NEWS2: The Evidence Behind the Early Warning Score

It is two in the morning on a medical ward. A 72-year-old woman admitted that afternoon with a urinary infection is “not herself”, the staff nurse says. She is picking at the sheet and does not know where she is. Her temperature is 101.3 F. Her breathing, pulse, blood pressure and saturation on air are all within the chart’s normal band. The nurse adds it up and gets a NEWS2 of 4, which is low. “Shall I just recheck in the morning?” This is an illustrative case, not a real patient.

By the end of this article you will know why the answer is no, and what the score can and cannot tell you on an Indian ward.

1. What NEWS2 Measures

Seven bedside observations, each scored 0 to 3 by how far it sits from normal, are summed to a total of 0 to 20. The Royal College of Physicians (RCP) published it in December 2017 for the ward, the emergency department and the ambulance. It grades how unwell a patient is now. Its value grows with repetition: a series of scores shows the direction of travel.

An aggregate of 0 to 4 is low risk, 5 to 6 medium and 7 or more high. A single parameter scoring 3 makes a low total low-medium risk.

2. The Red Score Beats the Total

A score of 3 in any one parameter calls for urgent review by the medical team, even when the total is 3. New confusion, including delirium and disorientation, scores 3 on its own.

Our patient’s temperature of 101.3 F is 38.5 C, which scores 1. Her new confusion scores 3. So her total of 4 is low, but she has a red score, and that makes her low-medium risk and calls for urgent review tonight. This is the common mistake: reading only the total. It is easy to see why, because the total is the number the chart asks you to write down.

“What if she is always like this?” Confusion that is the patient’s usual state scores nothing. When nobody can say, the RCP says to treat it as new. A patient responding only to voice is already a red score; the Glasgow Coma Scale grades it further.

3. Fahrenheit and the Chart’s Precision

Our wards still read temperatures in Fahrenheit, and the chart is printed in Celsius to one decimal. The tool rounds each entry to the chart’s precision, half up, before scoring, so every value falls in exactly one band:

4. The Two Oxygen Saturation Scales

Scale 1 penalises any saturation of 95 per cent or less. For a patient with hypercapnic respiratory failure, that pushes staff to give oxygen past a safe target. So Scale 2 scores low saturations from 87 per cent down and, on oxygen only, scores saturations of 93 per cent and above.

“Then why not use Scale 2 for every patient with COPD?” Because the evidence does not support it. Here are the two studies the tool used to carry:

In those only at risk of type 2 failure, Scale 2 lowered discrimination. That is why the RCP limits Scale 2 to failure confirmed on a blood gas, with the 88 to 92 per cent target prescribed and recorded by a competent decision maker, and the unused scale crossed out on the chart.

5. Sepsis at a Score of 5

The RCP asks that anyone with a known or suspected infection and a NEWS2 of 5 or more be assessed urgently by a team competent in sepsis. In the two studies below, NEWS and NEWS2 predicted death better than qSOFA.

• Among 30,677 patients with suspected infection outside the ICU in a US cohort, NEWS discriminated in-hospital death better than qSOFA (area under the curve 0.77 against 0.69).
• In 373 adults with sepsis in a Delhi emergency department, NEWS2 did the same (0.781 against 0.729), with a sensitivity for death of 83 per cent and a specificity of 57 per cent.

“An area under the curve of 0.77: is that good?” It is the chance that, of one patient who died and one who survived, the score ranks the one who died higher. A coin toss gives 0.5 and a perfect test 1.0. Neither score is perfect, and NEWS2 does a little better. Our patient’s 4 does not reach the sepsis threshold, but her red score has already brought the doctor to the bed.

6. Where NEWS2 Misleads

• Pregnancy and under 16: excluded by the RCP, because the physiological response differs.
• Spinal cord injury: autonomic disturbance can push the score either way.
• Beta blockers and rate-limiting drugs: the pulse may not rise with deterioration.
• Chronic hypoxaemia on Scale 1: the usual saturation scores every time.
• Oxygen: 2 points at any flow, so a rising requirement is invisible in the total.

7. Observation Frequency on Indian Wards

Hourly observations cost nursing time, and the RCP frequencies assume that time exists. The Indian Nursing Council norm is one staff nurse to three beds in a teaching hospital and one to five elsewhere, counted across the day with a 30 per cent leave reserve. A review of the national benchmarks found them outdated and not based on workload.

That review (Sharma and Rani 2020) concluded that the norms need revising on the evidence of workload.

So a ward with three medium-risk patients may not be able to deliver hourly observations. The honest response is to move the patient to where the observations can happen, not to observe less often. And name who takes the next set and when, in the handover and in the notes.

8. Validation of the Score

Does the score predict what it claims to? The tool used to carry the answer, from the derivation study of NEWS (Smith 2013) and what came after it:

In 35,585 acute medical admissions, NEWS discriminated death within 24 hours with an area under the curve of 0.894, better than 33 other early warning scores. NEWS2 kept those weights and changed the oxygen scale and the consciousness row. Indian data come mainly from emergency departments and from sepsis and COVID-19 cohorts; validation on Indian general wards, where the score is meant to be repeated, is limited.

Back to Our Patient

Her NEWS2 is 4, with a red score for new confusion. She needs urgent review by the medical team now, not a recheck in the morning. With a suspected infection, a rise to 5 or more calls for the sepsis team. Whoever reviews her should write down who takes the next set of observations, and when.

Try This

A 66-year-old man with COPD is on 2 L/min of oxygen, with a saturation of 90 per cent. There is no blood gas in his notes. Which saturation scale do you use? Illustrative case.

Answer. Scale 1. The RCP allows Scale 2 only for hypercapnic failure confirmed on a blood gas, with the 88 to 92 per cent target prescribed and recorded by a competent decision maker. On Scale 1 his saturation of 90 per cent scores, and his oxygen adds 2 at any flow. Get the blood gas. If it confirms hypercapnic failure, the target is prescribed and recorded, Scale 2 is used, and Scale 1 is crossed out on the chart.

Take-Home Points

• Read the red score before the total: a 3 in any one parameter calls for urgent review.
• New confusion scores 3; when nobody can say it is old, treat it as new.
• Use Scale 2 only for hypercapnic failure confirmed on a blood gas, with the target prescribed and recorded.
• Suspected infection with a NEWS2 of 5 or more calls for urgent review by a team competent in sepsis.
• If the ward cannot observe as often as the score asks, move the patient, not the frequency.

For the oxygen and the lung behind a falling saturation, see The SpO2/FiO2 Ratio and ARDS: The Evidence Behind the Tool; for the fever itself, Undifferentiated Fever: The Evidence Behind the Pathway.

Declaration: the author built and maintains the tool. It is free to use, and is still in administrator preview, pending clinical review. The figures in this article are those the tool carried in its Evidence tab on 5 November 2026, each from the source the tool names. The illustrative cases follow the tool’s rules; check the tool’s result for a real patient.

References

1. Royal College of Physicians. National Early Warning Score (NEWS) 2: standardising the assessment of acute-illness severity in the NHS. Updated report of a working party. London: RCP; 2017.
2. Royal College of Physicians. National Early Warning Score (NEWS) 2. Chart 1: the NEWS scoring system; Chart 2: NEWS thresholds and triggers; Chart 4: clinical response to NEWS trigger thresholds. London: RCP; 2017.
3. Smith GB, Prytherch DR, Meredith P, Schmidt PE, Featherstone PI. 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;84(4):465-70.
4. Pimentel MAF, Redfern OC, Gerry S, Collins GS, Malycha J, Prytherch D, et al. A comparison of the ability of the National Early Warning Score and the National Early Warning Score 2 to identify patients at risk of in-hospital mortality: a multi-centre database study. Resuscitation. 2019;134:147-56.
5. Echevarria C, Steer J, Bourke SC. Comparison of early warning scores in patients with COPD exacerbation: DECAF and NEWS score. Thorax. 2019;74(10):941-6.
6. Churpek MM, Snyder A, Han X, Sokol S, Pettit N, Howell MD, et al. Quick Sepsis-related Organ Failure Assessment, Systemic Inflammatory Response Syndrome, and early warning scores for detecting clinical deterioration in infected patients outside the intensive care unit. Am J Respir Crit Care Med. 2017;195(7):906-11.
7. Verma A, Farooq A, Jaiswal S, Haldar M, Sheikh WR, Khanna P, et al. National Early Warning Score 2 is superior to quick Sequential Organ Failure Assessment in predicting mortality in sepsis patients presenting to the emergency department in India: a prospective observational study. Int J Crit Illn Inj Sci. 2023;13(1):26-31.
8. Sharma SK, Rani R. Nurse-to-patient ratio and nurse staffing norms for hospitals in India: a critical analysis of national benchmarks. J Family Med Prim Care. 2020;9(6):2631-7.