University of Southern California Dissertations and Theses

Association of Pediatric Early Warning Score with early intensive care unit readmission

(USC Thesis Other)

Association of Pediatric Early Warning Score with early intensive care unit readmission

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content ii

ASSOCIATION OF PEDIATRIC EARLY WARNING SCORE WITH EARLY
INTENSIVE CARE UNIT READMISSION

By

Francine Denise Bynum, MD

A Thesis Presented to the
FACULTY OF THE USC KECK SCHOOL OF MEDICINE
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
CLINICAL AND BIOMEDICAL INVESTIGATION

May 2019

Copyright 2019 Francine Denise Bynum
ii

Acknowledgements

I would like to thank my mentors, Dr. Britni Belcher, Dr. Sarah Rubin, and Dr. Lynn Miller who
have been so patient and supportive with assisting me with my manuscript. A special thanks to
Dr. Rubin, who helped to guide me when this project was just an idea and sparked my interest in
research.

iii

Table of Contents

Acknowledgements ii
List of Tables iv
List of Figures v
Abstract vi
Introduction 1
Chapter One: Background 3
Early Warning Scores
Pediatric Early Warning Scores
PEWS at Children's Hospital Los Angeles
Research Objectives

Chapter Two: Methods 9
Data Collection
Data Analysis

Chapter Three: Results 12

Chapter Four: Discussion 15

Chapter Five: Conclusion 18

References 19

iv

Lists of Tables

Table 1. Comparisons of PEWS Systems
Table 2. Normal Vital Signs for Age
Table 3. Descriptive statistics by ICU Readmission vs. No ICU Readmission
Table 4. Association of PEWS with ICU readmission
Table 5. Utility of cutoff PEWS scores (PEWS prior to ICU

v

Lists of Figures

Figure 1: CHLA Modified Pediatric Early Warning System (PEWS) Score
Figure 2: PEWS Scores and ICU Readmission

vi
Abstract

Introduction: There are currently no clear guidelines to determine risk of ICU readmission when
a child is transferred from intensive care units (ICU) to a Pediatric Ward. It is known that
children readmitted to the ICU within a single hospital stay are at increased risk for poor clinical
and financial outcomes. The development of a clinical tool to help medical staff assess a child’s
risk of deterioration, requiring ICU readmission has been a common goal among pediatricians.

Objective: To evaluate the association of a Pediatric Early Warning Score (PEWS) with early
unplanned readmission to the Cardiothoracic Intensive Care Unit (CTICU) or Pediatric ICU
(PICU) from a General Care Pediatric Ward.

Study Design: Single-center case-control study at a tertiary care children's hospital. Children ≤
18 years transferred from the PICU or CTICU to the Pediatric Ward from 1/1/10-3/30/13 were
screened. Cases were defined as unplanned PICU or CTICU readmission from the Pediatric
Ward within 48 hours of transfer out of the ICU. There were 84 ICU readmissions (PICU=38,
CTICU=46). Those in the control group (N= 253) were not readmitted. Goal case to control
ratio was 1:3. A PEWS score assigned in the PICU or CTICU (≤4 hours prior to transfer) and
first PEWS assigned on the Pediatric Ward (≤4 hours after transfer) were collected for all
patients. Median PEWS for cases vs. controls were compared using the Wilcoxon Rank Sum
test. Logistic regression was used to calculate odds ratios and 95% confidence intervals and
determine area under the receiver operating characteristic curve (ROC).

vii
Results: Age and sex were similar among cases and controls (p>0.05). Among patients with
complex chronic conditions: Cardiovascular p-value= 0.001; Respiratory p-value <0.001; and
Gastrointestinal p-value= 0.003, indicating PEWS parameters leading to increased likelihood of
ICU readmission. PEWS prior to ICU discharge and first PEWS on the Pediatric Ward were
higher for cases vs. controls, [median(IQR), 3(2,3) vs. 2(1,3); p<.001] and [2.5(2,3) vs.
2(1,3);p<.001], respectively.

Conclusions: Higher PEWS prior to ICU discharge and first PEWS on the Pediatric Ward were
associated with increased risk of unplanned ICU readmission and may help identify children at
risk for early PICU or CTICU readmission. Including chronic disease components and patient
origin in future iterations of PEWS scores may increase predictive ability and warrants further
investigation.

1

Introduction

There is no protocol in place to determine when a child is ready to be transferred from an
Intensive Care Unit (ICU) to the Pediatric Ward. While individual physician evaluation is
usually appropriate, there is a movement towards more objective indicators of readiness for
transfer, such as elevation or change in PEWS score. The development of objective evaluation
tools, such as PEWS, that when combined with clinical action, may improve recognition and
treatment of children to prevent clinical deterioration is a common goal.
Pediatric Early Warning Scores (PEWS) have been developed as screening tools to assist
with improved early recognition of children at risk for clinical deterioration (usually defined as
need for ICU admission) on the Pediatric Ward (Parshuram, Duncan, Joffe, Farrell, Lacroix,
Middaugh, Hutchison, Wensley, Blanchard, Beyene, Parkin, 2011; Akre, Finkelstein, Erickson,
Liu, Vanderbilt, Billman, 2010). Failure to recognize impending deterioration often results in
urgent transfers to the ICU. Previous studies have shown early warning scores assigned to
patients admitted to a Pediatric Ward can identify high-risk patients, enabling early intervention
to prevent clinical decompensation (Subbe, Kruger, Rutherford, Gemmel, 2001). However, there
is little information on how effective PEWS is in an ICU environment, particularly when used to
determine when patients are appropriate for transfer to a general Pediatric Ward.
Patients discharged from the ICU to the Pediatric Ward are at higher risk of clinical
deterioration and/or life-threatening events compared to other ward patients (Andréasson,
Herlitz, Bång, Ekström, Lindqvist, Lundström, Holmberg, 1998; Penk, Loke, Waloff, Frank,
Stockwell, Spaeder, Berger, 2015). In addition, patients with unplanned admissions have been
shown to have increased mortality and morbidity (Odetola, Clark, Dechert, Shanley, 2007).

2

Determining when a patient is stable to be transferred out of the ICU requires thorough
clinical evaluation. PEWS assists with these evaluations and has been shown to predict early
deterioration of patients on the Pediatric Ward (Subbe et al, 2001; Skaletzky, Raszynski,
Totapally, 2011). We are investigating whether PEWS may also assist in identifying children
that are transferred out of the ICU to the Pediatric Ward who are at high risk for early (within 48
hours) ICU readmission. Early identification of high-risk patients may impact decisions about
which children are safely able to be transferred from the ICU to the Pediatric Ward. In addition,
adding complex chronic conditions (CCC) to PEWS scores may better identify children who are
at increased risk for readmission (Duncan, Hutchison, & Parshuram, 2006).

3

Chapter One: Background

EARLY WARNING SCORES. Early warning scores were initially developed in the adult
population after it was clear that changes in physiological parameters preceded catastrophic
events (Subbe et al, 2001; Sax, & Charlson, 1987; Smith & Wood 1998). It has been shown that
prior to being transferred to the ICU, more than 50% of patients experience between 8 to 48
hours of life threatening changes in physiologic parameters. (Hillman, Chey, Daffurn, Jacques,
Norman, Bishop, Simmons, 2002). Common vital sign changes include hypotension and
tachycardia (Hillman et al, 2002). These vital sign changes often go unnoticed. Subsequent
studies showed that an objective scoring system may help identify patients at risk for clinical
decompensation (Subbe et al, 2001; Whittington, White, Haig, Slock, 2007; Smith, Chiovaro,
O'Neil, Kansagara, Quiñones, Freeman, Motu'apuaka, Slatore, 2014). Multiple different versions
of early warning scores have since been developed or adapted over time for various at-risk
populations (e.g., those with CCC, oncology patients, and cardiac patients). A systematic review
of adult early warning scores revealed that moderate changes in vital signs may predict cardiac
arrest and death within 48 hours (Smith, et al, 2014). The addition of a rapid response team to
prompt early clinical interventions in response to these abnormalities has been shown to decrease
cardiopulmonary arrests (Buist, Moore, Bernard, Waxman, Anderson, Nguyen, 2002; Chan,
Khalid, Longmore, Berg, Kosiborod, Spertus, 2008).
Pediatric studies report that failure to recognize signs and symptoms of clinical
decompensation, lack of knowledge, and failure to seek advice were listed as major causes of
urgent ICU transfer and poor outcomes (McQuillan, Pilkington, Allan, Taylor, Short, Morgan,
Nielsen, Barrett, Smith, Collins, 1998.). In response to this, pediatric early warning scores, which

4
account for age-based metrics, have since been developed (Duncan, Hutchison, Parshuram,
2006; Parshuram, Bayliss, Reimer, Middaugh, and Blanchard, 2011).

PEDIATRIC EARLY WARNING SCORES. The first PEWS score was developed by Alan
Monaghan in 2005 and was referred to as the Brighton PEWS score. The Brighton PEWS both
assisted in early detection of deteriorating children and improved communication amongst the
staff to facilitate timely clinical care (Monaghan, 2005). The Monaghan scoring system included
nurse evaluation of behavior (playing, sleeping, irritability, or lethargy), cardiovascular status
(capillary refill, color, heart rate), and respiratory status (rate, effort, and amount of oxygen),
assigning a score from 1-3 in each category based on the number of abnormal characteristics
observed. Unlike more contemporary PEWS scores, the Brighton score included additional
points for persistent vomiting after surgical intervention or use of persistent nebulizers. A total
PEWS score greater than four or a score of three in any of the categories would dictate actions of
the nursing staff which could include notifying the charge nurse, increasing the frequency of
clinical assessment, calling for a medical review, or notifying the medical team. This study
showed that after implementation of PEWS, providers had an increased confidence in
recognition of children at risk of clinical deterioration (Monaghan, 2005).

PEWS AT CHILDREN’S HOSPITAL LOS ANGELES. Since 2005, multiple PEWS scores
with different components have been developed across children’s hospitals. This high degree of
variability makes comparison of outcomes based on PEWS difficult. Most PEWS include three
core components: heart rate, respiratory rate, and oxygen saturations (Roland, Oliver, Edwards,
Mason, Powell, 2014). Children’s Hospital Los Angeles (CHLA) implemented its version of

5

PEWS in 2011. When determining what components would most accurately assist in identifying
patients at risk for deterioration in our specific patient population, several factors needed to be
considered. We understood that the prevalence of children with complex chronic conditions
and/or a reliance on technology (tracheostomies with or without ventilator dependence and
gastrostomy tubes) represents a population of children that has been growing among hospitals
across the country (Graf, Montagnino, Hueckel, McPherson, 2008; Cohen, Kuo, Agrawal, Berry,
Bhagat, Simon, Srivastava, 2011). In addition, the validated Duncan PEWS score that
incorporated complex chronic conditions, demonstrated a high prediction level for acute clinical
deterioration with an AUC of 0.90 (Duncan et al, 2006). However, the complexity of the
Duncan PEWS provided obstacles for many institutions, and thus it has not been widely adopted
(Table 1).

6

Table 1: Comparisons of PEWS Systems
BRIGHTON DUNCAN PEWS CHLA PEWS
Behavior Behavior
Cardiovascular: Heart rate,
Color, Perfusion
*Cardiovascular: Heart Rate *Cardiovascular: Heart rate,
Color, Perfusion
Respiratory: Rate, Effort,
amount of oxygen
*Respiratory: Respiratory Rate *Respiratory: Rate, Effort,
amount of oxygen
Medical History: Add 2 points *Blood Pressure Medical History: Add 1 point
a) For Nebulizers q 4 hours Pulses a) RRT or code in last 2 weeks
b) Or persistent vomiting
following surgery
O2 saturations b) Transfer to/from CTICU or
PICU
Capillary refill
Loss of consciousness - GCS
Oxygen therapy
Bolus Fluid
Temperature
Abnormal airway
Home oxygen
Any previous admission to ICU
Central venous line in situ
Transplant recipient
Severe cerebral palsy
Gastrostomy tube
Greater than 3 medical
specialties

Max Score prior to additions: 9 Max Score: 34 Max Score prior to additions: 9
* Age related

The PEWS tool adopted at CHLA has four components (Figure 1) that include most
Brighton PEWS components with the exception of postoperative vomiting and use of nebulizers.
The CHLA PEWS includes evaluation of a child’s behavior (playful, irritable, lethargic),
cardiovascular system (abnormal heart rate, prolonged capillary refill, color of skin, or single
ventricle physiology), respiratory system (respiratory rate, respiratory effort, supplemental
oxygen required), and also adds medical history [Rapid Response Team (RRT) consult, Code
Blue event or transfer from PICU or Cardiothoracic Intensive Care Unit (CTICU) at CHLA

7

within the previous two weeks]. The maximum score for the behavior, cardiovascular, and
respiratory system components is 3. The maximum medical history component score is 1. The
PEWS committee at CHLA determined that a clinically relevant event was described as
decompensation resulting in need for transfer to the Pediatric Intensive Care Unit (PICU) or
Cardiothoracic Intensive Care Unit (CTICU). Total PEWS scores could range from 0 to 10 with
higher PEWS scores indicating greater risk of clinical deterioration and subsequent ICU
admission. Increased scores mandated medical team evaluation and discussion of the medical
plan amongst physicians, nurses and respiratory care providers.
Figure 1

ICU READMISSIONS. The transfer of patients out of the ICU is often a difficult decision.
While every effort is made to ensure that patients are stable for transfer, there are circumstances
that necessitate early readmission. The national average of ICU readmissions within 48 hours

8
ranges from 2-8% (Brown, Ratcliffe, Kahn, Halpern, 2012; Linton, Grant, Pellegrini, Davidson,
2009; Campbell, Cook, Adey, Cuthbertson, 2008). ICU readmission is associated with a longer
ICU length of stay and increased mortality (Odetola et al, 2007). ICU readmission is also
associated with increased costs per patient (Lone, Gillies, Haddow, Dobbie, Rowan, Wild,
Murray, Walsh, 2016). PEWS have been utilized to detect acute deterioration of patients
(Parshuram et al, 2011; Duncan et al, 2006; Monaghan, 2005), but there is little information as to
how PEWS scores prior to ICU transfer can be utilized to identify if patients to be transferred to
the Pediatric Ward are at risk for early ICU readmission within 48 hours.

Research Objective: The primary objective of this study was to evaluate the association
between PEWS assigned at ICU discharge and first PEWS assigned on the Pediatric Ward with
early, unplanned ICU (PICU and CTICU) readmission. The secondary objective is to evaluate
the association of PEWS when chronic disease components, including technology dependence,
are included with early, unplanned ICU readmission.

9

Chapter Two: Methods

After institutional review board appr oval, we conducted a single-center case-control
study to evaluate the association between PEWS at Pediatric Intensive Care Unit (PICU) or
Cardiothoracic Intensive Care Unit (CTICU) transfer with to the Pediatric Ward and first PEWS
on the Pediatric Ward after ICU transfer with risk of early unplanned PICU or CTICU
readmission at a tertiary care, academic, free-standing children’s hospital. All patients ≤ 18
years of age transferred from a 24-bed PICU or 24-bed CTICU to the Pediatric Ward from
January 1, 2010 through March 31, 2013 were screened. Patients who were discharged directly
from either ICU to home or transferred to another ICU were excluded. For patients who were
readmitted to either ICU multiple times during the study period, only the first readmission was
included.
Cases and control patients were identified from a hospital ICU billing database. Cases
were defined as having unplanned PICU or CTICU readmission within 48 hours of transfer to
the Pediatric Ward. Controls were defined as children who were not readmitted to either ICU
within 48 hours after transfer from an ICU to the pediatric ward. Potential controls were
identified from the entire cohort of children transferred from either ICU to the pediatric ward
during the study period who were not readmitted within 48 hours. Case to control ratio was 1:3.
PEWS assigned in the PICU or CTICU (≤4 hours prior to transfer) and first PEWS assigned on
the Pediatric Ward (≤4 hours after transfer) were collected and extracted from the electronic
medical record. This was done to determine relative stability of patients immediately after
leaving the ICU, once patients arrived on the Pediatric and to ensure that different nurses were
assigning the score similarly.

10

PEWS scores are routinely assigned for each patient within 4 hours prior to PICU
discharge by a PICU nurse. When a child is transferred from the PICU to the Pediatric Ward, a
second PEWS score is assigned on the Pediatric Ward within the first hour after transfer from the
ICU by a ward nurse. Each PEWS score is documented in the electronic health record.
However, PEWS scores are not routinely done on patients prior to being transferred from the
CTICU to the Pediatric ward in our hospital. Therefore, study investigators did chart reviews to
ascertain vital signs and other PEWS components to determine the first PEWS on the Pediatric
Ward floor, using vital signs and nursing documentation (Table 2). Nursing assessments
included behavior (level of consciousness), capillary refill, skin color, and respiratory effort.
PEWS were assigned and compared between investigators to ensure consistency.
Table 2. Normal Vital Signs for Age
Heart Rate Respiratory Rate
0-3 months 95-160 30-60
4-11 months 85-150 25-50
1-4 years 80-130 20-40
5-12 years 70-110 20-30
≥13 years 60-100 12-20

Electronic health records were reviewed by two investigators to ensure completeness and
accuracy in extracting acute and chronic diagnoses. Information was initially extracted from the
Cerner-EHR based problem list and verified through clinical note review by both investigators.
When there were differences of opinion, mutual agreement was determined after careful
assessment and review. Complex chronic conditions were categorized according to those

11

established by Feudtner et al (Feudtner, Christakis, & Connell, 2000). Feudtner categories
include cardiovascular, malignancies, neuromuscular, genetic, respiratory, renal, hematologic
and immunologic, metabolic, and gastrointestinal. Additional complex chronic diagnoses
included technology dependence (tracheotomy, gastrostomy), and home nasal gastric tubes.

Data Analysis: The goal of our analysis was to determine the association between PEWS
assigned at ICU discharge and first PEWS on the Pediatric ward with unplanned ICU
readmission. Since PEWS was not normally distributed, we compared median scores for cases
and controls using Wilcoxon Rank Sum test. The secondary goal was to evaluate if chronic
conditions contributed to unplanned ICU readmission. Univariate analysis identified potential
risk factors for early ICU readmission. Variables with p-values ≤ 0.05 in the univariate analysis
were eligible for inclusion in the multiple logistic regression model. Backward stepwise logistic
regression was used to calculate odds ratios, 95% confidence intervals and determine area under
the receiver operating characteristic curve (AUC) for PEWS association with unplanned ICU
readmission. Statistical analysis was performed using Stata 12 (StataCorp. 2011. Stata Statistical
Software: Release 12. College Station, TX: StataCorp LP).

12

Chapter Three: Results

There were 84 patients readmitted to the ICU after transfer to the Pediatric ward (N
(PICU)= 38, N (CTICU)= 46). Age and sex were not statistically different among cases and
controls (p>0.05; Table 3). In patients with complex care disease, those with Cardiovascular (p=
0.001); Respiratory (p <0.001); and Gastrointestinal disease (p= 0.003) are more likely to be
readmitted to the ICU. Patient origin refers to where the patient was likely to be readmitted
from. Patients from the Pediatric Ward (p= 0.002) were most likely to be readmitted to the ICU.
PEWS prior to ICU discharge and first PEWS on the Pediatric Ward were higher for cases vs.
controls, [median(IQR), 3 (2,3) vs. 2 (1,3); p <0.001] and [2.5 (2,3) vs. 2 (1,3); p <0.001],
respectively (Figure 2).
Univariate analysis demonstrated that higher PEWS scores prior to ICU discharge and on
the Pediatric Ward were associated with increased risk of ICU readmission (OR (95% CI): 1.62
(1.29-2.03) and OR 1.76 (1.39-2.22), respectively; Table 4). Univariate analysis also
demonstrated that weight, gastrostomy, complex chronic diseases (e.g., cardiovascular,
respiratory, and gastrointestinal disease and patient origin, meaning where the patient was
originally transferred to the ICU from such as the ED, outside hospital, pediatric ward, etc.) were
all associated with ICU readmission. However, when we included all of these variables in our
multivariate model, only chronic cardiovascular, respiratory and gastrointestinal disease, and
patient origin remained statistically significant.

13

Table 3: Descriptive statistics by ICU Readmission vs. No ICU Readmission
No ICU Readmission
(n=253)
Had ICU Readmission
(n=84)
P value
Pediatric Intensive Care
Unit
151 (60%) 38 (45%)
Cardiothoracic Intensive
Care Unit
102 (40%) 46 (55%)
Age (months) 41.1 (8.2, 129.8) 15.3 (3.5, 133.6) 0.06
Weight (kg) 13.5 (6.8, 29.8) 9.1 (4.7, 26.3) 0.04
Sex (male) 146 (58%) 49 (58%) 0.92
Technology Department
Gastrostomy 39 (15%) 24 (29%) 0.007
Home NGT feeds 2 (1%) 3 (4%) 0.07
Tracheostomy 9 (4%) 4 (5%) 0.62
Complex Chronic Disease
Neuromuscular 50 (20%) 22 (26%) 0.21
Cardiovascular 100 (40%) 51 (61%) 0.001
Respiratory 13 (5%) 16 (19%) <0.001
Renal 6 (2%) 4 (5%) 0.26
Gastrointestinal 12 (5%) 12 (14%) 0.003
Hematologic/Immune 28 (11%) 12 (14%) 0.43
Metabolic 12 (5%) 6 (7%) 0.4
Other congenital/genetic 37 (15%) 20 (24%) 0.05
Malignancy 34 (13%) 8 (10%) 0.35
Patient Origin
ED/Direct admission = 5 32 (13%) 8 (10%) 0.44
OSH=2 58 (23%) 13 (16%) 0.15
Other ICU in same
hospital=4
1 (0.4%) 1 (1%)
0.41
Peds Ward=1 41 (16%) 27 (32%) 0.002
OR=3 121 (48%) 35 (42%) 0.33
Ventilated 1st Admission 120 (47%) 45 (54%) 0.32

After adjusting for these four variables, our multivariate model had a significantly higher
predictive value than our unadjusted PEWS score, with an AUC of 0.754 compared to 0.649.
This suggested that adding these 4 variables to our original PEWS score improves the clinical
performance of the score (Table 4).

14

Figure 2 PEWS Scores and ICU Readmission

NO CTICU or
PICU
Readmission
(n=253)
Had CTICU or PICU
Readmission
(n=84)
p value
PEWS prior to ICU discharge 2 (1, 3) 3 (2, 3) < 0.0001
First PEWS on peds ward 2 (1, 3) 2.5 (2, 3) < 0.0001

Table 4 Association of PEWS with ICU Admission

Adjusted for chronic cardiovascular, respiratory, gastrointestinal disease and origin. Weight, gastrostomy, and ICU admission
category were not associated with ICU readmission after adjustment for other variables.
When evaluating cutoffs values for PEWS at ICU discharge, the best combination of
sensitivity and specificity occurred at a cutoff PEWS score of 2 (sensitivity 77.4%, specificity
39.9%; (Table 5). However, the positive likelihood ratio at a cutoff PEWS of 2 was low (LR+ =
1.6) and negative likelihood ratio was high (LR- = 0.6). Spearman’s correlation of PEWS prior to
ICU discharge with first PEWS on the pediatric ward was 0.7.
TABLE 5. Utility of cutoff PEWS scores (PEWS prior to ICU discharge)
PEWS Cutoff Score Sensitivity Specificity LR+ LR-
≥ 0 100 0 1
≥ 1 100 0.4 1 0
≥ 2 77.4 39.9 1.3 0.6
≥ 3 54.8 72.3 2 0.6
≥ 4 20.2 91.7 2.4 0.9
≥ 5 3.6 98 1.8 1

15

Chapter Four: Discussion

This study demonstrates a positive association between PEWS scores assigned prior to
ICU discharge and first PEWS on the Pediatric Ward with early unplanned ICU readmission.
Both PEWS at ICU discharge and first PEWS on the Pediatric Ward were higher in children
readmitted to the ICU within 48 hours compared to children who were not readmitted. For each
one-point increase in the PEWS score, there was an associated 40% and 48% respective increase
in risk of unplanned ICU readmission. These findings suggest a PEWS scoring system may be
helpful in identifying children at risk of early, unplanned readmission to the PICU and CTICU.
These findings are consistent with a previous study that also demonstrated a positive
association between PEWS assigned prior to the PICU discharge and the first PEWS on the
Pediatric Ward (Mandell, Bynum, Marshall, Bart, Gold, Rubin, 2015). However, this study only
looked at early, unplanned PICU readmissions. When developing a scoring system to assist in
identifying patients at risk for deterioration, a universal tool is preferred because of the potential
versatility and adaptability in different clinical environments. Utilization of a tool that can assist
in providing an objective means of identifying children at risk for decompensation is a common
goal. Being able to identify a PEWS cut off score would assist in this endeavor. In this study, no
threshold score could be identified. This was likely due to the low prevalence of ICU
readmissions in this sample. The design and implementation of our PEWS was an attempt to
create a tool that can be generalized across our patient population to identify those at highest risk
for deterioration. However, in making the PEWS adaptable, certain components of the scoring
system may in fact make it difficult to capture some patients who are known to be at higher risk
for acute deterioration. Alone, a PEWS score used as a predictor of ICU readmission is not very

16

strong (AUC= 0.649 and AUC= 0.670, respectively). Close attention was given to identifying
characteristics that would assist in making the PEWS score a stronger predictor of clinical
deterioration. It has been previously determined that patients with complex chronic conditions
and with technology dependence may also be associated with early ICU admission (Krmpotic,
Lobos, 2013). Therefore, the addition of these chronic disease and technology-dependent
diagnoses to our regression model may improve the discrimination ability of PEWS scores.
According to the results in Table 4, the conclusion that adding these criteria to the current PEWS
score may improve its predictive performance is strengthened by showing that the positive
likelihood ratio of any PEWS score cutoff was low (<10) and the negative likelihood ratio of any
PEWS score cutoff was high (>0.1). The best combination of sensitivity and specificity occurred
at a cutoff PEWS score of 2 (sensitivity 77.4%, specificity 39.9%). The sensitivity and
specificity were never both acceptably high for any PEWS cutoff to be identified. Therefore, no
PEWS score cutoff that predicted ICU readmission could be identified.
Unplanned readmissions to the ICU have been associated with an increase in mortality
and longer length of stays (Rosenberg, Hofer, Hayward, Strachan, Watts, 2001). Children
admitted to the ICU are already at a greater risk of clinical decompensation than children who
were never admitted to the ICU. Thus, identifying patients who are appropriate for stable
transfer out of the ICU is important and a screening tool would be helpful. A study looking at
unplanned readmissions to the pediatric cardiac unit examined if PEWS prior to transfer to the
Pediatric Ward predicted readmission (Kroeger, Morrison, Smith, 2018). This single center
study used a modified Monaghan early warning score system and looked at all readmissions to
the cardiac ICU. They found a 70% increase in odds of unplanned readmission for each one
point increase in pre-transfer cardiac PEWS. However, the PEWS utilized in their study was

17

specifically designed for cardiac patients. Our study is the first study to evaluate the use of
PEWS in the setting of unplanned readmissions to both the PICU and CTICU, demonstrating
versatility in the use of our PEWS system. The utilization of PEWS have assisted in making
clinicians more aware of the potential risk for clinical deterioration and for possible ICU
readmission. However, it is still recommended that PEWS be used in conjunction with the
clinical assessment of patients.
There were several limitations to our study. Since the ICU readmission rate is low
(approximately 4%), there was not a large number of cases to evaluate. This led to a case-control
study design instead of a potentially more informative cohort study design. Expanding the study
to include readmissions of several children's' hospitals in a multicenter study would be ideal.
This may also assist in identifying other variables that can increase the discrimination ability of
PEWS scores for ICU readmission. Our PEWS scoring system was modified from the Brighton
PEWS. PEWS score criteria may vary between hospitals as hospitals may utilize a different
scoring system designed for their specific patient population. Therefore, our results may not be
applicable to all institutions, particularly for hospitals caring for patients with low prevalence of
chronic disease.
Despite our limitations, we were able to demonstrate a positive association between
PEWS assigned at ICU discharge and first PEWS on the pediatric ward with early unplanned
ICU readmission. Future studies should evaluate the modified Brighton PEWS with the addition
of complex, chronic conditions, particularly cardiac, gastrointestinal, and respiratory. These
findings should assist in the design of future studies that evaluate PEWS components and
likelihood of ICU readmission that can potentially be standardized across hospitals.

18

Chapter Five: Conclusion

Development of a system to assist providers in rapidly identifying potential high-risk children is
vital in helping to decrease morbidity and mortality. PEWS prior to ICU discharge and first
PEWS on the Pediatric Ward may assist in alerting providers of children at risk of early
unplanned ICU readmission. Adding chronic disease variables such as cardiac, gastrointestinal
and respiratory conditions to the PEWS criteria, may improve the generalizability and predictive
performance of the PEWS scores in future iterations of the score and should be evaluated in a
multi-center prospective study.

19

References

Akre, M., Finkelstein, M., Erickson, M., Liu, M., Vanderbilt, L., & Billman, G. (2010).
Sensitivity of the pediatric early warning score to identify patient deterioration.
Pediatrics, 125(4), e763-769.
Andréasson, A. C., Herlitz, J., Bång, A., Ekström, L., Lindqvist, J., Lundström, G., & Holmberg,
S. (1998). Characteristics and outcome among patients with a suspected in-hospital
cardiac arrest. Resuscitation, 39(1-2), 23-31.
Brown, S. E., Ratcliffe, S. J., Kahn, J. M., & Halpern, S. D. (2012). The epidemiology of
intensive care unit readmissions in the United States. Am J Respir Crit Care Med, 185(9),
955-964
Buist, M, Moore, G., Bernard, S., Waxman, B., Anderson, J., & Nguyen, T. (2002). Effects of a
medical emergency team on reduction of incidence of and mortality from unexpected
cardiac arrests in hospital: preliminary study. BMJ : British Medical Journal., 324(7334),
387–390.
Campbell, A. J., Cook, J. A., Adey, G., & Cuthbertson, B. H. (2008). Predicting death and
readmission after intensive care discharge. Br J Anaesth, 100(5), 656-662.
Chan PS, Khalid A, Longmore LS, Berg RA, Kosiborod M, Spertus JA.(2008) Hospital-wide
code rates and mortality before and after implementation of a rapid response team.
JAMA. Dec 3;300(21):2506-13

Cohen E, Kuo DZ, Agrawal R, Berry JG, Bhagat SK, Simon TD, Srivastava R. (2011). Children
with medical complexity: an emerging population for clinical and research initiatives.
Pediatrics. Mar;127(3):529-38. Epub 2011 Feb 21. Review.
Duncan, H., Hutchison, J., & Parshuram, C. S. (2006). The Pediatric Early Warning System
score: a severity of illness score to predict urgent medical need in hospitalized children. J
Crit Care, 21(3), 271-278.
Feudtner, C., Christakis, D. A., & Connell, F. A. (2000). Pediatric deaths attributable to complex
chronic conditions: A population-based study of washington state, 1980–1997. Pediatrics,
106(Supplement 1), 205-209.
Goldhill, D. R., White, S. A., & Sumner, A. (1999). Physiological values and procedures in the
24 h before ICU admission from the ward. Anaesthesia, 54(6), 529-534.
Graf JM, Montagnino BA, Hueckel R, McPherson ML.(2008). Pediatric tracheostomies: a
recent experience from one academic center. Pediatr Crit Care Med. Jan;9(1):96-100.

20

Krmpotic K, Lobos AT. (2013). Clinical profile of children requiring early unplanned
admission to the PICU. Hosp Pediatr. 3(3):212-218.

Kroeger, A., Morrison, J., & Smith, A. (2018). Predicting unplanned readmissions to a pediatric
cardiac intensive care unit using predischarge Pediatric Early Warning Scores.
Congenital Heart Disease, 13(1), 98–104.
Linton, S., Grant, C., Pellegrini, J., & Davidson, A. (2009). The development of a clinical
markers score to predict readmission to paediatric intensive care. Intensive Crit Care
Nurs, 25(6), 283-293.
Lone NI, Gillies MA, Haddow C, Dobbie R, Rowan KM, Wild SH, Murray GD, Walsh (2016).
TS Five-Year Mortality and Hospital Costs Associated with Surviving Intensive Care.
Am J Respir Crit Care Med. 2016 Jul 15;194(2):198-208.
Mandell, I., Bynum, F., Marshall, L., Bart, R., Gold, J., & Rubin, S. (2015). Pediatric Early
Warning Score and unplanned readmission to the pediatric intensive care unit. Journal of
Critical Care., 30(5), 1090–1095.
McQuillan, P., Pilkington, S., Allan, A., Taylor, B., Short, A., Morgan, G., . . . Collins, C. H.
(1998). Confidential inquiry into quality of care before admission to intensive care. BMJ,
316(7148), 1853-1858.
Monaghan, A. (2005). Detecting and managing deterioration in children. Paediatr Nurs, 17(1),
32-35.
Odetola, F. O., Clark, S. J., Dechert, R. E., & Shanley, T. P. (2007). Going back for more: an
evaluation of clinical outcomes and characteristics of readmissions to a pediatric
intensive care unit. Pediatr Crit Care Med, 8(4), 343-347.
Parshuram, C. S., Bayliss, A., Reimer, J., Middaugh, K., & Blanchard, N. (2011). Implementing
the Bedside Paediatric Early Warning System in a community hospital: A prospective
observational study. Paediatr Child Health, 16(3), e18-22.
Parshuram, C. S., Duncan, H. P., Joffe, A. R., Farrell, C. A., Lacroix, J. R., Middaugh, K. L., . . .
Parkin, P. C. (2011). Multicentre validation of the bedside paediatric early warning
system score: a severity of illness score to detect evolving critical illness in hospitalised
children. Crit Care, 15(4), R184.
Penk, J. S., Loke, Y. H., Waloff, K. R., Frank, L. H., Stockwell, D. C., Spaeder, M. C., & Berger,
J. T. (2015). Unplanned admissions to a pediatric cardiac critical care unit: a review of 2
years' experience. Pediatr Crit Care Med, 16(2), 155-160.
Reis, A., Nadkarni, V., Perondi, M., Grisi, S., & Berg, R. (2002). A prospective investigation
into the epidemiology of in-hospital pediatric cardiopulmonary resuscitation using the
international Utstein reporting style. Pediatrics, 109(2), 200–209.

21

Roland, D., Oliver, A., Edwards, E. D., Mason, B. W., & Powell, C. V. (2014). Use of paediatric
early warning systems in Great Britain: has there been a change of practice in the last 7
years? Arch Dis Child, 99(1), 26-29. doi:10.1136/archdischild-2012-302783
Sax, F. L., & Charlson, M. E. (1987). Medical patients at high risk for catastrophic deterioration.
Crit Care Med, 15(5), 510-515.
Skaletzky SM, Raszynski A, Totapally BR. (2011) Validation of a modified pediatric early
warning system score: a retrospective case–control study. Clin Pediatr 51(5):431–5.
Smith, A. F., & Wood, J. (1998). Can some in-hospital cardio-respiratory arrests be prevented?
A prospective survey. Resuscitation, 37(3), 133-137.
Smith ME, Chiovaro JC, O'Neil M, Kansagara D, Quiñones AR, Freeman M, Motu'apuaka ML,
Slatore CG. (2014). Early warning system scores for clinical deterioration in hospitalized
patients: a systematic review. Ann Am Thorac Soc. 2014 Nov;11(9):1454-65.
Subbe, C. P., Kruger, M., Rutherford, P., & Gemmel, L. (2001). Validation of a modified Early
Warning Score in medical admissions. QJM, 94(10), 521-526.
Whittington J, White R, Haig KM, Slock M. (2007). Using an automated risk assessment report
to identify patients at risk for clinical deterioration. Jt Comm J Qual Patient Saf
2007;33:569–5.

Abstract (if available)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Validation of a modified pediatric early warning system (PEWS) score

PDF

Improving health equity after discharge from the neonatal intensive care unit: a qualitative study

PDF

The effect of cytomegalovirus on gene expression of pediatric acute lymphoblastic leukemia

PDF

Simulation-based training is associated with lower risk-adjusted mortality in ACS Pediatric TQIP centers

PDF

Validation of the pancreatitis activity scoring system in a large prospective cohort

PDF

Alkaline phosphatase levels in infants with spontaneous intestinal perforation vs. necrotizing enterocolitis with perforation

PDF

The association between inhaled nitric oxide treatment and ICU mortality and 28-day ventilator free days in children with acute respiratory distress syndrome

PDF

Cerebral oxygen saturation and cerebrovascular instability in newborn infants with congenital heart disease compared to healthy controls

PDF

The cost of opioid use in high-risk hospitalized infants

PDF

Assessing the utility of wearable fitness trackers for objective capture and quantification of activity levels among patients undergoing radical cystectomy

PDF

Refined screening guidelines for radiation-related late effects in childhood cancer survivors: a dose-volumetric guided approach

PDF

Alpha globin and vascular reactivity in humans: loss of alpha globin genes in human subjects is associated with improved nitric oxide-mediated vascular perfusion

PDF

Predictors of thrombosis in hospitalized children with central venous catheters: a multi-center predictive study from the CHAT Consortium

PDF

The Filipino family initiative: preliminary effects of an evidence-based parenting intervention offered in churches on parent and child outcomes

PDF

Survival trends and related outcomes of survivors of childhood and young adult cancer

PDF

The association between Echocardiography use in the Intensive Care Unit and 28-day mortality among patients with septic shock: a systematic review

Asset Metadata

Creator Bynum, Francine Denise (author)

Core Title Association of Pediatric Early Warning Score with early intensive care unit readmission

School Keck School of Medicine

Degree Master of Science

Degree Program Clinical, Biomedical and Translational Investigations

Publication Date 04/28/2019

Defense Date 04/28/2019

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag early warning scores,ICU readmission,OAI-PMH Harvest,Pediatrics,pews,PICU,readmission

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Belcher, Britni (committee chair), Miller, Lynn (committee member), Rubin, Sarah (committee member)

Creator Email fdbynum@gmail.com,fdbynummd@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-158230

Unique identifier UC11660647

Identifier etd-BynumFranc-7336.pdf (filename),usctheses-c89-158230 (legacy record id)

Legacy Identifier etd-BynumFranc-7336.pdf

Dmrecord 158230

Document Type Thesis

Format application/pdf (imt)

Rights Bynum, Francine Denise

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA