Research in practice

A systematic review of the characteristics of published paediatric early warning systems

Early identification of children at risk of deterioration is a priority for patients, families and healthcare professionals.

Photo: iStock

The paediatric early warning system (PEWS) score has been introduced in many hospitals to assist staff with the recognition of children who may be deteriorating. However, many differing systems exist and there is a lack of agreement on which components and characteristics are most effective.

Design and methods

This study formed part of a wider systematic review of published PEWS scores. The review followed the grading of recommendations assessment, development and evaluation (GRADE) approach.

Paediatric track and trigger systems (PTTS) were categorised as scoring or trigger systems. Within a scoring system, different parameters (for example, heart rate or oxygen therapy) would be assessed against a scoring matrix. Values with increased deviation from normal would be allocated a higher score. The scores for all the different parameters would then be cumulated into a single numeric value, which, depending on the cut off level, determines a patient’s risk for clinical deterioration. In a trigger system the patient is considered at risk if any one of the parameters is positive.

Systems were then classified as being either age-dependent (multiple systems with differing age-related thresholds) or age-independent (a single system applied regardless of age).


Overall, 55 papers describing 33 PEWS scores were identified. A total of 21 systems were classified as scoring systems, and 12 as trigger systems. Fourteen were age-dependent and 19 age-independent.

There was a wide variety in the number and type of parameters. The median number of parameters per system was six, with a range of three to 19. Some broader parameters shared the same name (such as respiratory or cardiovascular), but were constituted from differing component parts or had differing thresholds for scoring/triggering.

All PEWS included one or more vital signs. Method of vital sign utilisation varied, but largely took one of three forms. At the simplest level, a single vital sign was assessed against a scoring matrix with clear thresholds for normal and abnormal values. Other PEWS scores assessed vital signs against subjective criteria where some interpretation or evaluation was required by the clinician. Examples include ‘acute change’ in vital sign or a recording ‘above the baseline’. At its most complex level, the vital sign was one part of a parameter requiring assessment of two or more component parts simultaneously. These complex parameters most commonly include subjective measures of vital signs combined with other related clinical features and interventions, such as oxygen therapy and respiratory effort.

Most PEWS appeared to be developed from expert opinion. Differences in vital sign thresholds were often minor and their source was unreferenced. Age categories varied as did the thresholds for systolic blood pressure, heart and respiratory rate, which resulted in marked differences in scoring for some PEWS. Some PEWS provided different thresholds dependent on whether the child was awake or sleeping, male or female. Although trigger thresholds for oxygen saturation were largely consistent, the method of measurement varied, with some requiring measurement in oxygen, others in air or detecting a decrease despite non-specified first line interventions. Thresholds for capillary refill time and temperature were broadly similar; however, they were less frequently incorporated into the PEWS compared to other vital signs.

Implications for practice

A high number of published PEWS exist, often with minor differences to differentiate between differing systems. However, these minor differences can create difficulties in conducting collaborative research and benchmarking practice.

Reflections and lessons learnt

The GRADE Working Group website and the GRADE Handbook were useful, and the website had a link to a series of videos, The Grade Approach. Watching these gave me a good understanding of the methodology but also made me realise the importance of keeping notes of how and why methodological decision were made. Overall, the process was far more time consuming than I anticipated, but also far more rewarding. 

Susan Chapman is Gulf regional clinical manager, international and private patients division at Great Ormond Street Hospital for Children NHS Foundation Trust, on behalf of  the RCN's Research in Child Health community


Research supervisors: Jo Wray, Kate Oulton, Mark Peters.

This article is for subscribers only