National Early Warning Score (NEWS)

Clinical background

The National Early Warning Score (NEWS) was developed to standardise the assessment of acute clinical deterioration in hospitalised adults. The score is a track-and-trigger instrument: vital signs that are already measured routinely at the bedside are scored and combined into a single number that triggers a defined level of escalation. The decision the instrument serves is not diagnostic but operational, namely when a patient requires increased monitoring, urgent medical review or immediate critical care input. Without an aggregate score, deterioration tends to be recognised late, because individual vital signs often drift gradually and are interpreted in isolation by changing staff.

NEWS was launched in 2012 by the Royal College of Physicians (RCP) in London as a national standard instrument for the NHS. In 2017 it was revised into NEWS2, chiefly by introducing an alternative oxygen saturation scale for patients with hypercapnic respiratory failure [3]. The calculator here uses the original NEWS structure, with the AVPU scale for level of consciousness and without the dual SpO₂ scale.

Calculating the National Early Warning Score

NEWS is calculated as the sum of the points from seven components:

$$\text{NEWS} = P_{\text{RR}} + P_{\text{SpO}$$2} + P{\text{temp}} + P_{\text{SBP}} + P_{\text{HR}} + P_{\text{O}2} + P{\text{AVPU}}

where each $P$ is the score for the respective vital sign according to fixed ranges:

The total score ranges from 0 to 20. Supplemental oxygen gives 2 points and reduced level of consciousness (VPU) gives 3 points, regardless of the degree of impairment.

The derivation cohort consisted of 35,585 consecutive acute medical admissions at four British hospitals, with a total of 198,755 observation sets [1]. The outcomes modelled were cardiac arrest, unanticipated intensive care admission or death within 24 hours of a given NEWS score. In this cohort NEWS performed better than 33 other early warning systems for the composite outcome, with an AUROC of 0.873 (95% CI 0.866 to 0.879) [1]. For death alone the AUROC was 0.894, for unanticipated intensive care admission 0.857 and for cardiac arrest 0.722.

Interpretation in practice

The calculator defines three risk bands that govern the clinical response:

The single parameter scoring 3, which triggers independently of the total score, is central: a patient with a total score of 2 but a respiratory rate of 28 falls into the medium band and requires increased monitoring, because a single markedly deranged vital sign may precede rapid deterioration even when the other parameters are normal.

Validation and performance

A large external validation was carried out in Singapore in 58,809 patients on a general ward at a tertiary hospital [4]. NEWS2 (in practice the same scoring structure as NEWS for patients without hypercapnic respiratory failure) showed an AUROC of 0.898 (95% CI 0.881 to 0.915) for the composite outcome of death, unanticipated intensive care admission or cardiac arrest within 24 hours. Discrimination was excellent for death (AUROC 0.963) but only moderate for unanticipated intensive care admission (AUROC 0.761) and cardiac arrest (AUROC 0.795). Calibration was excellent for mortality but poor for intensive care admission and cardiac arrest, reflecting that these outcomes are harder to predict from vital signs alone [4].

At a threshold of NEWS ≥5, 82.2% of all deterioration events were captured, but 22.9% of patients triggered an alert. At a threshold of ≥7 the alert burden fell to 9.2% but only 67.1% of events were captured [4]. This illustrates the inherent trade-off between sensitivity and specificity when choosing an alert threshold.

A Danish–Dutch multicentre study in emergency departments showed that NEWS performed worse with increasing age [2]. In a Danish development cohort (14,809 patients) the AUROC for in-hospital mortality was 0.82, and in a Dutch validation cohort (50,448 patients) 0.75. In patients over 80 years of age, NEWS markedly underestimated mortality risk. When age was added to the model the AUROC rose to 0.86 and 0.82 respectively, and the underestimation was eliminated [2]. Age was the single strongest predictor of death, but added no information for predicting intensive care admission.

A French cohort study of 202 hospitalised patients with COVID-19 found an AUROC of 0.68 (95% CI 0.60 to 0.77) for the composite outcome of intensive care admission or death, with an optimal threshold at NEWS ≥6 and an adjusted odds ratio of 3.78 (95% CI 1.94 to 7.09) [5]. The lower discrimination in this cohort probably reflects that patients with COVID-19 may deteriorate rapidly with relatively modest derangement of vital signs initially.

A broad evaluation of NEWS2 against 36 other early warning systems across 123 diagnostic groups found that NEWS2 performed best, or joint best, in 120 of 123 groups [6]. In three groups it was marginally outperformed, and only for the outcome of unanticipated intensive care admission.

Limitations

NEWS is validated for adult inpatients on acute medical and surgical wards. It does not apply to children, pregnant women or patients in intensive care, where other systems are used. Patients with chronic respiratory failure, particularly those with chronic hypercapnia, may have a systematically elevated NEWS because of a low oxygen saturation that represents their stable baseline. This was the reason NEWS2 introduced an alternative SpO₂ scale for this group [3]. A critical analysis of NEWS2 found, however, that in a cohort of 2,361 patients with acute exacerbation of COPD the alternative scale performed no better than the original NEWS, and that 44% (27 of 62) of patients who had a NEWS ≥7 and subsequently died would have fallen into a lower risk band using the dual scale of NEWS2 [3]. This raises questions about the safety of lowering the score in patients with chronic hypoxia.

The age effect is a systematic weakness: NEWS underestimates risk in patients over 80 years of age, who may have serious illness with vital signs that still fall within normal ranges [2]. Nor does the instrument capture subjective signs of deterioration such as the patient's general appearance, distress or confusion, which may precede objective derangement of the parameters.

A practical pitfall is that NEWS is a snapshot. The score must be measured repeatedly and trended; a single normal measurement can give false reassurance if the patient is deteriorating rapidly. The frequency of measurement should be matched to the risk band, and a rising trend within the medium band warrants escalation even if the threshold for high risk has not been reached.

References

1. Smith GB, Prytherch DR, Meredith P, et al. The ability of the National Early Warning Score (NEWS) to discriminate patients at risk of early cardiac arrest, unanticipated intensive care unit admission, and death. Resuscitation 2013. PMID: 23295778
2. Nissen SK, Candel BGJ, Nickel CH, et al. The Impact of Age on Predictive Performance of National Early Warning Score at Arrival to Emergency Departments: Development and External Validation. Ann Emerg Med 2022. PMID: 34742589
3. Hodgson LE, Congleton J, Venn R, et al. NEWS 2: too little evidence to implement? Clin Med (Lond) 2018. PMID: 30287428
4. Chen L, Sun Y, Kuek SY, et al. Predictive performance and temporal dynamics of National Early Warning Score 2 (NEWS2) in detecting clinical deterioration in general ward: A large-scale validation study in Singapore. Resusc Plus 2025. PMID: 41142216
5. Pokeerbux MR, Yelnik CM, Faure E, et al. National early warning score to predict intensive care unit transfer and mortality in COVID-19 in a French cohort. Int J Clin Pract 2021. PMID: 33650136
6. Price C, Prytherch D, Kostakis I, et al. Evaluating the performance of the National Early Warning Score in different diagnostic groups. Resuscitation 2023. PMID: 37931891