Skip to Main Content
Skip Navigation Links
Magnet Logo

Nursing Excellence

The Online Newsletter for Children's Nurses
e-Edition, Issue 6

Pediatric Early Warning Tools

Mary-Ann Robson    Carole Cooper    Lori Medicus    Mary Jo Quintero

Written By: (from left to right)
Mary-Ann Robson, BSN, RN, CCRN
Carole Cooper, MHA, BSN, RN, CPN, NE-BC
Lori Medicus, MN, RN, CPNP/CNS
Mary Jo Quintero, AS, AA, RN, PLN, CCRN

Only 10% of pediatric patients who suffer a cardio-pulmonary arrest survive intact one year post event.1,2 With this poor survival rate it is imperative that we do all we can to prevent cardio-pulmonary arrests in our patients. Research has shown that children who suffer a cardio-pulmonary arrest or unplanned Pediatric Intensive Care Unit (PICU) admission have physiological abnormalities 1-24 hours before the event.3,4 Based on these physiological changes, Pediatric Early Warning (PEW) tools were developed during 2005-2009 by international healthcare providers to help identify acute care patients at risk for deterioration.

There are a variety of tools currently identified in the literature and in use at other children’s hospitals. Three have been validated through clinical studies at major institutions. A team of Children’s Hospital Central California clinical nurse researchers conducted a retrospective case-control study of these tools to help identify the most effective tool for Children’s Hospital. The team evaluated three PEW tools: the Toronto Children’s Hospital PEW System Score; the Toronto Children’s Hospital Bedside PEW System Score; and, the Bristol Children’s PEW Tool. These tools were used to score 96 patients (cases) who triggered a Code Blue call in the acute care areas and 96 randomly selected patients (controls) of similar age, gender, diagnosis, residing patient unit and time frame.

The study identified that children who suffered a cardiopulmonary arrest at Children’s Hospital were primarily comprised of males (59%) and that 76% of the children were less than four years of age, with 57 % less than one year of age (See Table 1). The primary diagnosis of children who suffered a cardiopulmonary arrest was respiratory (42.7%). (See Table 2)

Table 1

Table 2

An aggregate score of five (sensitivity 86.6%, specificity 72.2%) was identified as the trigger score to enact a rapid response team call (See Table 3). This would allow for early recognition and intervention in 86.6% of children developing critical illness in acute care.

Based on the results of the Receiver Operating Characteristic Curve (See Figure 1) and logistical regression, the Toronto PEW System Score was the strongest predictor of cardio-pulmonary arrest in this patient population. The comparison of the three ROC curves showed the Toronto PEW System Score demonstrated a significantly greater amount of accuracy (p<.05). The study identified that in the sample population the Toronto tool demonstrated the greatest sensitivity and specificity in identifying acute care children at risk for cardiopulmonary arrest.

Figure 1


In a Receiver Operating Characteristic (ROC) curve the true positive rate (Sensitivity) is plotted in function of the false positive rate (1 -Specificity) for different cut-off points. A test with perfect discrimination has a ROC plot that passes through the upper left corner (100% sensitivity, 100% specificity). The nearer the ROC plot is to the upper left corner, the higher the overall accuracy of the test.5  The larger the area under the ROC the better the diagnostic test. If the area is 1.0, this is a perfect test, because it achieves both 100% sensitivity and 100% specificity. If the area is 0.5, then this test is no better than flipping a coin as it has only 50% sensitivity and 50% specificity.


 An aggregate score of five (sensitivity 86.6%, specificity 72.2%) was identified as the trigger score to enact a rapid response team call (See Table 3). This would allow for early recognition and intervention in 86.6% of children developing critical illness in acute care.


Figure 3


The Toronto PEW System Score is a tool that is comprised of two components, static and dynamic. When the tool is used consistently as part of a routine nursing assessment, it allows the bedside nurse to quantify a score. As the patient deteriorates, the score rises, indicating an increased risk for cardio-pulmonary arrest. Subtle changes in the child can be trended over time providing an early warning for the healthcare team.

The static or more concrete variables do not change with every assessment. These ask whether the child has: an abnormal airway; been on home oxygen; a Central Venous line; had a transplant; Cerebral Palsy; a Gastrostomy tube; been in ICU this admission or greater than three medical specialties involved in their care.3  Each of the variables requires a Yes/No answer. One point is scored for every “Yes” response. The maximum score a child can have from the static component is eight points.

The second component utilizes age appropriate dynamic or physiological variables. The variables that make up the dynamic component of the tool are:  vitals signs, oxygen saturations, oxygen therapy, pulses, capillary refill, level of consciousness, whether a fluid bolus has been given and amount of medications the child is on.3 As children have different normal parameters for vital signs based on age, there are age specific criteria embedded within the PEW System Score. The maximum score the child can have from the dynamic variables is 24 points.

These two components when scored together assess indicators that the child is at risk for deterioration. These indicators include co-morbidities and physiological responses to illness. 

The Toronto PEW System Score will be implemented this fall in the acute care areas. The aim is to eradicate acute care codes entirely, with a short term goal of a 50% reduction in the first year. A prospective study will be undertaken to evaluate clinical significance and if modifications may improve this tool’s efficacy in identifying deterioration in our patients.



1.  Nadkarni, V., Larkin, G., Peberdy, M., Carey, S., Kaye, et al. (2006). First documented rhythm and clinical outcome from in-hospital cardiac arrest among children and adults. Journal of the American Medical Association, 295 (1), 50-55.

2.  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, 200-209.

3.  Duncan, H., Hutchison, J., & Parshuram, C. (2006). The pediatric early warning system score: a severity of illness score to predict urgent medical need in hospitalized children. Journal of Critical Care, 21, 271-279.

4. Tume, L. (2007). The deterioration of children in ward areas in a specialist children’s hospital. Nursing in Critical Care, 12 (1), 12-19.

5.  Zweig, M. and Campbell, G. (1993) Receiver-operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine. Clinical Chemistry, 39, pp. 561-577.



A very special note of appreciation is extended to Beverly Hayden-Pugh, MOB, BSN, RN, NE-BC, Vice President and Chief Nursing Officer of Children’s Hospital Central California and Dr. Samuel M. Lehman for their ongoing support of this study.

Statistical support provided by Stephen Zuniga, PhD and Rhonda Pisk, MS.

In This Issue

Great Moments

Shining Stars

Pediatric Early Warning Tools

Patient Safety Survey

Informatics: The Language of Nursing

Nursing Peer Review

NICU Outreach Education

What is “Just Culture”?

Patient Satisfaction Comments