Your browser is no longer supported

For the best possible experience using our website we recommend you upgrade to a newer version or another browser.

Your browser appears to have cookies disabled. For the best experience of this website, please enable cookies in your browser

We'll assume we have your consent to use cookies, for example so you won't need to log in each time you visit our site.
Learn more


How helpful are early warning scores?

  • 1 Comment

Despite the use of early warning scoring systems, deteriorating patients are often not identified, suggesting these systems may not be sufficiently robust


This article discusses a literature review examining UK practice and the origins, benefits and limitations of early warning scores. An accompanying article discusses the introduction of clinical-based trigger questions to help ward-based nurses to identify patients whose condition is deteriorating.

Citation: Carberry M et al (2014) Early warning systems 1: how helpful are early warning scores? Nursing Times; 110: 1/3, 12-14.

Authors: Martin Carberry is a nurse consultant for critical care; Pauline Clements is a hospital emergency care team (HECT) senior charge nurse; Elaine Headley is a HECT senior charge nurse; all at NHS Lanarkshire.


In 2004, junior doctors were restricted to working 58 hours a week, with a target of 48 hours a week by 2009 (European Commission, 2000). These hours were previously unrestricted and varied across the UK. Traditionally, junior doctors covered out-of-hours (OOH) medical care, and the reduction in their working hours forced the NHS to address how OOH acute hospital care is delivered.

NHS Lanarkshire responded by developing hospital emergency care teams (HECT), which involved recruiting and training senior critical care nurses as the first responders to the acute deterioration of ward patients. The HECT nurses assume responsibility for all ward-based OOH clinical duties; during the day their main focus is delivering an outreach service to help identify patients at risk of deterioration.

To carry out their role, the HECT nurses visit all the acute hospital wards and ask the nurses to identify patients with high modified early warning score (MEWS) or those who are causing concern.

A review of the team’s OOH activity showed that many patients being referred had some recognisable signs of deterioration or had undergone significant clinical interventions in the previous 24 hours of care. Despite this, they were not highlighted to the HECT nurses unless the patient had a MEWS of four or more. We therefore decided to review our use of early warning scores, so a systematic review of the literature was carried out.

Literature review

One of the Audit Commission’s (1999) highest priority recommendations was to develop an outreach service through which critical care specialists could support ward staff in managing patients at risk of deterioration.

These recommendations were followed by the Comprehensive Critical Care review (Department of Health, 2000), which stated that outreach services should be an integral part of comprehensive critical care. However, Goldhill (2001) reported that critical care facilities in the UK were underprovided, and there was evidence that critical care patients were more unwell in the UK than those in comparable countries.

Goldhill (2001) suggested that critical care bed numbers should increase at least twofold to meet requirements. Comprehensive Critical Care (DH, 2000) had already outlined a concept for delivering services to patients based on need, rather than locality. “Critical care without walls” places patients at the front of service provision. The aim of outreach is to extend critical care services beyond the usual physical limits of the intensive care unit and act as a service and educational partnership between critical care and general wards (Leary and Ridley, 2003).

In-hospital cardiac arrests

There is strong evidence that the majority of patients who experience cardiorespiratory arrests in hospital have abnormal vital signs in the hours before the event (Goldhill and McGinley, 2005). The government initiatives discussed above focused on identifying deteriorating patients early so interventions can be given, potentially improving outcomes.

Some seminal studies from the 1990s are worthy of comment. Schein et al (1990) studied a group of 64 consecutive in-hospital cardiopulmonary arrests and reported that 84% of patients had signs of clinical deterioration or new complaints documented within eight hours before the arrest. Franklin and Mathew (1994) supported these findings, reporting that clinicians noted clinical deterioration in 66% of patients included in their study before arrest. In the UK, McQuillan et al (1998) reviewed the quality of care in two groups of 50 consecutive patients before admission to ICU. They found evidence of suboptimal care, often defined as a failure to recognise clinical signs of deterioration or failure to act on these signs.

Early warning scores

The concept of using early warning scores (EWS) systems was introduced by the DH (2000) as part of the recommendations in the comprehensive critical care report.

The EWS, also known as “track-and-trigger systems”, is the calculation of an aggregate trigger score based on physiological abnormalities of heart rate, blood pressure, respiratory rate, temperature, urine output and level of consciousness. It is intended to support objective decision making to help staff identify deteriorating patients.

However, despite the introduction of EWS, a report investigating patient outcome and death (National Confidential Enquiry into Patient Outcome and Death, 2012; 2005) and the National Patient Safety Agency (2007) found failures to recognise and act on signs of cardiopulmonary arrest (NCEPOD, 2012; NPSA, 2007). These findings raised concerns that patients’ physiological changes were still being overlooked or misinterpreted and questions about whether the clinical application of multiple EWS systems was consistent.

A systematic review found limited evidence of the validity, reliability and utility of track-and-trigger systems and suggested they should only be used in combination with clinical judgement (Gao et al, 2007). This is supported by Jansen and Cuthbertson (2010) who claim there is poor evidence behind the use of EWS systems and further research is needed to validate their use. However, they also suggest EWS can promote good practice by reinforcing the need for regular physiological observations.

Worryingly, failure to rescue deteriorating patients is often linked to staff having difficulty asking for advice and relaying or interpreting information due to occupational, professional and hierarchical boundaries (NCEPOD 2012; NPSA, 2007). The Critical Care Stakeholder Forum (2005) identified three key problems that could cause delays in identification and referral:

  • Low standards of documentation and observations on general wards;
  • Poor knowledge of critical illness and its presentation;
  • Suboptimal treatment of at-risk patients due to inadequate skills and knowledge, and organisational failings.

However, Andrew and Waterman (2001) argued that EWS can be used as a tool to overcome these clinical and communication barriers, as the tool gives nurses the opportunity to present their concerns as evidence to medical staff.

A review by Smith et al (2008) identified 72 EWS or track-and-trigger systems, highlighting a point raised by Subbe (2010) that the many adaptations of EWS questions makes their validity questionable. In an attempt to standardise practice, the Royal College of Physicians (2012) published a report recommending the use of a national early warning score (NEWS) in the UK.

Electronic warning systems

O’Kane et al (2012) identified two issues that need to be taken into account to maximise the benefits of EWS: the need to collect vital signs data frequently enough to detect patient deterioration; and the importance of recording data accurately. Electronic warning systems have been developed to overcome these issues.

Computer programmes can chart variables, calculate scores and immediately alert the appropriate teams to deteriorating patients (Nwulu et al, 2012). However, as with all EWS systems, they remain reliant on timely and complete observations being carried out (Nwulu et al, 2012).

Improving outcomes

The multicentre MERiT study in Australia investigated the effectiveness of medical emergency teams (MET) hospitals against non-MET hospitals. The study noted an increase in calls to MET, but no statistical difference in the number of cardiac arrests, unplanned ICU admissions or unexpected deaths (Hillman et al, 2005). The authors also found that MET calls did not occur quickly when appropriate physiological triggers occurred.

Endacott et al (2009) conducted a meta-analysis to explore the activities and outcomes of UK-based ICU liaison nurses and outreach teams. They found these teams reduced intensive care mortality, hospital mortality, unplanned admissions and readmissions to ICU, delayed discharges and rates of adverse events. However, limitations to the study make it difficult to conclude that intensive care liaison outreach services resulted in improved outcomes. The study also identified non-measured improvements in communication, confidence, knowledge and critical care skills (Endacott et al, 2009).

These finding are supported by McGaughey et al (2009); despite limitations on meta-analysis of their work due to heterogeneous influence, two-cluster randomised control trials showed no evidence of effectiveness of outreach services or reduction in mortality of patients receiving outreach services.

Trigger questions

Evidence from the literature and anecdotal feedback from HECT nurse practitioners suggests that the EWS alone is often not enough to identify and aid the handover of patients at risk of clinical deterioration (Jansen and Cuthbertson, 2010).

The need to improve patient safety by improving the early identification of deteriorating patients led us to develop four clinical trigger questions to be used with the MEWS (Box 1). The questions were selected following a case study review of patients who had not been highlighted to the HECT during routine ward checks but then deteriorated out of hours; and previous studies and reports of why patient deterioration had been missed (NCEPOD, 2012; NPSA, 2007; Goldhill and McGinley, 2005; Schein et al, 1990). The trigger questions were tested through short ward-based pilots and agreed before hospital wide implementation.

To our knowledge, there has been no research into the use of specific clinical trigger questions to support identifying deteriorating ward-based patients. This apparent gap in the literature was the main reason for this service evaluation. The accompanying article (page 15) describes the aims, methods, results and discussion of this evaluation.

Box 1. Clinical trigger questions

  • Do you have any patients with a high or increased frequency of MEWS?
  • Are any patients on high-flow oxygen (>40%)?
  • Are any of your patients on fast fluids, blood products or continuous drug infusions?
  • Have any of your patients required medical review out with normal ward round review? 

Key points

  • Inpatients who experience cardiorespiratory arrests often have abnormal vital signs in the hours before the event
  • Noting and escalating clinical deterioration early can help reduce the likelihood of cardiac arrest
  • A lack of communication between health professionals can prevent vital information being passed on and advice being sought
  • There is a lack of evidence of the efficacy of early warning systems
  • Early warning scoring should only be used alongside clinical judgement on patient care 
  • 1 Comment

Readers' comments (1)

  • Hate to point this out, but EWSs are only as good as the research underpinning them. In spite of this, all of them have some value as mnemonics, serving to remind tired and harried staff about clinically important factors.

    In practice, I have seen many fail utterly in practice because nurses fail to measure physiological data properly. More commonly, HCAs may be delegated such observation tasks, but these staff lack the necessary understanding of underpinning physiology or pathology to make sense of the data.

    As a result, I've seen records where staff added up the factors, but simply left out the data they hadn't collected. The higher score level for alerting colleagues was never reached, so consequently the patients monitored either collapsed or recovered without help!

    I've also seen wards where every patient (irrespective of their age, health status or medical condition) were all consistently recorded as breathing at 18 respirations per minute!

    The best designed and tested EWS is useless in such hands.

    Unsuitable or offensive? Report this comment

Have your say

You must sign in to make a comment

Please remember that the submission of any material is governed by our Terms and Conditions and by submitting material you confirm your agreement to these Terms and Conditions. Links may be included in your comments but HTML is not permitted.