Use of the National Early Warning Score in community nursing: a scoping review

In the past decade, national health and social care policy drivers have redirected the onus of care away from acute settings and firmly back into the community, with the aim of caring for patients in or near to their own homes while maintaining safety and, where possible, avoiding hospital admission (Department of Health and Social Care (DHSC), 2012; DHSC, 2013a; NHS, 2014; 2019). In parallel with the population growth and the fact that people are living longer with multi-morbidities (Divo et al, 2014), the size, acuity and complexity of community nursing caseloads have risen drastically.

The very nature of community nursing means that the infrequency, inconsistency, task-focus and isolation of visiting individuals makes objective judgement and decision-making difficult. This is of particular concern since sepsis alone is thought to account for up to 52 000 deaths annually in the UK (Sepsis Research, 2018), and community patients are arguably most vulnerable to sepsis by virtue of their comorbidities, chronic wounds or recent surgery (Jones, 2017). This further highlights the need for sensitive, objective and consistent assessment and monitoring tools which have become crucial in the early recognition and escalation of physical deterioration (Health Service Executive (HSE), 2017).

Early warning scores (EWS) or track-and-trigger systems, are aggregate scores based on physiological measurements of vital signs that allow for an objective indication of actual or potential deteriorating physical condition, enabling escalation on the basis of concrete evidence of early concerns instead of reliance on nursing intuition (Jarvis et al, 2015). These scores, in numerous and modified forms, have long been used in acute care settings as an adjunct to clinical judgement in aiding the recognition and monitoring of the deteriorating patient.

The National Early Warning Score (NEWS) was developed to address concerns around the lack of standardisation of these tools in acute settings (Royal College of Physicians (RCP), 2012). This assessment tool is simple and quick to complete in that it applies parameters and risk scores to the observations of six vital signs (respiration rate, oxygen saturation, blood pressure, heart rate, temperature and alertness), giving an aggregate score which prompts further action or escalation. NEWS was later improved and updated, becoming NEWS2 (RCP, 2017) in order to better define concerning observation parameters and allow for the ‘normalisation’ of scoring outside of some parameters where underlying morbidity would expectedly lead to that observation.

As a tool, NEWS2 has been well validated in acute inpatient settings and, to some extent, in residential care (Alrawi et al, 2013; Smith et al, 2013) where identifiably ill or frail patients on a treatment pathway can be monitored regularly; in fact, its use is now not only recommended but mandated within acute and ambulance care (NHS England and NHS Improvement, 2018). However, to date, the use of NEWS2 for routine monitoring prior to treatment commencement has not yet been established or evaluated in clinical evidence, neither has it been validated in primary or community care, where the risk of undetected deterioration is arguably greater, due to the infrequency of individual patient contact and the lack of continuity. Of further consideration is the isolation of the community nursing workforce, which is required to make evidence-based clinical decisions rapidly, confidently and autonomously (Nursing and Midwifery Council, 2015; DHSC, 2013b). On investigation, the reasons for this gap are unclear-anecdotally, there appears to be a perceived ‘lack of fit’ or inability to incorporate the tool into community practice in the recommended fashion or intended frequency. This is perhaps because of the task-based nature or intensity of the workload (Tucker and Lusher, 2018); yet, the risk of unrecognised deterioration remains.

This scoping review was undertaken to explore the possible benefits and limitations of using early warning scores for identification of physical deterioration in community nursing caseloads.

Aim and design

Based on the framework described by Arksey and O'Malley (2005) and including later enhancements from others (Levac et al, 2010; Colquhoun et al, 2014), a scoping review of the literature was undertaken to explore the use of EWS or NEWS in community nursing in the UK and its possible benefits and outcomes for patients and community nursing staff in recognition of physical deterioration. It was hoped that this would enable recommendations for quality improvements in community nursing practice that are up to date and evidence based.

Search strategy

An initial search of the medical and nursing databases was undertaken, including CINAHL, BNI, Medline, EMCARE and EMBASE, using terms in isolation and then combination, applying Boolean operators to expand and refine the results as necessary. The terms used included ‘deteriorating patient’, deteriorat*, ‘physical deteriorat*’ combined with community, primary or district and nurs* or care. The intervention was searched as ‘early warning s*’, EWS, NEWS, ‘track and trigger’ or ‘recognise and rescue’. Outcomes were not defined in the search in order to maintain breadth, and the database search was later compared to a brief online search (Google Scholar) with a scan of article references to ensure inclusivity.

Inclusion and exclusion criteria

Searches very quickly highlighted the dearth of evidence pertaining to ‘true’ community nursing, that is, domiciliary nursing caseloads, and so no limitations were applied to the search process. In the final selection of papers for review, some were included for their apparent proximity or comparability to community nursing, but inclusions were limited more by lack of evidence than by single researcher or time constraints.

Papers were included where they represented original UK-based work and addressed any staff or patient aspect involving the use of EWS/NEWS in community or domiciliary settings where patient contact is infrequent. NEWS, in particular, is already well validated in hospital and residential settings, so studies based in hospital, community inpatient units or residential care were excluded in favour of those perceived to include nursing assessment in or near to the patient's own home. Older studies involving track-and-trigger or recognise-and-rescue systems were also excluded in favour of the need for a contemporary, objective, vital-signs based system. Other exclusions were narrative commentaries, opinion pieces or other reviews, which were used instead to inform the discussion. The flow diagram in Figure 1 provides an overview of the search and selection process.

Figure 1. Flow diagram to illustrate literature search and selection

Data extraction and quality appraisal

Descriptors of scoping reviews usually provide no expectations for the quality assessment of evidence (Arksey and O'Malley, 2005; Aveyard et al, 2016), but more recent studies have argued that, without some weighting of evidence, no recommendation can be made regarding its generalisability or direction for future practice, policy and research (Daudt et al, 2013; Colquhoun et al, 2014).

In this case, the data and value of the included studies were extracted, studied and compared for discussion on the basis of validated and published frameworks (Benton and Cormack, 2000; Caldwell et al, 2005; Holland and Rees, 2010; Moule and Hek, 2011; Critical Appraisal Skills Programme (CASP), 2014).

Findings

The findings of this review, where succinctly comparable, are presented in Table 1, and then discussed more fully in narrative organised around the themes that emerged from using the constant comparative method for mixed-method studies in the data extraction.

Two of the included studies were qualitative in nature, exploring staff and service experience of using NEWS in community settings. Of the remaining quantitative studies, three examined patient treatment and outcomes in the light of a NEWS score calculated in differing community or domiciliary settings prior to hospital admission, while the last explored the distribution of NEWS use across differing community services following its introduction regionally.

NEWS in communicating acuity

The majority of studies included recognised the value of NEWS as a communication aid. This may be either as a consistent ‘universal language’ (Pope, 2020), by which to convey severity of concern concisely and with clarity across different services, disciplines and hierarchies, or as a referral tool to add leverage for hospital admission (Brangan et al, 2018), pre-alerting secondary care to the acuity of the patient (Silcock et al, 2015; Inada-Kim et al, 2020). However, Brangan et al (2018) also found that inconsistent use across services and areas could lead to the score being discounted as evidence of illness, due to scepticism or lack of understanding.

NEWS in prioritisation of care

Where used effectively in practice, NEWS was found to be valuable in the prioritisation of care, stratification of risk and mitigation of potential for missed recognition (Pope, 2020). This was particularly true in referral to and arrival at secondary care centres, where Scott et al (2020) found a clear correlation between higher referral scores and reduced times for transfer and clinician review on arrival in the secondary care setting. This supports the suggestion that scores should be highlighted in the handover from community assessment (Silcock et al, 2015; Brangan et al, 2018; Inada-Kim et al, 2020; Scott et al, 2020).

Other qualitative findings of value in terms of prioritisation included the use of the tool to support or refute confusing clinical appearances, giving objective justification and confidence in decisions that may otherwise be challenged (Brangan et al, 2018).

Inada-Kim et al (2020) found the aggregate score to be advantageous in lending weight to minor derangements that may otherwise be overlooked across multiple physical parameters, although Scott et al (2019) and Silcock et al (2015) suggested that, in a community population, chronic long-term conditions may pre-empt a high baseline score, making it more pertinent to monitor changes over time than relying on an absolute score in isolation. With relative strength of sample size in their own study, Scott et al (2019) went on to disprove this notion, finding that only 11% of patients with such conditions had NEWS>5. Both studies used the first version of NEWS (RCP, 2012) during their period of data collection, and Scott et al (2019) further acknowledged that NEWS2 (RCP, 2017) goes some way to mitigating this risk.

One study outlined, by experience, the possible use of NEWS in deprioritisation, in that the referral to secondary care, using a low score alone, could be blocked, even where clinical concern is higher (Brangan et al, 2018).

NEWS vs clinical judgement

All of the studies included in the review touched on the issue of objective tool versus clinical judgement and the pitfalls of using one or the other in isolation.

Many experienced practitioners reportedly resisted the use of a tool, preferring to rely on their own clinical assessment (Pope, 2020), or experienced discomfort when the score disputed their judgement. Brangan et al (2018) described a further ‘tension’ between the two, with the greatest concern being that the sole use of a score increases the risk of missing ‘veiled symptoms’ or comorbidities, with Silcock et al (2015) observing that serious pathology cannot be excluded despite a normal score. In a strong, sizable study, Inada-Kim et al (2020) asserted that the issue of whether NEWS adds any prognostic value to clinical assessment is a ‘key unanswered question’. However, the authors were unanimous in their belief that NEWS is a useful tool for the support of triage and decision making, to be used as an adjunct to clinical judgement and alongside other sources of information, and they were mindful of the RCP's guidance that concern should always override score (RCP, 2012).

Integration of NEWS into clinical practice

From the point of view of addressing the need for use in community nursing, the integration of NEWS into clinical practice showed one of the largest gaps in the evidence. Four of the studies included suggested a need for adaptation or modification of the score to suit a community caseload, and, while two of these studies recommended changes in parameters for those with chronically altered physiology or specific illness (Silcock et al, 2015; Brangan et al, 2018), another lacked clarity in the stated adaptation, with the only identifiable alteration being the addition of a category for new confusion (Pope, 2020), a change that has already been made in NEWS2 (RCP, 2017). Scott et al (2019) acknowledged a difference in physiologies among chronic comorbidities, but suggested that NEWS2 adaptations are not only enough to mitigate this but sensitive enough to identify and contextualise acute illness or exacerbation from normally higher baselines.

Alternatively, they suggested that the trigger thresholds, used to indicate the degree of action or escalation required, may not be consistent both in and out of hospital, and acknowledged that inference either way was beyond the scope of their study (Scott et al, 2019).

Brangan et al (2018) found that community and mental health services within their study were using NEWS in an adaptive way and even ‘beyond its design’ to make it relevant to the caseload or practice schedule, where patients were not acutely ill. Indeed, the frequency of scoring is a significant variable throughout these studies, where some have used a one-off score or episode in isolation to examine effectiveness (Silcock et al, 2015; Inada-Kim et al, 2020; Scott et al, 2020), and community use in practice was found to vary radically in frequency from daily (on inpatient or residential units) to every 3 months (Brangan et al, 2018). In recognition of barriers to use, Silcock et al (2015) highlighted that effectiveness of NEWS as an indicator of deterioration depends on scoring at an appropriate frequency, and, in community settings, this may be limited by workload. This, in turn, may go some way towards explaining why community uptake of NEWS has been slow. Two of the studies observed a degree of GP resistance to use, for reasons of over-medicalisation and need to be selective about appropriate use (Brangan et al, 2018; Scott et al, 2020), and another found that primary and community services had been much slower to adopt the practice than ambulance and emergency departments (Scott et al, 2019).

Sensitivity of community NEWS in terms of outcome

Despite concerns that NEWS has not effectively been validated for use outside of the hospital environment (Brangan et al, 2018; Scott et al, 2019), the three studies that examined clinical outcome in the context of NEWS prior to admission show a clear correlation between high scores/risk strata and poor outcomes in terms of ICU transfer and mortality (Silcock et al, 2015; Inada-Kim et al, 2020; Scott et al, 2020) and, conversely, lower risk of adverse outcome in the lower-risk groups (Inada-Kim et al, 2020). Two of the studies also conferred, to a lesser extent, additional correlation with length of hospital stay (Scott et al, 2019; 2020). The strength of these quantitative studies lies in their robust methodologies and comparative findings, which should indicate more generalisability than sole findings from single trusts or organisations.

The study by Inada-Kim et al (2020) was less conclusive than others about the prognostic value of NEWS, with both Silcock et al (2015) and Scott et al (2020) finding it to be a good predictor of poor outcomes. A strength of the findings is that all three of these studies found comparatively low incidence of high-risk scores in a chronically unwell community population, contrary to the original concern of Scott et al (2019) that this group might show increased secondary care referrals with inflated baseline scores. This, in turn, would strengthen the suggestion, in contrast, that NEWS is sensitive enough to contextualise acute exacerbations in chronic or long-term illness (Scott et al, 2019).

Silcock et al (2015) also made an incidental observation that monitoring of respiratory rate alone is an effective indicator of deterioration, and they found that monitoring of this vital sign was increased by the introduction of NEWS.

NEWS in different patient populations

In the search for evidence in settings outside of hospital or residential care, studies have included primary, community and pre-hospital care, with ‘pre-hospital’ limited to ambulance, GP or advanced clinical practitioner (ACP) referrer. By their very nature, patients presenting to these services would have already self-identified a health issue needing treatment, and NEWS is used as an indicator of the severity of that illness, thereby differing from community nursing caseloads and the need to monitor long-term or chronic illness, identifying acute exacerbations from that background.

In the studies that examined outcome quantitively (Silcock et al, 2015; Inada-Kim et al, 2020; Scott et al, 2020), the decision to admit for secondary care assessment had already been made, so the influence of the score over that decision making is unclear. However, as noted in the latter study, the score was being used previously, so it may have influenced decision making and outcome earlier in the pathway (Scott et al, 2020).

The aforementioned studies were limited to single trusts and one or two hospitals or services, so the findings can only be transferable to similar populations; however, in combination, they have far stronger generalisability given their large samples and geographical diversity.

Discussion

To the author's knowledge, this is the first review of the potential benefit of the use of the NEWS tool within community nursing practice using recognised scoping review methods.

The review itself was limited by the paucity of evidence and lack of validation for community use of NEWS, particularly where any evidence from the only study based entirely in a true community nursing setting (Pope, 2020) is limited by its small size and lack of robust detail. Despite this, the existing literature does give some reassurance that use of NEWS outside of acute settings is at least safe, with some proven benefits in prioritisation of care, effective communication of acuity, speed of transfer and treatment and, particularly, as a predictor of poor outcome.

In fact, the communication benefits of a universal score seem to be the one point on which most authors agree, as long as it is used consistently. However, one of the barriers to consistent community use of NEWS would appear be reticence in GP adoption, perhaps because of a perceived need to be selective about application (Burns, 2018). Another explanation may be the tension experienced in reconciling the score with clinical judgement (Brangan et al, 2018), with some persistently arguing that a score would not augment their decision making but rather interfere with it in creating a ‘cognitive shortcut’ (Finnikin and Wilke, 2020). In general, the authors of studies included in this review agree on the value of NEWS as part of an assessment and in conjunction with clinical judgement—raising awareness of health issues and lending confidence to decision making in the reduction of clinical uncertainty (Fullerton et al, 2012; Patel et al, 2018).

Of the literature searched for this review, very little pertained to actual community nursing or domiciliary care settings, with ‘pre-hospital’ being a common term to describe a singular patient contact in a community setting prior to hospital transfer. This identifies a clear gap in the evidence and raises a question of uncertain difference in use of a score to rate severity in those presenting as already unwell versus routine monitoring of the chronically unwell. Some authors have asserted a need to adapt the parameters or trigger thresholds in NEWS to better suit community services (Pope, 2020; Brangan et al, 2018; Scott et al, 2019), but such adaptations would be at the expense of the much valued universal language and are precluded by the creators of the tool (RCP, 2012; 2017).

It would seem that these studies in combination provide further reassurance on the sensitivity of the NEWS2 score to differentiate acute exacerbations of illness from a chronic background. The relatively low incidence of high scores in community settings may limit its use as a pre-hospital triage tool (Brangan et al, 2018) but can also be interpreted as reassuring in refuting the suspicion that community use of NEWS would increase hospital admission contrary to the aim of admission avoidance.

The challenges for community and domiciliary settings are clearly identified in the literature, not least of which is the lack of validation and clear guidance for integration into practice, where patient contact is infrequent and often task based, in addition to a lack of funding for the resources, equipment and training needed for quality improvement (Tucker and Lusher, 2018).

Perhaps, it is logistically and ethically too difficult to prove improved patient outcomes or reduced morbidity/mortality solely on the strength of one intervention. More importantly, as the weight of evidence shifts, some authors go as far as to assert that there are risks in not using NEWS in terms of treatment delay, mortality/morbidity and litigation (Tucker and Lusher, 2018). This further supports the suggestion that, given the simplicity of the NEWS tool and its successful adoption into so many other pathways (Inada-Kim et al, 2020), it might be irresponsible not to use such an invaluable tool in the identification of preventable illness (Jones, 2017).

Conclusion and recommendations

With increasing acuity of patients in community settings, it would seem crucial, now more than ever, to use an objective assessment tool to identify deterioration and aid the decision making of isolated healthcare workers with infrequent or inconsistent patient contact.

As such, a tool NEWS has been well evidenced, validated and mandated for acute and ambulance use in the monitoring of patients who have already presented with a health issue. However, although recommended, there remains very little evidence for use outside of those settings. This review has elicited some reassurance around the sensitivity, effectiveness and application of NEWS as an adjunct to clinical judgement and communication aid in community nursing, but much more work needs to be done for its successful and consistent implementation.

The implications for practice are clear, in that, the dangers of not using NEWS outweigh the doubts. However, NEWS itself can only be effective in reducing morbidity and mortality if the response to it is appropriate. Therefore, the development of local policy and process must include some clarity around frequency of monitoring and the response or escalation of the score.

Further research needs to establish effective use in routine monitoring of a community patient cohort, adverse or other outcome at earlier stages or lower risk scores, and the trigger thresholds or limits for treatment and maintenance in or near the patient's home in the avoidance of acute admission.

KEY POINTS

• The size and acuity of community nursing caseloads is increasing as people live longer with multi-morbidities
• Nurses visiting infrequently and inconsistently, often on a task-focused basis, need an objective assessment tool by which to identify and quantify physical deterioration
• The National Early Warning Score (NEWS) is validated and mandated for acute and ambulance use, but is not well evidenced in the nursing literature for use in primary care and community caseloads
• NEWS is sensitive enough to identify exacerbations or acute episodes from a background of chronic illness and is a particularly accurate indicator of poor outcome
• Use of NEWS in community nursing might mitigate the risk of unidentified deterioration

CPD REFLECTIVE QUESTIONS

• Why do nurses need an objective assessment tool in addition to their own clinical judgement?
• What are the advantages and limitations of using the same assessment tools in acute and community settings?
• What is a realistic/relevant frequency for completing observations of vital signs and a NEWS score in a community caseload?