Online Clinical Investigations
Comparison of International Pediatric Sepsis
Consensus Conference Versus Sepsis-3
Definitions for Children Presenting With Septic
Shock to a Tertiary Care Center in India:
A Retrospective Study*
Jhuma Sankar, MD; Nitin Dhochak, MD; Kiran Kumar, MD; Man Singh, MD;
M. Jeeva Sankar, MD, DM; Rakesh Lodha, MD
Objectives: To evaluate the proportion of children fulfilling
“Sepsis-3” definition and International Pediatric Sepsis Consensus
Conference definition among children diagnosed to have septic
shock and compare the mortality risk between the two groups.
Design: Retrospective chart review.
Setting: PICU of a tertiary care teaching hospital from 2014 to
2017.
Patients: Children (≤ 17 yr old) with a diagnosis of septic shock at
admission or during PICU stay.
Interventions: None.
Measurements and Main Results: We applied both International
Pediatric Sepsis Consensus Conference and the new “Sepsis-3”
definition (sepsis with hypotension requiring vasopressors and
a lactate value of ≥ 2 mmol/L) to identify cases of septic shock
by these definitions. Key outcomes such as mortality, propor-
tion attaining shock reversal at 24 hours and organ dysfunction
were compared between those fulfilling “Sepsis-3” definitions
(“Sepsis-3” group) and those fulfilling “International Pediatric
Sepsis Consensus Conference” definition (“International Pe-
diatric Sepsis Consensus Conference” group). A total of 216
patients fulfilled International Pediatric Sepsis Consensus Con-
ference definitions of septic shock. Of these, only 104 (48%;
95% CI, 42–55) fulfilled “Sepsis-3” definition. Children fulfilling
*See also p. 299.
All authors: Department of Pediatrics, All India Institute of Medical
­Sciences, New Delhi, India.
Supplemental digital content is available for this article. Direct URL
­citations appear in the printed text and are provided in the HTML and PDF
versions of this article on the journal’s website (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 © 2019 by the Society of Critical Care Medicine and the World
Federation of Pediatric Intensive and Critical Care Societies
DOI: 10.1097/PCC.0000000000001864
e122 www.pccmjournal.org March 2019 • Volume 20 • Number 3
Copyright © 2019 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
“Sepsis-3 plus International Pediatric Sepsis Consensus Confer-
ence definitions” (“Sepsis-3 and International Pediatric Sepsis
Consensus Conference” group) had lower proportion with shock
resolution (61% vs 82%; relative risk, 0.73; 95% CI, 0.62–0.88)
and higher risk of multiple organ dysfunction (85% vs 68%; 1.24;
1.07–1.45) at 24 hours. The mortality was 48.5% in “Sepsis-3
and International Pediatric Sepsis Consensus Conference” group
as compared with 37.5% in the “International Pediatric Sepsis
Consensus Conference only” group (relative risk, 1.3; 95% CI,
0.94–1.75).
Conclusions: Less than half of children with septic shock iden-
tified by International Pediatric Sepsis Consensus Conference
definitions were observed to fulfill the criteria for shock as per
“Sepsis-3” definitions. Lack of difference in the risk of mortality
between children who fulfilled “Sepsis-3” definition and those
who did not fulfill the definition raises questions on the appropri-
ateness of using this definition for diagnosis of septic shock in
children. (Pediatr Crit Care Med 2019; 20:e122–e129)
Key Words: Goldstein-B; International Pediatric Sepsis Consensus
Conference; organ dysfunction; outcomes; Sepsis-3; septic shock
S
evere sepsis and septic shock cause significant mortality
and morbidity in children (1, 2). The reported mortality
rates due to sepsis/ septic shock are about 5–32% for chil-
dren (2–5). In 2005, the International Pediatric Sepsis Con-
sensus Conference (IPSCC) definitions for sepsis and septic
shock for children with variables based on age-appropriate
values were proposed (6). These definitions—developed largely
by consensus for use in clinical trials—are widely used for diag-
nosis of sepsis and septic shock in acutely ill children (6).
Despite concerns with the nonspecific nature of systemic
inflammatory response syndrome (SIRS) and the nonspe-
cific nature of definitions on organ dysfunction in the 1991

definitions for adult patients with septic shock (7), the con-
sensus definitions remained unchanged (8). To address these
concerns with SIRS, in 2016, new definitions of sepsis—Sep-
sis-3—were proposed which defined sepsis as life-threatening
organ dysfunction caused by a dysregulated host response to
infection (9). Organ dysfunction in Sepsis-3 was defined as in-
crease in Sequential Organ Failure Assessment (SOFA) score of
2 or more from baseline or greater than 2 in patients with no
baseline score available (9). For defining septic shock, hypo-
tension (mean arterial pressure > 65 mm Hg) requiring vaso-
pressors and increased lactate (> 2 mmol/L) despite adequate
volume resuscitation were found to be the best predictors of
mortality in adults with sepsis (9, 10).
Sepsis-3 definitions were originally proposed for adults (9).
However, there is a growing interest in the pediatric critical
care community (developed or developing) if the definitions
could be applicable to children due to lack of gold standard for
identifying septic shock and the fact that the IPSCC definitions
have not been extensively validated in children. For clinicians,
it is relevant to know which definitions work and how to rec-
ognize shock (11–13).
However, there are important physiologic differences be-
tween children and adults that make it important to eval-
uate the validity of these definitions in children before being
put into use. For example, hypotension is a late feature of
septic shock in children, and the stringent Sepsis-3 criteria
are likely to miss many pediatric septic shock patients (14–
17). The limited data available from high-income countries
suggests that the Sepsis-3 definition identifies sicker chil-
dren with septic shock when compared with other defini-
tions for septic shock (3, 4, 18). No such data are available
from resource-limited settings in low- and middle-income
countries (LMICs) where children often present late in the
course of their illness and have high case fatality rates (19).
We therefore planned to evaluate whether Sepsis-3 defini-
tion identifies only the sicker children with septic shock and
whether these children have worse clinical outcomes as com-
pared with IPSCC definitions.
MATERIALS AND METHODS
Setting and Participants
This was a retrospective chart review from January 2014
to 2017 in a tertiary care PICU of a teaching hospital. We
screened case records of all admissions for identifying chil-
dren with a diagnosis of septic shock. The definitions of
septic shock used in the unit during the study period were
the IPSCC definitions. However, we wanted to ensure if all
the cases diagnosed to have septic shock actually fulfilled the
IPSCC definitions. Therefore, we applied both IPSCC and the
Sepsis-3 criteria to these children. The variables used in both
definitions were collected from the case records at the time the
child was diagnosed to be in shock (within the first 1 hr of rec-
ognition of shock). Children having PICU stay of less than 6
hours were excluded. The study was approved by the Institute
Ethics Committee.
Pediatric Critical Care Medicine www.pccmjournal.org e123
Copyright © 2019 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 Investigations
Objectives and Outcome Measures
Our primary objective was to evaluate the proportion of chil-
dren fulfilling “Sepsis-3” definition of those who were diag-
nosed to have septic shock by IPSCC definitions and compare
the mortality rates between those fulfilling “Sepsis-3 and
IPSCC” criteria (“Sepsis-3 and IPSCC” group) with those ful-
filling “IPSCC criteria” only (“IPSCC only” group). Our sec-
ondary objectives were to compare the difference between the
two groups with regard to proportion attaining shock reversal
in first 24 hours from the time of recognition of shock, use of
fluid boluses in first 24 hours, maximum Vasoactive-Inotrope
score in first 48 hours, ventilation for shock during ICU stay,
sepsis-induced myocardial dysfunction diagnosed in the first
6 hours of shock recognition, acute respiratory distress syn-
drome and acute kidney injury within first week, pediatric
SOFA (pSOFA) score on day 1 (d1 pSOFA—worst values from
1 to 24 hr of shock recognition) and day 2 pSOFA (worst values
from 24 to 48 hr of shock recognition), day 1 Pediatric Logistic
Organ Dysfunction (PELOD)–2 scores and day 2 PELOD-2
scores (same as for pSOFA scores), and PICU stay and hospital
stay between the two groups.
Study Definitions
Sepsis was defined as confirmed or suspected infection with
an acute rise in the pSOFA score of 2 points or more from 48
hours before infection to 24 hours after infection and who re-
ceived antimicrobial therapy (3, 9). Previous organ dysfunc-
tion and thereby the pSOFA score was assumed to be zero in
those in who the data was not available. Infection time was
defined as the time when the first microbiological study or
antimicrobial therapy was ordered by a physician, which-
ever was earlier (3). Septic shock was defined as patients with
sepsis and hypotension requiring vasopressor therapy and
lactate greater than 2 mmol/L despite adequate fluid resusci-
tation (9). Hypotension was defined as per pSOFA/PELOD-2
age-adapted cut-offs for mean arterial blood pressure (MBP)
(same cut-offs as per age in both, the pSOFA score had used
the age-adapted cut-offs of PELOD-2 for defining hypoten-
sion) (3, 20). We used MBP as has been proposed in original
Sepsis-3 definitions and did not try to change the same to
allow appropriate comparison between originally proposed
Sepsis-3 and IPSCC definitions.
The IPSCC and other definitions used in the study are pro-
vided in the Electronic supplement 1 (Supplemental Digital
Content 1, http://links.lww.com/PCC/A860). For defining hy-
potension as per IPSCC criteria, we used the cut-offs proposed
in the corrigendum of the original definitions by Goldstein
et al (6, 21).
Methodology
Those children whose records had the necessary information
on exposure and outcome variables for both definitions were
eligible for inclusion. We assumed the values to be normal if
two or less variables were missing. We did not want to assume
normal data for too many missing variables. If more than two
variables in pSOFA score (bilirubin, creatinine, platelets) were
 Sankar et al

missing, we did not include the data of these patients. If the RESULTS
variables for defining Sepsis-3 shock (blood pressure or lactate) We screened the records of 875 children admitted during the
or use of vasopressors were missing, we excluded those patients’ study period for suspected sepsis with or without administration
data as well. of boluses or inotropes. Of them, 289 (33%) required administra-
The management of these children was as per unit proto- tion of bolus/inotropes for suspected sepsis. Data on lactate/blood
cols, which is based on international guidelines for manage- pressures/SIRS criteria or two or more variables of pSOFA score
ment of septic shock in children (22, 23). Our unit protocol was missing for 61 (Fig. 1) who were subsequently excluded from
based on the International Guidelines for management of chil- analysis. Another 12 children (4%) who died in the first 6 hours
dren with septic shock is briefly described in the Electronic of initiating resuscitation were also excluded. The remaining 216
supplement 1 (Supplemental Digital Content 1, http://links. children in whom data were available for IPSCC and/or Sepsis-3
lww.com/PCC/A860). definitions for septic shock were included in the study.
Of the enrolled 216 children, all fulfilled IPSCC definitions
Statistical Analysis of septic shock while only 104 (48%; 95% CI, 42–55) fulfilled
Data were entered into Microsoft Excel 2007 and analyzed Sepsis-3 definitions for septic shock (Fig. 1). Key baseline char-
using Stata 11.2 (Stata Corp, College Station, TX). Catego- acteristics including age, gender, duration of illness, Pediatric
rical data are presented as n (%), whereas continuous vari- Index of Mortality-3 score, SOFA score at admission, clinical,
ables are presented as mean (sd), if normally distributed and laboratory variables were not different between “Sepsis-3
and median (interquartile range [IQR]), if skewed. Statis- and IPSCC” group and “IPSCC only” groups (Table 1; and
tical analysis was performed using Student t test/Wilcoxon Electronic supplement 2, Supplemental Digital Content 2,
rank-sum test/sign rank test and chi-square test for contin- http://links.lww.com/PCC/A861). Baseline pSOFA score was
uous and categorical variables, respectively. We calculated available for 55 patients in “Sepsis-3 group.” Median (IQR) lac-
the relative risk (RR)/mean difference with 95% CI for the tate was high in the “Sepsis-3 and IPSCC group” as compared
outcomes. We also calculated the area under the receiver with “IPSCC only” group (4 mmol/L [2.8–6.2 mmol/L] vs 1.4
operating characteristics (AUROCs) curve and sensitivity/ mmol/L [1–1.7 mmol/L]; p < 0.0001).
specificity at different cut-offs for pSOFA and PELOD scores
on days 1 and 2. We compared the time to mortality between Primary Outcome
the two groups using Kaplan-Meier survival analysis and The overall in-hospital mortality of the enrolled children was
log-rank test. 42.5% (the corresponding mortality in children with suspected
sepsis but “no” septic shock
was 12% [70/586]). In-hospital
mortality was 37.5% in “IPSCC
only” and 48.5% in “Sepsis-3
and IPSCC group” (RR, 1.3;
95% CI, 0.94–1.75; p = 0.1)
(Table 2). The median time
to death was 11 days and 15
days in the “IPSCC group” and
“Sepsis-3 and IPSCC” groups
respectively (p = 0.15) (Fig. 2).
Patients with meningitis/ en-
cephalitis were in greater pro-
portion in the “IPSCC only”
group as compared with the
“Sepsis-3 and IPSCC group”
(22 [20%] and 17 [16.5%], re-
spectively) (Table 1). Fourteen
(64%) and nine (53%) died of
refractory raised intracranial
pressure (ICP) in the “IPSCC
only” and “Sepsis-3 and IPSCC”
groups, respectively. Healthcare-
associated infections were also
more prevalent in the “IPSCC
only” group as compared with
Figure 1. Study flow. IPSCC = International Pediatric Sepsis Consensus Conference, pSOFA = pediatric
“Sepsis-3 and IPSCC” group
Sequential Organ Failure Assessment, SIRS = systemic inflammatory response syndrome. (35% vs 22%, respectively).

e124 www.pccmjournal.org March 2019 • Volume 20 • Number 3

Copyright © 2019 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 Investigations

TABLE 1. Baseline Characteristics of the Two Study Groups

Sepsis 3 and IPSCC Only
IPSCC Group Group
Variables (n = 104) (n = 112) p

Age (mo), median (IQR) 84 (27–135) 60 (24–120) 0.19

Male gender, n (%) 55 (53) 65 (58) 0.44
Weight for age (z score), median (IQR) –1.7 (–1.1 to –2.5) –1.5 (–1.2 to –2.4) 0.42
Body mass index for age (z score), median (IQR) –0.6 (–0.2 to 1.5) –0.8 (–0.3 to 1.8) 0.39
Duration of illness (d), median (IQR) 4 (2–7) 4 (2–10) 0.74
Pediatric Index of Mortality-3 score 28 (16–42) 26 (14–39) 0.99
(predicted probability of death %), median (IQR)
Pediatric Sequential Organ Failure Assessment 8 (5–10) 7 (4–9) 0.59
score at the time of diagnosis of shock, median (IQR)
Time to diagnosis/onset of shock from PICU 1 (0–26) 0 (0–22) 0.09
admission (hr), median (IQR)
  Community acquired infection, n (%) 81 (88) 84 (65) 0.62
  Nosocomial infection, n (%) 23 (22) 28 (35)
Source of infection, n (%) 0.71
 Pneumonia 33 (32) 40 (36)
  Abdominal infection 31 (30.1) 32 (29)
 Meningitis 17 (16.5) 22 (20)
  Skin and soft-tissue infection 3 (2.9) 3 (2.7)
 Tuberculosis 4 (3.8) 1 (0.9)
 Malaria 1 (0.9) 0
 Hepatitis 1 (0.9) 0
 Urosepsis 1 (0.9) 2 (1.8)
  Septicemia without focus 13 (12.5) 11 (10)
Clinical variables, mean (sd)
 Tachycardia 86 (83) 88 (79) 0.44
  Heart rate 140 (28) 143 (32) 0.52
 Tachypnea 71 (64) 78 (75) 0.19
  Respiratory rate 38 (15) 38 (16) 0.80
  Blood pressure
   Systolic blood pressure 78 (19) 96 (21) < 0.0001
   Diastolic blood pressure 37 (8) 49 (17) < 0.0001
   Mean arterial blood pressure 51 (10) 65 (16) < 0.0001
  Prolonged capillary refill time 77 (74.7) 76 (67.8) 0.26
   Capillary refill time 3 (0.9) 3 (0.8) 0.4
  Low GCS (< 15) 39 (39.4) 48 (44.4) 0.46
  GCS 10.2 (4) 10.5 (3.2) 0.54
  Core-periphery temperature differencea 77/98 (69) 64/101 (61) 0.25
Central venous pressure, median (IQR) 8 (4–11) 8.2 (6–11) 0.54
(Continued )

Pediatric Critical Care Medicine www.pccmjournal.org e125

Copyright © 2019 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

Sepsis 3 and IPSCC Only
IPSCC Group Group
Variables (n = 104) (n = 112) p

Scvo2 lowb, n (%) 28/46 (61) 19/48 (40) 0.04

 Scvo2 % (baseline) 68 (13) 75 (10) 0.01
Oxygen saturation, mean (sd) 94 (12) 96 (13) 0.01
Pao2, median (IQR) 70 (58–98) 71 (47–122) 0.10
Paco2, mean (sd) 40 (13) 37 (12) 0.11
  Proportion with base deficit > 5 mmol/L, n (%) 37 (37) 28 (25) 0.06
  Base deficit, median (IQR) 2.2 (–7.4 to 8) 0.7 (–6.6 to 5.3) 0.29
  Lactate in mmol/L, median (IQR) 4 (2.8–6.2) 1.4 (1–1.7) < 0.0001
Underlying comorbidities, n (%) 27 (41.9) 36 (39.5) 0.32
  Nephrotic syndrome 8 (12.1) 11 (10)
  Hematological malignancies 7 (10.6) 10 (11.6)
  Chronic liver disease 2 (6.1) 2 (2.3)
  Underlying immunodeficiency 7 9 (2.3)
  Neurometabolic disorders 3 (3.0) 4 (10.5)
Type of shock, n (%)
  Cold shock 77 (69) 64 (61) 0.11
  Warm shock 6 (5.4) 14 (13.5)
GCS = Glasgow Coma Scale, IPSCC = International Pediatric Sepsis Consensus Conference, IQR = interquartile range, Scvo2 = superior vena cava oxygen
saturation.
a
Data available for 98 and 101 patients in Sepsis-3 and IPSCC group and IPSCC only group respectively.
b
Data available for 46 and 48 patients in Sepsis-3 and IPSCC group and IPSCC only group respectively.

Secondary Outcomes
Proportion of children achieving therapeutic endpoints at 24
hours in the “IPSCC only” group were 82.1% (n = 92) com-
pared with 61% (n = 63) in the “Sepsis-3 and IPSCC” group
(RR, 0.73; 95% CI, 0.62–0.88; p = 0.0004) (Table 2).
pSOFA scores at 24 hours were higher in the “Sepsis-3
and IPSCC” group as compared with “IPSCC only” group
(median of 9 vs 6; p = 0.0008). Median PELOD scores at 24
hours were also higher in “Sepsis-3 and IPSCC” group (12
vs 11; p = 0.04). The AUROC curve, sensitivity, specificity,
positive predictive and negative predictive values for pSOFA
at diagnosis of shock, d1 pSOFA, d2 pSOFA, d1 PELOD,
and d2 PELOD are provided in Electronic supplement 3
(Supplemental Digital Content 3, http://links.lww.com/PCC/
A862). A cut-off of greater than or equal to 8 for pSOFA at
diagnosis of shock had a sensitivity and specificity of 80%
and 73%, respectively and a cut-off of greater than or equal
to 12 for d1 PELOD had a sensitivity and specificity of 80%
and 75%, respectively.
Maximum Vasoactive-Inotrope score in first 48 hours was
higher in the “Sepsis-3 and IPSCC” group as compared with
“IPSCC only” group (30 vs 20; p = 0.005). Proportion requir-
ing mechanical ventilation was also higher in “Sepsis-3 and
IPSCC” group as compared with “IPSCC only” group (73% vs
59%; p = 0.03). Median duration of ICU stay was similar in
both groups (p = 0.71) (Table 2).
The amount of fluid boluses and total fluids received in mL/kg
in the first 24 hours was not significantly different between
the groups in the first 24 hours. However, greater propor-
tion of children in the “Sepsis-3 and IPSCC” group had fea-
tures of fluid overload (16% vs 6%; 2.8; 1.01–7.6; p = 0.036)
at 24 hours as compared with the “IPSCC only” group. Need
for dialysis and blood transfusion was also greater in “Sepsis-3
and IPSCC” group as compared with “IPSCC only” group
(Electronic ­supplement 4, Supplemental Digital Content 4,
http://links.lww.com/PCC/A863).
DISCUSSION
Our study findings suggest that the new Sepsis-3 definition for
septic shock identifies only sicker children with septic shock—
only 48% of patients among those identified by the IPSCC
definition were classified as having septic shock by the new
Sepsis-3 definition. This may limit the utilization of Sepsis-3
definition as operating definitions for diagnosis and treat-
ment of septic shock in children. Although the IPSCC defini-
tions may be oversensitive and may result in overtreatment
of patients which have nonseptic disease (24), the Sepsis-3

e126 www.pccmjournal.org March 2019 • Volume 20 • Number 3

Copyright © 2019 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 Investigations

TABLE 2. Primary and Secondary Outcomes in the Study Groups

Sepsis 3 and IPSCC Only
IPSCC Group Group Relative Risk
Variables (n = 104) (n = 112) (95% CI) p

Primary outcome, n (%)

  In-hospital mortality 50 (48.5) 42 (37.5) 1.3 (0.94–1.75) 0.10
Secondary outcomes
  Proportion achieving therapeutic end points in first 6 hr, n (%) 26 (25.3) 46 (41) 0.60 (0.4–0.9) 0.006
  Proportion achieving therapeutic end points in first 24 hr, n (%) 63 (61) 92 (82.1) 0.73 (0.62–0.88) 0.0004
  Time to achieving therapeutic end points of shock, hr, 5.5 (5–23) 5 (3.5–22) — < 0.0001
median (IQR)
  Number of boluses in 1st hour, median (IQR) 1 (1–2) 1 (1–2) — 0.47
  Number of boluses in 24 hr, median (IQR) 1 (1–2) 2 (1–2) — 0.35
  Vasoactive-Inotrope Score, median (IQR) 30 (20–41) 20 (10–36) — 0.005
Organ system failure
  Proportion with Multiple Organ Dysfunction Syndrome in 88 (85) 76 (68) 1.24 (1.07–1.45) 0.0008
first 24 hr, n (%)
  Number of organ system involved at 6 hr, median (IQR) 2 (2–3) 2 (1–3) — 0.01
  Number of organ system involved at 24 hr, median (IQR) 2 (2–3) 2 (1–3) — 0.005
  Number of organ system involved at 48 hr, median (IQR) 2 (1.5–3) 2 (1–3) — 0.017
  pSOFA score at 24 hr, median (IQR) 9 (6–11) 6 (4–8) — 0.0008
  pSOFA score at 48 hr, median (IQR) 8 (4–12) 5 (2–7) — 0.0003
  PELOD score at 24 hr, median (IQR) 12 (2–22) 11 (1–21) — 0.04
  PELOD score at 48 hr, median (IQR) 12 (1–30) 10 (1–11) — 0.003
  Sepsis-induced myocardial dysfunction , n (%)
a 10/16 (63) 5/18 (28) 0.44 (0.19–1.02) 0.04
  Ejection fraction, median (IQR) 0.45 (0.42–0.57) 0.6 (0.54–0.6) — 0.02
  Acute respiratory distress syndromea, n (%) 21/94 (22.3) 13/102 (13) 0.57 (0.30–1.07) 0.07
  Acute renal failure, n (%) 50 (48) 27 (24) 0.50 (0.34–0.73) < 0.0001
  Need for mechanical ventilation during ICU stay, n (%) 76 (73) 66 (59) 1.24 (1.02–1.5) 0.03
  Duration of mechanical ventilation, hr, median (IQR) 111 (72–192) 72 (48–120) — 0.07
  Clinical features of fluid overload at 24 hr, n (%) 13/82 (16) 5/85 (6) 2.8 (1.01–7.6) 0.036
  Duration of inotrope therapy, hr, median (IQR) 48 (24–96) 48 (24–96) — 0.37
  Duration of ICU stay, d, median (IQR) 5 (3–9) 5 (3–9) — 0.71
IPSCC = International Pediatric Sepsis Consensus Conference, IQR = interquartile range, PELOD = Pediatric Logistic Organ Dysfunction, pSOFA = pediatric
Sequential Organ Failure Assessment.
a
Row percentage.

Boldface values indicate significant results.

definition may result in underdiagnosis of children with septic
shock and may lead to undertreatment.
Our study findings are similar to a previous report by
Schlapbach et al (4) that identified only 10% of patients as
having septic shock of all the patients identified to have shock
as per IPSCC criteria (75/748) (17). Matics and Sanchez-Pinto
(3) and Leclerc et al (18) have reported the prevalence of
septic shock as per Sepsis-3 criteria to be 4% and 8%, respec-
tively. Matics and Sanchez-Pinto (3) used an Electronic Health
Record derived single institution PICU database from 2009 to
2016, whereas Leclerc et al (18) used the historic PELOD mul-
ticenter cohort (18) and Schlapbach et al (4) used a binational
ICU registry from 2012 to 2015. In the present study, we used
retrospective data from a busy PICU in a LMIC.
Studies in adult patients with septic shock have demon-
strated slightly higher sensitivity of Sepsis-3 criteria among
Sepsis-1 or Sepsis-2 diagnosed septic shock patients from
42.5% to 79.2% (25–28). The possible explanation lies in the

Pediatric Critical Care Medicine www.pccmjournal.org e127

Copyright © 2019 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
Figure 2. Kaplan-Meier survival analysis. IPSCC = International Pediatric
Sepsis Consensus Conference.
differences in pathophysiology of septic shock among children
and adults. Although hypotension is an early feature in adults
(29), it is often a late feature in pediatric septic shock (15, 16).
Also, defect in oxygen extraction has been demonstrated in
adults with septic shock while children may predominantly
have a defect in oxygen delivery. This may lead to anaerobic
glycolysis and lactate elevation much earlier in adults than in
children (15–17). Sepsis-3 definition, which is based on hypo-
tension and lactate, may therefore not be sensitive in identify-
ing pediatric septic shock as in adults.
The mortality in the “Sepsis-3 and IPSCC” group was 48.5%
as compared with 37.5% in the “IPSCC only” group (RR, 1.3;
95% CI, 0.94–1.75). The mortality rates reported in children are
13% in the IPSCC group (4) and about 32–32.5% in the Sepsis-3
group (3, 18). In comparison to septic shock by either of these
definitions, mortality was 12% in the nonshock patients.
The lack of a statistically significant difference in mortality
in our study could be explained by the following: 1) the pro-
portion of children with meningitis/encephalitis and raised
ICP was higher in the “IPSCC only” group than the “Sepsis-3
and IPSCC” group; 2) the proportion with HCAI was also
higher eventually leading to greater proportion of children
dying in the “IPSCC only” group; and 3) the small numbers
studied. Larger studies would be required to evaluate this asso-
ciation in children.
Although the mortality difference of 10% may not be statis-
tically significant, it is still clinically important. But the more
concerning observation was that the mortality was almost
40% even in the “IPSCC only” group. This implies that these
patients would be underdiagnosed had we applied only the
Sepsis-3 definitions to the cohort and would be at risk of poor
prognosis due to late diagnosis of shock. In comparison to
septic shock by either of these definitions, mortality was 12%
(n = 70) in the nonshock patients. With the literature avail-
able till date, it is becoming increasingly apparent that directly
using these definitions for recognition and treatment of septic
shock in children may not be appropriate. We need further
e128 www.pccmjournal.org March 2019 • Volume 20 • Number 3
Copyright © 2019 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
research on what variables should be/could be used to define
septic shock in children.
The area under the curve (AUC) for pSOFA at admission
and in first 24 hours in the “Sepsis 3 group” in our cohort
was greater than 0.80 suggesting that the pSOFA was a good
tool to predict outcomes based on organ dysfunction cohort.
The AUC for d1 PELOD-2 score was also good (> 0.80). Our
findings are similar to Schlapbach et al (30) who reported
AUROC of 0.829 for age-adjusted SOFA (pSOFA) and 0.816
for PELOD-2. Matics and Sanchez-Pinto (3) reported AUC of
greater than 0.90 for both pSOFA and PELOD-2 scores. Leclerc
et al (18) reported the AUC of the d1 PELOD-2 score to be 0.88
in those with low systolic blood pressure and hyperlactatemia
and 0.91 in those with low MBP and hyperlactatemia (Sepsis-3
criteria) (further evaluation of both scores would be desirable
before they can be widely accepted as reliable tools for defining
organ dysfunction in children with septic shock). However, the
laboratory intensive nature of the scores may preclude their
use in non-PICU settings, in particular in LMICs even they are
found to be reliable (13).
The present study is one of the few studies that made head-
to-head comparison of IPSCC definitions (which have been the
standard definitions in children for clinical or research purposes
till date) with the new Sepsis-3 definitions, particularly from a
LMIC setting. It is also one of the first few studies to test the pe-
diatric adaptation of Sepsis-3 septic shock definitions using the
pSOFA for organ dysfunction. The other strengths are—focus
on shock recognition and use of data from a busy Indian unit.
The findings are not very encouraging even in a setting where
patients often present late in the course of their illness. With the
published literature available till date, it is becoming increas-
ingly apparent that directly using these definitions for recog-
nition and treatment of septic shock in children may not be
appropriate. We need further research on what variables should
be/could be used to define septic shock in children.
The major limitation of our study is that it was a retro-
spective, single-center study; but our unit is fairly representa-
tive of tertiary care ICUs from the Indian subcontinent. Data
were missing or incomplete for 61 patients. We used the pSOFA
score which has not been extensively validated in pediatrics. The
results of the present study may not be generalizable to the de-
veloped nations where the case mix may be different, and the di-
sease burden is also low. Finally, the small sample size could have
resulted in lack of a statically significant difference in mortality
despite an absolute difference of 11% between the two groups.
CONCLUSIONS
Less than half of children with septic shock identified by
IPSCC definitions were observed to fulfill the criteria for shock
as per the “Sepsis-3” definitions. These children were sicker as
compared with “IPSCC only” group and required more rig-
orous supportive therapies. Lack of significant difference in
the risk of mortality between children who fulfilled “Sepsis-3”
definition and those who did not fulfill the definition raises
questions on the appropriateness of using this definition in
children with septic shock.

ACKNOWLEDGMENTS
We thank Rohit Sharma and Jatin Singh for their help with the
study.
REFERENCES
1. Weiss SL, Fitzgerald JC, Pappachan J, et al; Sepsis Prevalence,
Outcomes, and Therapies (SPROUT) Study Investigators and
Pediatric Acute Lung Injury and Sepsis Investigators (PALISI)
Network: Global epidemiology of pediatric severe sepsis: The sepsis
prevalence, outcomes, and therapies study. Am J Respir Crit Care
Med 2015; 191:1147–1157
2. Wolfler A, Silvani P, Musicco M, et al; Italian Pediatric Sepsis Study
(SISPe) group: Incidence of and mortality due to sepsis, severe sepsis
and septic shock in Italian Pediatric Intensive Care Units: A prospec-
tive national survey. Intensive Care Med 2008; 34:1690–1697
3. Matics TJ, Sanchez-Pinto LN: Adaptation and validation of a Pediatric
Sequential Organ Failure Assessment score and evaluation of the
sepsis-3 definitions in critically ill children. JAMA Pediatr 2017;
171:e172352
4. Schlapbach LJ, MacLaren G, Festa M, et al; Australian & New Zealand
Intensive Care Society (ANZICS) Centre for Outcomes & Resource
Evaluation (CORE) and Australian & New Zealand Intensive Care
Society (ANZICS) Paediatric Study Group: Prediction of pediatric
sepsis mortality within 1 h of intensive care admission. Intensive Care
Med 2017; 43:1085–1096
5. Wiens MO, Larson CP, Kumbakumba E, et al: Application of sepsis
definitions to pediatric patients admitted with suspected infections in
Uganda. Pediatr Crit Care Med 2016; 17:400–405
6. Goldstein B, Giroir B, Randolph A; International Consensus
Conference on Pediatric Sepsis: International pediatric sepsis con-
sensus conference: Definitions for sepsis and organ dysfunction in
pediatrics. Pediatr Crit Care Med 2005; 6:2–8
7. American College of Chest Physicians/Society of Critical Care
Medicine Consensus Conference: Definitions for sepsis and organ
failure and guidelines for the use of innovative therapies in sepsis. Crit
Care Med 1992; 20:864–874
8. Levy MM, Fink MP, Marshall JC, et al: 2001 SCCM/ESICM/ACCP/
ATS/SIS international sepsis definitions conference. Intensive Care
Med 2003; 29:530–538
9. Singer M, Deutschman CS, Seymour CW, et al: The third international
consensus definitions for sepsis and septic shock (sepsis-3). JAMA
2016; 315:801–810
10. Shankar-Hari M, Phillips GS, Levy ML, et al; Sepsis Definitions Task
Force: Developing a new definition and assessing new clinical criteria
for septic shock: For the third international consensus definitions for
sepsis and septic shock (sepsis-3). JAMA 2016; 315:775–787
11. Piva JP, Garcia PC: Sepsis: From the stone age to nowadays without
a precise definition. Pediatr Crit Care Med 2016; 17:794–795
12. Kawasaki T: Update on pediatric sepsis: A review. J Intensive Care
2017; 5:47
13. Schlapbach LJ, Kissoon N: Defining pediatric sepsis. JAMA Pediatr
2018; 172:312–314
14. Davis AL, Carcillo JA, Aneja RK, et al: American College of Critical
Care Medicine clinical practice parameters for hemodynamic
Pediatric Critical Care Medicine www.pccmjournal.org e129
Copyright © 2019 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 Investigations
support of pediatric and neonatal septic shock. Crit Care Med 2017;
45:1061–1093
15. Pollack MM, Fields AI, Ruttimann UE: Distributions of cardiopulmo-
nary variables in pediatric survivors and nonsurvivors of septic shock.
Crit Care Med 1985; 13:454–459
16. Pollack MM, Fields AI, Ruttimann UE: Sequential cardiopulmonary
variables of infants and children in septic shock. Crit Care Med 1984;
12:554–559
17. Carcillo JA, Pollack MM, Ruttimann UE, et al: Sequential physiologic
interactions in pediatric cardiogenic and septic shock. Crit Care Med
1989; 17:12–16
18. Leclerc F, Duhamel A, Deken V, et al; Groupe Francophone de
Réanimation et Urgences Pédiatriques (GFRUP): Can the Pediatric
Logistic Organ Dysfunction-2 score on day 1 be used in clin-
ical criteria for sepsis in children? Pediatr Crit Care Med 2017;
18:758–763
19. Sankar J, Lodha R, Kabra SK: Management of septic shock: Where
do we stand? Indian J Pediatr 2008; 75:1167–1169
20. Leteurtre S, Duhamel A, Salleron J, et al; Groupe Francophone de
Réanimation et d’Urgences Pédiatriques (GFRUP): PELOD-2: An
update of the PEdiatric logistic organ dysfunction score. Crit Care
Med 2013; 41:1761–1773
21. Brahm G, Brett G, Adrienne R: Reply: Values for systolic blood pres-
sure. Pediatr Crit Care Med 2014; 15:e157–e167
22. 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
23. 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 se-
vere sepsis and septic shock, 2012. Intensive Care Med 2013;
39:165–228
24. Klompas M, Calandra T, Singer M: Antibiotics for sepsis-finding the
equilibrium. JAMA 2018; 320:1433–1434
25. Sterling SA, Puskarich MA, Glass AF, et al: The impact of the sepsis-3
septic shock definition on previously defined septic shock patients.
Crit Care Med 2017; 45:1436–1442
26. Besen BAMP, Romano TG, Nassar AP Jr, et al: Sepsis-3 definitions
predict ICU mortality in a low-middle-income country. Ann Intensive
Care 2016; 6:107
27. SepNet Critical Care Trials Group: Incidence of severe sepsis and
septic shock in German intensive care units: The prospective, multi-
centre INSEP study. Intensive Care Med 2016; 42:1980–1989
28. Driessen RGH, van de Poll MCG, Mol MF, et al: The influence of a
change in septic shock definitions on intensive care epidemiology and
outcome: Comparison of sepsis-2 and sepsis-3 definitions. Infect Dis
(Lond) 2018; 50:207–213
29. Burgdorff AM, Bucher M, Schumann J: Vasoplegia in patients with
sepsis and septic shock: Pathways and mechanisms. J Int Med Res
2018; 46:1303–1310
30. Schlapbach LJ, Straney L, Bellomo R, et al: Prognostic accuracy of
age-adapted SOFA, SIRS, PELOD-2, and qSOFA for in-hospital
mortality among children with suspected infection admitted to the in-
tensive care unit. Intensive Care Med 2018; 44:179–188