Health professional reported experience and published evidence (Kause et al, 2004) indicate that patients demonstrate signs of physiological deterioration before cardio-respiratory arrest. It is well known that these patients may be identified whether they are adults (Kause et al, 2004) or children (Tume, 2007); and a variety of ‘early warning systems’ (EWS) have been developed to enable health care staff to recognize pre-arrest physiological changes. EWS have typically been developed for use in the hospital setting.
The Appendix demonstrates an example of the components of an EWS which are standard basic observations (temperature, respiratory rate, heart rate, blood pressure etc.), but may also include scores for subjective assessment (such as work of breathing).
| TEWS | 3 | 2 | 1 | 0 | 1 | 2 | 3 |
|---|---|---|---|---|---|---|---|
| Mobility | Walking | With help | Stretcher immobile | ||||
| RR | <9 | 9-14 | 15-20 | 21-29 | > 30 | ||
| HR | ? 40 | 41-50 | 51-100 | 101-110 | 111-129 | > 130 | |
| Syst BP | ? 70 | 71-80 | 81-100 | 101-199 | > 200 | ||
| Temp | <35 | 35-38.4 | >38.5 | ||||
| AVPU | Alert | Voice | Pain | Unresponsive |
Typically, EWS are adjacent or integral to observation charts. Observations outside the normal range require a graded response by nursing or medical teams and sometimes the activation of a specialist outreach team.
Definitions
The term track and trigger systems (TTS) may be more appropriate than EWS. TTS rely on periodic observation of selected basic physiological signs (‘tracking’) with predetermined calling or response criteria (‘trigger’) for requesting the attendance of staff who have specific competencies in the management of acute illness and/or critical care. This definition appears to have initially been described by the Department of Health (DH) via the NHS Modernization Agency (2003).
It is important to recognize that TTS is not triage. Triage, in its truest sense, is an algorithm to decide how long a patient can wait before being treated. It was initially used in military settings to enable to the review of large numbers of casualties simultaneously. This still occurs in major incidents where a variety of validated systems are in use. A similar principle is now employed at point of access to emergency departments (ED) (Subbe, 2006).
The use of triage in prehospital care is well developed in major incidents, but currently used ED triage tools (Manchester Triage group, 1997; Canadian Association of Emergency Physicians; and Australasian College of Emergency Medicine); were not validated for use in the prehospital setting.
The Cape Triage Group in South Africa has developed a system for in-hospital and prehospital staff (Gottschalk et al, 2006). Although validated in South Africa (Bruijns, 2008), it has not be validated in a North European or American setting—which is relevant as the case mix of patients may be substantially different.
TTS is also not an illness severity score (ISS). These are used to predict outcome at the point of admission such as APACHE II (Knaus, 1985), predict organ failure or relate severity of injury/illness to subsequent resource use. They are commonly used in intensive care settings and trauma centres to enable comparison of the performance of units. With regard to trauma, they have been extensively used to determine that variation in survival is dependent on the skills and experience of the trauma team working in a particular unit. However, the criteria and algorithms used in the scoring systems may be quite complex and are not suitable for ‘at scene’ use as they would be far too cumbersome and time consuming. While triage, EWS and ISS are all different, the terms are sometimes used interchangeably, creating mis-understanding as to their true purpose.
Development of early warning systems
In July 2007, the National Institute for Clinical Excellence (NICE) released a short clinical guideline entitled ‘Acutely ill patients in hospital: Recognition of and response to acute illness in adults in hospital’. Key priorities for implementation included:
1.2.1……Physiological track and trigger systems should be used to monitor all adult patients in acute hospital settings…
It was realized that such tools may have a role outside of the ward environment and shortly after the NICE guidance was released, the Royal College of Physicians published a report on acute medical care. They suggested the creation of a NHS early warning score, otherwise known as NEWS. An extract from this report reads:
‘…We recommend that the NEWS score be used at all stages in the acute care pathway, including prehospital assessment, e.g. by the GP, ambulance service or other healthcare professionals seeking advice on acute medical care….’
It can be argued that a national system is counter-intuitive as the demographic base in terms of physiology, age and underlying conditions would be so diverse inter-regionally and between different services as to make one single system impossible to validate (Roland, 2011). However, multiple different systems may cause confusion, especially when transferring patients between different health care providers. Validation of a specific system with discriminators may be a solution to this problem (Earley, 2010). One of the difficulties of having one system across primary and secondary care providers is defining which outcome you are wishing to improve. The actual frequency of cardiorespiratory collapse is low as an inpatient, especially so in a paediatric cohort (20 per 105 000 ward patients over 41 months) (Tibballs et al, 2005).
This may make current scoring systems inapplicable to an ED or prehospital cohort where the incidence of mortality is higher in the acute phase of illness. The importance of ambulance services being able to distribute patients to alternative providers means clear outcomes of EWS systems must be developed and relevant to the purposes of the tool. It may be that the number of EWS measurements do not serve the prehospital environment well. For example, EDs tend to use triage systems to identify at the point of presentation those patients needing urgent care. Research or audit from EDs with regard to ongoing use of a scoring system, in addition to triage, to predict deterioration is limited.
A comparison of the available scoring systems vs the triage score at predicting the need for intensive care for adult patients showed they added little benefit (Subbe at al, 2006). In this study, only the physiological parameters collected by the nursing staff on admission to the ED were used and resulting deterioration was not assessed. Sequential scoring has been used in an ED with the use of a modified patient at risk score (Rees and Mann, 2004) (Table 1). Thirty triage category 1 (immediate) or 2 (within 10 minutes) patients were enrolled and scores were taken at 0 and 15 minutes.
| Score | 3 | 2 | 1 | 0 | 1 | 2 | 3 | Score |
|---|---|---|---|---|---|---|---|---|
| ONS response | Alert | Drowsy | Pain | |||||
| Respiratory/min | <8 | 9–14 | 15–20 | 21–30 | >30 | |||
| Pulse rate | <40 | 40–50 | 51–100 | 100–110 | 111–130 | >130 | ||
| Systolic blood | <70 | 71–80 | 81–100 | 101–199 | >200 |
The study suggested sequential scoring could be used to define those at need of further resuscitation or intensive care support. However, with the limited time paramedic crews have with patients, it may not be possible to obtain multiple measurements or validate a system based on the small number of formal observations taken.
Prehospital systems
So where does this leave prehospital care providers? An argument could be made that in some situations the use of a scoring system is limited as decisions often need to be made immediately and on gut instinct. There may not be time to take a full set of observations at all, or at least until the patient has had a number of interventions performed.
For this reason, you may expect that initial ‘scores’ performed outside of hospital do not completely relate to the true initial picture of the patient and may not correspond well with validating outcomes. Although, in one study, a modified early warning score (MEWS) (which includes blood pressure, pulse rate, respiratory rate, temperature and AVPU) taken on presentation to an ED by the ED nursing staff did correlate with the need for admission, length of stay and mortality (Burch et al, 2008) Therefore, an argument can be made that MEWS scores would also correlate as long as the time observations are initially performed is not significantly before arrival in the ED.
Evidence on scoring systems used prehospital does exist and to demonstrate differences, they have been divided up into separate categories as their uses have been quite specific; mainly, medical, trauma and disease outbreaks.
Medical
Critically ill medical, i.e. non-injured, patients may be currently triaged using a number of techniques—such as, dispatch criteria (Sporer et al, 2008) and initial paramedic assessment and medical command office co-ordination (Brown et al, 2009). Initial paramedic assessment of eventual disposition (admission to hospital or discharge from ED) was shown to be better in medical than trauma patients in a UK study of one ambulance service (Clesham et al, 2008).
Predicting eventual discharge from the ED is only one outcome an ambulance providing service may wish to assess with a scoring system. Others may include length of stay in hospital, appropriateness of transfer, mortality etc.
The selection of outcome is important as the range of acuity prehospital providers deal with means it is difficult to maintain good sensitivity and specificity of the tool across all patients. When patients were retrospectively offered alternatives to transfer to hospital, many said they would have chosen other options (Yarris et al, 2006). This latter group of patients is clearly different in outcome than those who need resuscitation at the roadside and yet a scoring system may need to cope with both.
A number of factors, including non-clinical considerations, play a role in the final decision to transport a patient to hospital—therefore this may not be a useful outcome measure either. Researchers have also looked at disease specific triage systems such as those used in coronary artery disease which include the use of prehospital ECG (Brown et al, 2008) and prehospital triage (Sivagangabalan et al, 2009).
A recent meta-analysis concluded more information was needed on the use of triage based direct transfer for percutaneous coronary intervention as there was currently insufficient evidence (Brooks et al, 2009). Prehospital stroke screens also exist and have been validated at indentifying patients who meet criteria for thrombolysis (Kothari et al, 1999) Table 2). A ‘discriminator list’ therefore seems essential in the development of prehospital EWS systems.
| The CPSS evaluates for facial palsy, arm weakness, and speech abnormalities. Items are scored as either normal or abnormal |
|
Facial droop (the patient shows teeth or smiles) |
|
Arm drift (the patient closes their eyes and extends both arms straight out for 10 seconds) |
|
Speech (the patient repeats ‘the sky is blue in Cincinnati’) |
However, even with a discriminator list, it would not be plausible to have a scoring system prepared for every conceivable medical condition. A recent study has examined the use prehospital predictors of subsequent critical illness and noted:
‘…subjective and disease-specific assessments alone may not be sufficient for triage in general populations at risk of critical illness.’
The system developed by the authors predicted with a reasonable degree of accuracy those patients who eventually died, developed severe sepsis or were ventilated.
Although it is obviously important to determine which patients who appear initially well subsequently develop critical illness requiring intervention, prehospital providers may wish to know whether their patient is critically unwell at the moment of their arrival. This is an important point as one retrospective analysis demonstrated that only half of all critically ill patients arriving to the ED had been flagged to the ED as an alert prior to their arrival. The authors suggest that modified EWS scoring may be an aid to improving this (Brown and Bleetman, 2006). Conversely, it is also of crucial importance to recognize which patients do not need urgent hospital transfer and can be referred elsewhere.
Pandemics
In times of serious disease outbreaks, prehospital teams must deal with large numbers of potentially unwell patients rapidly and effectively. A pandemic medical early warning score has been developed for those with a respiratory complaint and correlated will with subsequent disposition from hospital (Gray et al, 2010). However, review of this score compared with other triage techniques did raise concerns about the potential to miss some cases (Goodacre et al, 2010). This again highlights the difficulties of developing one score to cover a variety of eventualities.
Trauma
In the prehospital setting, triage for trauma has long been established. Using a number of criteria, including mode of injury and physiological scoring, decisions are made if the casualty needs transport to a specialist trauma centre, normal ED or needs no medical intervention. The boundaries in these cases between triage, identification of time to be seen, TTS and identification of current or developing critical illness becomes increasingly blurred. This is reflected in the case of military triage where very brief assessments are needed (Eastridge et al, 2010).
‘Decisions made outside of hospital must be accurate as long transfers to reach definitive care may prove to be unnnecessary’
The increasing development of regional trauma centres means decisions made outside of the hospital must be accurate as long transfers to reach definitive care may prove to be unnecessary. A thorough examination of the current system in Australia demonstrated potential deficiencies in the system and the need for review (Cox et al, 2010).
Conclusion
A number of scoring systems exist in prehospital care. They may prove very helpful in guiding the need for transport to secondary care or in identifying the most critically ill patients. EWS may provide prehospital care providers with more objective data than just their subjective assessment. Therefore, EWS may empower and aid decision-making by prehospital care providers; however, all paramedic teams already have the autonomy to alert, differentiate, transport and refer to the most appropriate receiving facility for their patient.
The number of tools available is large and the scope for using one tool in all situations is unlikely. Further work is needed in producing and defining the exact aims of these tools by prehospital care providers.