 16,000 (10,900–21,000) 15,000 (7,500–20,000) 0.67
 C-reactive protein (semiquantitative, positive—6 mg/L), n/N (%) 14/23 (61) 13/31 (42) 0.27
  Blood culture positive 7 (16) 11 (22) 0.78
Organisms isolated 7 (16) 11 (22) 0.78
  Escherichia coli 2 3
  Klebsiella pneumoniae 1 3
  Haemophilus influenzae 1 2
  Streptococcus pneumoniae 2 2
  Staphylococcus aureus 1 1
Type of shock
  Cold shock 37 (82) 43 (85) 0.7
IQR = interquartile range.
Data presented as n (%) unless otherwise indicated. Pediatric Logistic Organ Dysfunction score includes a total of 12 variables for six key organ dysfunctions
(cardiovascular, respiratory, hematologic, neurologic, renal, and hepatic).

over the first 24 hours. The proportion developing hepatomeg-
aly was also higher in this group.
Data from retrospective studies in critically ill children
with or without shock have shown early fluid overload to have
worse outcomes in terms of respiratory morbidity and mor-
tality (10, 15, 16). Studies by Arikan et al (10) and Sinitsky et
al (16) showed that peak FO% and cumulative fluid overload
at 48 hours, respectively, were associated with higher peak
OI, duration of ventilation, and ICU stay. However, none of
these studies had compared the effect of speed of administra-
tion of fluid boluses during initial resuscitation on develop-
ment of fluid overload and subsequent respiratory morbidity
in children with septic shock. Similar to the findings in these
studies, we observed not the cumulative volume, but the speed
of administration to be associated with increased respiratory
morbidity such as increase in peak OI over first 24 hours, fall in
saturations, Pao2 levels, and increased proportions of children
getting ventilated in the 5–10 minutes group. Thus, it appears
that it is not only the cumulative accumulation of fluids over
a period of time that affects the lungs but also the speed with
which it is administered. And in the latter case, the effects are
early and pronounced.
In contrast to a previous study (6) from a similar setting
in which the authors found higher occurrence rate of hepa-
tomegaly (70% vs 35%) but not intubation rates in the 5–10
minutes group as compared to the 15–20 minutes group, we
found higher rates of both hepatomegaly and intubation in the
5–10 minutes group in the first hour in our study. All those
who were intubated after the fluid resuscitation had other
features of fluid overload as well in addition to worsening of
respiratory parameters. The discrepancy between the results of
present study and previous study could simply be due to usage
of different thresholds for intubation and ventilation in the
two units. However, given the protocol used in our study was
a reasonably standard one—intubation for worsening respira-
tory rate or fall in saturation in the presence of signs of fluid
overload, the findings highlight the need to exercise caution
while administering fluid boluses rapidly in children with sep-
tic shock.
Also, there are important differences between this study and
ours which need to be kept in mind while interpreting results
of both of these studies. The lower intubation rates in the pre-
vious study could have been attributed to lower volumes used
in the 5–7 minutes group by the authors. Only 40 mL/kg was
given compared with 60 mL/kg in the 20 minutes group (6).
In our study, we used similar volumes in both groups (of up
to 60 mL/kg) to adhere to the guidelines (1, 12) as well as to
ensure similar volumes in both groups so that the results are
comparable. It appears from the above that it is not only the
rapidity with which the bolus is administered but also the vol-
umes administered that influence the development of tissue
edema in these children.
Studies have attributed the worsening of lung function and
organ dysfunction in those with iatrogenic fluid overload to
1) pathologic accumulation of interstitial fluid due to iatro-
genic infusion of fluids; 2) excessive “leakiness” or “capillary
leak syndrome” in children with sepsis (10, 17–21); 3) rapid
loss of the vasoconstrictor response in shock (21) that con-
fers protection by reducing perfusion to nonvital tissues; and
4) ischemia-reperfusion injury following rapid resuscitation

6 www.pccmjournal.org XXX 2017 • Volume XX • Number XXX

Copyright © 2017 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies.
Unauthorized reproduction of this article is prohibited
 Online Clinical Investigation

TABLE 2. Primary and Secondary Outcomes in the Study Groups

15–20 Min Group 5–10 Min Group Relative Risk
Variable (n = 45) (n = 51) (95% CI) p

Primary outcome, n (%)

  Proportion of children having composite
 outcome
  At the end of first hour 8 (18) 22 (43) 0.41 (0.20–0.83)a 0.008
  In first 6 hr 16 (36) 29 (57) 0.62 (0.39–0.99)b 0.04
  In first 24 hr 19 (43) 34 (68) 0.63 (0.42–0.93)c 0.02
  Proportion of children newly ventilated
  At the end of first hour 5 (11) 18 (35) 0.31 (0.13–0.78) 0.006
  In first 6 hr 13 (32) 25 (52) 0.59 (0.34–1.01) 0.05
  In first 24 hr 16 (36) 30 (59) 0.60 (0.38–0.95) 0.02
  Increase in oxygenation index by 5 in those
  already ventilated
  At the end of first hour 3/4 (75) 4/5 (80) 1.3 (0.75–2.34) 0.44
  In first 6 hr 7/16 (43) 12/29 (41) 0.9 (0.47–1.91) 0.9
  In first 24 hr 6/19 (32) 16 /34 (47) 1.5 (0.70–3.16) 0.29
Secondary outcome
  Proportion of children newly ventilated for
  fluid overload, n (%)
  At the end of first hour 5 (11) 18 (35) 0.31 (0.13–0.78) 0.006
  In first 6 hr 8/40 (18) 20/46 (39) 0.46 (0.23–0.92) 0.02
  In first 24 hr 8/37 () 20/41 (50) 0.44 (0.22–0.88) 0.01
  Proportion achieving therapeutic endpoints in 20 (44) 23 (45) 0.98 (0.63–1.53) 0.95
first 6 hr, n (%)
  Time to achieving therapeutic endpoints of 36 (24–48) 29 (12–42) 17 (–72.5 to 106.5) 0.43
shock (hr), median (IQR)
  Duration of mechanical ventilation (hr), 48 (24–180) 64 (48–102) 12.03 (–28.9 to 52.9) 0.07
median (IQR)
  Time to initiating mechanical ventilation (min), 215 (52–340) 25 (15–180) 114 (152–227) < 0.0001
median (IQR)
  Fluid overload % in first 24 hr 7 (3) 9 (4.5) –2 (–3.53 to –0.46) 0.01
  Inotrope score at 24 hr, mean (sd) 398 (153) 403.7 (167) –5 (–70.2 to 60.2) 0.54
  Duration of inotrope therapy (hr), median 68 (44–134) 82 (42–112) 0.62 (–49.5 to 50.7) 0.97
(IQR)
  Duration of ICU stay (d), median (IQR) 5 (2–8) 6 (3–8) 0.52 (–1.72 to 2.76) 0.64
  Mortality, n (%) 18 (40) 17 (33) 1.2 (0.70–2.03) 0.50

(Continued)

Pediatric Critical Care Medicine www.pccmjournal.org 7

Copyright © 2017 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies.
Unauthorized reproduction of this article is prohibited
 Sankar et al

TABLE 2. (Continued). Primary and Secondary Outcomes in the Study Groups

15–20 Min Group 5–10 Min Group Relative Risk
Variable (n = 45) (n = 51) (95% CI) p

Treatments received
  Fluids in mL/kg, mean (sd)
  End of resuscitation in first hour 57 (7) 59 (7) –2 (–4.84 to 0.84) 0.16
  Between 1 and 6 hr 95 (20) 97 (15) 1.21 (–6.04 to 8.48) 0.40
  Between 6 and 24 hr 158 (20) 151 (17) 7.14 (–0.54 to 14.82) 0.06
  Total no. of additional boluses in first 6 hr, 0 (0–1) 0 (0–1) –0.07 (–0.19 to 0.04) 0.82
median (IQR)
  No. of units of RBCs transfused, median (IQR) 0 (0–1) 0 (0–1) –0.05 (–0.08 to 0.02) 0.76
  Need for dopamine in first 6 hr, n (%) 38 (85) 46 (91) 0.93 (0.80–1.09) 0.41
  Time of administration of first dose of 15 (7) 16 (8) –1 (–4.03 to 2.03) 0.36
antimicrobials (min), mean (sd)
Details of boluses administered, min
  Fastest bolus 15 5 — —
  Slowest bolus 20 8 — —
  Duration of administration of first bolus
  Mean (sd) 17 (1.5) 6 (0.8) 10.9 (10.48–11.43) < 0.0001
  Range (minimum–maximum) 15–20 5–8 — —
  Duration of administration of second bolus
  Mean (sd) 17 (1.2) 6 (1.1) 10.8 (10.35–11.3) < 0.0001
  Range (minimum–maximum) 15–20 5–8 — —
  Duration of administration of third bolus
  Mean (sd) 17 (1.4) 7.5 (1) 9.8 (9.30–10.36) < 0.0001
  Range (minimum–maximum) 15–20 5–10 — —
  First bolus completed by 17 (1.5) 6 (0.8) 10.9 (10.48–11.43) < 0.0001
  Second bolus completed by 35 (2.2) 13 (1.7) 21.8 (21.0–22.6) < 0.0001
  Third bolus completed by 53 (3) 22 (2) 31 (30.0–32.0) < 0.0001
IQR, interquartile range.
Number needed to treat (NNT) 4 (95% CI, 2–13).
a

NNT 5 (95% CI, 2–56).

b

NNT 4 (95% CI, 2–20).

c

Data presented as n (%) unless otherwise indicated. Boldface font signifies p value significant or < 0.05.

with IV fluid boluses (18–20). Restoring the intravascular vol-
ume slowly may be more physiologic for the body to adjust
to. However, children with hypotension and severe ischemia
would need rapid restoration of intravascular volume as delay-
ing the process too much also would lead to end-organ damage
and mortality (13, 22–25). Therefore, fluid resuscitation has to
be balanced to address both of these concerns.
Few studies have reported early fluid overload to be associ-
ated with mortality (26–28). However, it remains unclear if the
mortality was related to excess fluid or increased disease sever-
ity in these studies. Most other studies have reported association
of fluid overload with respiratory morbidity (10, 15, 16) but not
mortality similar to what we found. Our study was not powered
to evaluate endpoints that might be more important, such as
mortality, which was higher in the slow bolus arm, although it
was nonsignificant. Mortality differences would outweigh the
intermediate outcome of ventilatory requirement. Therefore, it
remains unclear whether rapid fluid bolus administration affects
other clinically important outcomes such as mortality.
Strengths and Limitations
Our study is the first study to compare the speed of admin-
istration with similar volume of fluids in both groups as per
the current guidelines (1, 12). Our study addresses a very basic

8 www.pccmjournal.org XXX 2017 • Volume XX • Number XXX

Copyright © 2017 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies.
Unauthorized reproduction of this article is prohibited
 Online Clinical Investigation

TABLE 3. Comparison of Important Clinical and Laboratory Variables in the First 24 Hours
Between the Two Study Groups
Variable 15–20 Min Group (n = 45) 5–10 Min Group (n = 51) RR (95% CI); p

HR, mean (sd)

  Baseline (before fluid boluses) 135 (33) 131 (27) 0.51
  After the fluid boluses in first hour 142 (32) 131 (28) 0.09
  Maximum HR from 1 to 6 hr 137 (30) 133 (26) 0.47
  Maximum HR from 6 to 24 hr 132 (28) 129 (27) 0.60
RR, mean (sd)
  Baseline (before fluid boluses) 38 (13) 39 (14) 0.82
  After the fluid boluses in first hour 38 (12) 49 (14) 0.0003
  Maximum RR from 1 to 6 hr 35 (7) 42 (9) 0.003
  Maximum RR from 6 to 24 hr 28 (9) 35 (5) 0.005
Oxygen saturation %, mean (sd)
  Baseline (before fluid boluses) 97 (3) 97 (2) 0.13
  After the fluid boluses in first hour 89 (8) 82 (12) 0.006
  Lowest Spo2 from 1 to 6 hr 96 (4) 91 (7) 0.0001
  Lowest Spo2 from 6 to 24 hr 93 (4) 93 (5) 0.88
Pao2 in mm Hg, mean ± sd
  Baseline (before fluid boluses) 151 (91) 137 (67) 0.4
  After the fluid boluses in first hour 77 (16) 64 (17) < 0.0001
  Lowest Pao2 from 1 to 6 hr 88 (14) 76 (18) < 0.0004
  Lowest Pao2 from 6 to 24 hr 113 (47) 87 (28) 0.001
Fio2 %, mean (sd)
  Baseline (before fluid boluses) 72 (14) 70 (17) 0.44
  After the fluid boluses in first hour 78 (12) 88 (9) < 0.0001
  Maximum Fio2 from 1 to 6 hr 73 (12) 82 (21) 0.01
  Maximum Fio2 from 6 to 24 hr 68 (13) 77 (16) 0.004
OI, mean (sd)
  At baseline (in those ventilated before fluid bolus) 8 (0.8) 7.2 (0.8) 0.13
  After the fluid bolus 8.3 (4) 12 (5) 0.004
  Peak OI from 1 to 6 hr 9.5 (6) 13 (6.4) 0.04
  Peak OI from 6 to 24 hr 10.3 (6) 16 (6) 0.001
CVP in cm, mean (sd)
  Baseline (at the time of insertion of catheter)a 5 (3) 5 (3) 0.73
  Maximum CVP from 1 to 6 hr b 11 (4) 12 (4) 0.59
  Maximum CVP from 6 to 24 hr 11 (3) 13 (3) 0.05

(Continued)

Pediatric Critical Care Medicine www.pccmjournal.org 9

Copyright © 2017 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies.
Unauthorized reproduction of this article is prohibited
 Sankar et al

TABLE 3. (Continued). Comparison of Important Clinical and Laboratory Variables in the

First 24 Hours Between the Two Study Groups
Variable 15–20 Min Group (n = 45) 5–10 Min Group (n = 51) RR (95% CI); p

Mean arterial pressure in mm Hg, mean (sd)

  Baseline (before fluid bolus) 62 (13) 59 (12) 0.20
  After the fluid bolus 68 (16) 67 (15) 0.75
  Minimum MAP from 1 to 6 hr 66.5 (13.6) 65.1 (12) 0.55
  Minimum MAP from 6 to 24 hr 67.1 (15) 69.6 (15.7) 0.39
Superior vena cava oxygen saturation %, mean (sd)
  After the fluid bolusa 72 (11) 70 (12) 0.53
  Maximum Scvo2 from 1 to 6 hrb 75 (9) 73 (11) 0.57
  Maximum Scvo2 from 6 to 24 hr 78 (8) 69 (11) 0.05
Lactate in mmol/L, median (IQR)
  Baseline (before fluid boluses) 2.9 (1.1–3.6) 3.0 (0.9–3.8) 0.47
  At 6 hr 1.4 (0.9–3.8) 2.7 (1.2–3.8) 0.27
  At 24 hr 1.2 (0.9–3.3) 2.1 (1.1–3.1) 0.18
Base deficit in meq/L, median (IQR)
  Baseline (before fluid boluses) 8 (4.2–12.8) 7.8 (3.9–11.8) 0.90
  At 6 hr 5.8 (3.2–8.5) 6 (3.2–9.6) 0.73
  At 24 hr 5.4 (2.5–8.4) 5 (2.9–9.6) 0.85
Proportion developing hepatomegaly in first 6 hr, n (%) 8 (18) 21 (41) 0.43 (0.21–0.88);
0.01
CVP = central venous pressure, HR = heart rate, IQR = interquartile range, MAP = mean arterial pressure, OI = oxygenation index, RR = relative risk.
Mean time to establish CVP in both groups—42 (15) min.
a

CVP and superior vena cava oxygen saturation were available for 43 patients in 15–20 min group and 48 patients in 5–10 min group. Boldface font signifies
b

p value significant or < 0.05.

question of what should be the optimal duration of bolus fluid
administration and provides preliminary data for addressing
other important questions related to fluid bolus administration
in septic shock such as the effect of bolus duration on mortality.
Our study has major limitations too—1) we used a specific pro-
tocol for intubation and ventilation in the presence of fluid over-
load that may not be followed in all other centers. Some units may
prefer to use noninvasive ventilation in this scenario. Our results
may not be generalizable to these units. However, given the wors-
ening of respiratory parameters after the boluses such as increase
in respiratory rates, drop in Spo2, increase in Fio2, and increase in
OI from baseline (i.e., before administration) in the 5–10 minutes
group, it seems reasonable to exercise caution while administering
fluid boluses over 5–10 minutes each. The proportion with increase
in OI—not likely to be affected by the intubation protocol—was
not different between the two groups probably because of the small
number of ventilated patients in both groups. However, the mean
OI values increased over time in those already ventilated. Therefore,
this group needs to be studied in larger numbers to estimate the
true impact of fluid bolus administration in this group of patients;
2) the sample size was small and as mentioned, the proportion of
children already ventilated at enrollment was low—the effect of
duration of bolus administration in this subgroup could not have
been truly established; 3) we did not use any other parameters such
as inferior vena cava collapsibility/distensibility or passive leg rais-
ing methods to assess fluid responsiveness; and 4) we did not have
a formally constituted DSMB and the IEC takes the decision to ter-
minate the study after interim analysis in such cases.
CONCLUSION
Notwithstanding the lack of difference in risk of mortality and
the possibility that a lower threshold of intubation was used in
the presence of fluid overload, our results do raise concerns on
the current recommendation of administering boluses over 5–10
minutes each in children with septic shock. The effect of speed
of bolus fluid administration on OI needs further evaluation.
ACKNOWLEDGMENTS
We would like to sincerely acknowledge Drs. H. K. Kar and Bindu
Kulashreshtha, PGIMER Dr RML Hospital, and Drs. S. K. Kabra and
Rakesh Lodha, AIIMS, for their invaluable guidance and support.

10 www.pccmjournal.org XXX 2017 • Volume XX • Number XXX

Copyright © 2017 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies.
Unauthorized reproduction of this article is prohibited

REFERENCES
1. Dellinger RP, Levy MM, Rhodes A, et al; Surviving Sepsis Campaign
Guidelines Committee including The Pediatric Subgroup: Surviving
Sepsis Campaign: International guidelines for management of severe
sepsis and septic shock, 2012. Intensive Care Med 2013; 39:165–228
2. Carcillo JA, Fields AI; Comitê de Força-Tarefa: [Clinical practice
parameters for hemodynamic support of pediatric and neonatal
patients in septic shock]. J Pediatr (Rio J) 2002; 78:449–466
3. Carcillo JA, Davis AL, Zaritsky A: Role of early fluid resuscitation in
pediatric septic shock. JAMA 1991; 266:1242–1245
4. Han YY, Carcillo JA, Dragotta MA, et al: Early reversal of pediatric-
neonatal septic shock by community physicians is associated with
improved outcome. Pediatrics 2003; 112:793–799
5. Stoner MJ, Goodman DG, Cohen DM, et al: Rapid fluid resuscitation
in pediatrics: Testing the American College of Critical Care Medicine
guideline. Ann Emerg Med 2007; 50:601–607
6. Santhanam I, Sangareddi S, Venkataraman S, et al: A prospec-
tive randomized controlled study of two fluid regimens in the initial
management of septic shock in the emergency department. Pediatr
Emerg Care 2008; 24:647–655
7. Maitland K, Kiguli S, Opoka RO, et al; FEAST Trial Group: Mortality
after fluid bolus in African children with severe infection. N Engl J Med
2011; 364:2483–2495
8. Goldstein B, Giroir B, Randolph A: International pediatric sepsis con-
sensus conference: Definitions for sepsis and organ dysfunction in
pediatrics. Pediatr Crit Care Med 2005; 6:2–8
9. Nguyen HB, Rivers EP, Abrahamian FM, et al; Emergency Department
Sepsis Education Program and Strategies to Improve Survival
(ED-SEPSIS) Working Group: Severe sepsis and septic shock:
Review of the literature and emergency department management
guidelines. Ann Emerg Med 2006; 48:28–54
10. Arikan AA, Zappitelli M, Goldstein SL, et al: Fluid overload is associ-
ated with impaired oxygenation and morbidity in critically ill children.
Pediatr Crit Care Med 2012; 13:253–258
11. Leteurtre S, Martinot A, Duhamel A, et al: Validation of the Paediatric
Logistic Organ Dysfunction (PELOD) score: Prospective, observa-
tional, multicentre study. Lancet 2003; 362:192–197
12. Kleinman ME, Chameides L, Schexnayder SM, et al: Part 14: Pediatric
advanced life support: 2010 American Heart Association Guidelines
for Cardiopulmonary Resuscitation and Emergency Cardiovascular
Care. Circulation 2010; 122:S876–S908
13. Sankar J, Sankar MJ, Suresh CP, et al: Early goal-directed therapy in
pediatric septic shock: Comparison of outcomes “with” and “without”
intermittent superior venacaval oxygen saturation monitoring: A pro-
spective cohort study. Pediatr Crit Care Med 2014; 15:e157–e167
Pediatric Critical Care Medicine www.pccmjournal.org 11
Copyright © 2017 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies.
Unauthorized reproduction of this article is prohibited
Online Clinical Investigation
14. Saharan S, Lodha R, Kabra SK: Management of acute lung injury/
ARDS. Indian J Pediatr 2010; 77:1296–1302
15. Bhaskar P, Dhar AV, Thompson M, et al: Early fluid accumulation in
children with shock and ICU mortality: A matched case-control study.
Intensive Care Med 2015; 41:1445–1453
16. Sinitsky L, Walls D, Nadel S, et al: Fluid overload at 48 hours is asso-
ciated with respiratory morbidity but not mortality in a general PICU:
Retrospective cohort study. Pediatr Crit Care Med 2015; 16:205–209
17. Parikh SM, Mammoto T, Schultz A, et al: Excess circulating angiopoi-
etin-2 may contribute to pulmonary vascular leak in sepsis in humans.
PLoS Med 2006; 3:e46
18. Maitland K, George EC, Evans JA, et al; FEAST trial group: Exploring
mechanisms of excess mortality with early fluid resuscitation: Insights
from the FEAST trial. BMC Med 2013; 11:68
19. Funk DJ, Jacobsohn E, Kumar A: The role of venous return in critical ill-
ness and shock-part I: Physiology. Crit Care Med 2013; 41:250–257
20. Myburgh J, Finfer S: Causes of death after fluid bolus resuscitation:
New insights from FEAST. BMC Med 2013; 11:67
21. The SAFE Study Investigators: Saline or albumin for fluid resus-
citation in patients with traumatic brain injury. N Engl J Med 2007;
357:874–884
22. Rivers E, Nguyen B, Havstad S, et al; Early Goal-Directed Therapy
Collaborative Group: Early goal-directed therapy in the treatment of
severe sepsis and septic shock. N Engl J Med 2001; 345:1368–1377
23. Trzeciak S, Dellinger RP, Abate NL, et al: Translating research to
clinical practice: A 1-year experience with implementing early goal-
directed therapy for septic shock in the emergency department.
Chest 2006; 129:225–232
24. Kern JW, Shoemaker WC: Meta-analysis of hemodynamic optimiza-
tion in high-risk patients. Crit Care Med 2002; 30:1686–1692
25. de Oliveira CF, de Oliveira DS, Gottschald AF, et al: ACCM/PALS
haemodynamic support guidelines for paediatric septic shock: An
outcomes comparison with and without monitoring central venous
oxygen saturation. Intensive Care Med 2008; 34:1065–1075
26. Lalitha A, Vasudevan A, Shilpa R, et al: Impact of early fluid overload
in critically ill children–A prospective observational study. Pediatr Crit
Care Med 2014; 15(4 Suppl):59
27. Sutherland SM, Zappitelli M, Alexander SR, et al: Fluid overload and
mortality in children receiving continuous renal replacement therapy:
The prospective pediatric continuous renal replacement therapy reg-
istry. Am J Kidney Dis 2010; 55:316–325
28. Benakatti G, Singhi S, Jayshree M, et al: Conventional vs. restric-
tive maintenance fluid regime in children with septic shock after ini-
tial resuscitation: A randomized open label controlled trial. Arch Dis
Child 2012; 97:A5