Clinical deterioration on the ward: NEWS2 and escalation

It is three in the morning in a surgical ward. A 68-year-old man is on his second day after a colectomy. On rounds, the nurse records a respiratory rate of 24 per minute, measures oxygen saturation of 93 on room air, systolic pressure of 102, pulse of 112 and temperature of 38.4 degrees. He responds, but asks twice where he is and wants to remove the IV line. At the midnight rounds, he was oriented. No single number, on its own, seems dramatic. The scene is fictional and was written for teaching.

This text uses the picture as a clinical reasoning scenario for nursing. The trap is looking at each vital sign in isolation, finding each one "acceptable" and going back to the station. The sum of the signs tells another story, and that is exactly what an early warning score does: it turns the vague impression that something is wrong into a number that forces action. The text does not replace your institution's protocol. If you play the case of the day in NEXO, you will recognize the challenge: putting small clues together before the picture becomes obvious.

Key points

• NEWS2 adds up seven physiological parameters, and the sum weighs more than any single sign.
• New confusion scores as much as responding only to voice, and it should not be put down to age or the night without assessment.
• Respiratory rate is one of the strongest predictors of deterioration and one of the most poorly measured signs.
• The score range defines how often to reassess and who should be called, and with what urgency.
• Structured communication, such as SBAR, helps turn the number into action on the other end of the line.
• The nurse's concern is a legitimate trigger to call for help, even with a low score.

Why this is a reasoning scenario

Clinical deterioration on the ward rarely begins with a cardiac arrest. It usually comes announced by hours of small changes: breathing that speeds up, a pulse that rises, pressure that gives a little, a person who seems different. Those closest to seeing these changes are the nursing team, who measure the signs, talk with the patient and notice when he "is not well". The challenge is converting that perception into a response before the window closes.

Early warning scores were created for this. A narrative review of the British score compares the instrument to a fire alarm: it should be judged by the number of fires it helps prevent or put out, and not by predicting with precision who will die (Holland and Kellett, 2023). The authors note that the score was designed to identify who needs to be seen urgently by a trained professional, and not to diagnose a specific disease.

What the scene teaches

The scene trains three habits. The first is to add up before judging: each "acceptable" sign can add up to a high score. The second is to treat new confusion as a vital sign, and not as a personality trait of the older patient at night. The third is to escalate clearly, stating the number, what changed and what you need the other professional to do. The post on clinical decision-making discusses how to act with incomplete information.

What NEWS2 measures

The United Kingdom's national early warning score, NEWS, adds points for seven measurements: respiratory rate, oxygen saturation, use of supplemental oxygen, temperature, systolic pressure, pulse and level of consciousness. Each measurement receives from zero to three points as it moves away from the normal range (Mitsunaga and colleagues, 2019). In the table used in the study by Mitsunaga and colleagues, for example, a respiratory rate of 21 to 24 is worth 2 points and 25 or more is worth 3; saturation of 92 to 93 is worth 2 points; systolic pressure of 101 to 110 is worth 1 point; and a pulse of 111 to 130 is worth 2 points (Mitsunaga and colleagues, 2019).

The 2017 version, NEWS2, brought two main changes. It added new confusion to the consciousness scale, which went from AVPU to ACVPU, and created a second saturation scale for patients with hypercapnic respiratory failure or at risk of it (Kostakis et al., 2021). For most patients, according to the same authors, the two versions are practically identical (Kostakis et al., 2021). On the consciousness scale, responding only to voice, only to pain or not responding are each worth 3 points (Mitsunaga et al., 2019), and new confusion falls into that same group of alterations (Kostakis et al., 2021).

Saturation scale 2

Patients with type 2 respiratory failure, or at risk of it, usually have a lower saturation target, from 88 to 92, according to the oxygen therapy guidelines cited by Pimentel et al. (Pimentel et al., 2019). Scoring these patients on the common scale would generate false alerts and could lead to too much oxygen. Scale 2 of NEWS2 assigns points at lower saturation thresholds and also scores high saturation when the patient is on oxygen, reflecting the concern about carbon dioxide retention (Pimentel et al., 2019).

The detail matters for nursing because choosing the scale is a clinical decision that must be recorded. Using scale 2 in someone without this risk is an error. In the study by Pimentel et al., applying scale 2 to patients without risk of hypercapnia worsened the score's performance compared with the original NEWS (Pimentel et al., 2019). In our patient, with no known chronic lung disease, scale 1 is the correct one.

How accurate the score is

In the database analyzed by Pimentel et al., the area under the curve for in-hospital death within 24 hours was around 0.86 to 0.91, depending on the respiratory risk group (Pimentel et al., 2019). A study in patients assessed by rapid response teams in Canada found good accuracy of NEWS2 for in-hospital mortality, with sensitivity of 84.5 percent at the 5-point threshold, but with low specificity, which can fuel alarm fatigue (Fernando et al., 2019). During the covid-19 pandemic, the score's performance remained similar between groups with and without the infection (Kostakis et al., 2021).

How the picture presents

Let us return to the patient. Adding up the points with the table, a rate of 24 is worth 2, saturation of 93 on room air is worth 2, pressure of 102 is worth 1, pulse of 112 is worth 2, and temperature of 38.4 is worth 1 (Mitsunaga et al., 2019). New confusion adds 3 (Kostakis et al., 2021). The total reaches 11 points. Each sign on its own seemed tolerable. Together, they place the patient in the highest risk band.

The table with the response bands recommended by the Royal College of Physicians appears in the review by Holland and Kellett (Holland and Kellett, 2023). With 7 points or more, the recommendation is continuous monitoring of vital signs, immediate notification of the medical team and emergency assessment by a team with critical care competence (Holland and Kellett, 2023).

NEWS2 band	Minimum reassessment frequency	Recommended response
0	Every 12 hours	Continue routine monitoring
1 to 4	Every 4 to 6 hours	Nurse assesses and decides whether to increase the frequency or escalate
5 or more, or 3 points in a single parameter	Every hour	Urgently notify the medical team; urgent assessment by a trained professional; care in a monitored setting
7 or more	Continuous monitoring	Immediately notify the medical team; emergency assessment by a team with critical care competence

Table source: Royal College of Physicians escalation protocol, reproduced by (Holland and Kellett, 2023). The institution may adapt the bands, and the authors themselves advocate local protocols that state how to respond to the score and when to override it with clinical judgment.

Clues that point to the cause

The score says the patient is seriously ill, but it does not say why. That is why, alongside the number, the nurse records what was seen: abdominal pain stronger than the day before, distended abdomen, drain with cloudy fluid, scant urine output in recent hours, calf pain, sudden shortness of breath, dressing stained with blood. Each finding pushes the reasoning toward a different cause. The post on signs, symptoms and history shows how to read these clues.

Differential diagnosis on the ward

On the second day after intestinal surgery, a patient with fever, tachypnea, tachycardia, falling blood pressure and new confusion has sepsis until proven otherwise. But the list is longer, and nursing needs to know it in order to collect the data that help separate the hypotheses. The table summarizes the reasoning.

Hypothesis	Clues in favor	What nursing checks
Sepsis of abdominal origin	Fever, worsening abdominal pain, distension	Appearance of the drain and the wound, urine output, blood glucose
Pneumonia or atelectasis	Cough, pain on breathing, low oxygen saturation	Auscultation, breathing pattern, secretions
Pulmonary embolism	Sudden shortness of breath, tachycardia, chest pain	Calves, whether thrombosis prophylaxis was given
Bleeding	Falling blood pressure, pallor, rapid pulse	Dressing, drains, stools, recent hemoglobin
Effect of opioid or sedative	Drowsiness, slow breathing	Prescription, dose times, pupils
Hypoglycemia	Confusion, sweating	Capillary blood glucose at the bedside

The post on differential diagnosis with clues shows how to organize this list when time is short.

Tachypnea: the sign that arrives first

In a multicenter study of more than 269,000 ward admissions, respiratory rate, pulse, age and systolic blood pressure were the most important variables for predicting cardiac arrest, transfer to intensive care or death (Churpek et al., 2016). Risk had a U shape for respiratory rate, pulse and blood pressure, that is, it rose with both high and low values (Churpek et al., 2016). In our case, the rate of 24 is the first clue, and the easiest one to miss.

Confusion: the brain as a sentinel organ

New confusion in an older surgical patient is often called "postoperative delirium" and treated as a nuisance. It may be that, but NEWS2 gives it the same weight as responding only to voice precisely because an acute change in mental status can reflect hypoperfusion, hypoxia, hypoglycemia or infection (Kostakis et al., 2021). The nursing question is not "is he confused?" but "is he different from how he was?". Comparison with the previous assessment is the data that matters.

Nursing approach: assess and measure

The first task is to measure well. A score calculated from poorly measured signs gives a false sense of safety. The second is to measure again, at the frequency the range requires, to see the trend. The third is to record what the score does not capture.

Counting breathing the right way

A systematic review of manual measurement of respiratory rate found five sources of error (Kallioinen et al., 2021). When the patient notices they are being observed, the rate falls. Short counts of 15 or 30 seconds multiplied up underestimate the value. There is wide variation between observers and within the same observer. Certain numbers appear too often in the records, suggesting estimation instead of counting. And the rate is sometimes not recorded at all (Kallioinen et al., 2021). The authors' recommendation for nursing is to count, not estimate, to use 60-second counts whenever possible, to do this without the patient noticing and to record the value obtained (Kallioinen et al., 2021).

In practice, a good moment is right after taking the pulse, keeping your fingers on the wrist and watching the chest for a full minute. The patient thinks you are still counting the pulse and breathes normally.

What to measure beyond the seven parameters

• Capillary blood glucose, because hypoglycemia causes confusion and is corrected in minutes.
• Urine output in recent hours, because the drop usually comes before frank hypotension.
• Pain, with scale and location, compared with the day before.
• Appearance of the wound, drains and tubes.
• Skin: temperature of the extremities, color, capillary refill.
• Medications received in recent hours, especially opioids and sedatives.

These data do not enter the score sum, but they enter the conversation with whoever will assess the patient. In an Australian study with more than 10,000 audits, when there was a rapid response team trigger, the nursing interventions most associated with correct escalation included electrocardiogram, Glasgow Coma Scale and capillary blood glucose (Considine et al., 2023).

First measures within the scope of nursing

While help arrives, nursing can position the patient with the head of the bed raised, offer oxygen according to the institution's protocol and the appropriate saturation target, ensure a patent venous access, prepare the material for collecting tests and maintain continuous monitoring. In the Australian study, when there was a rapid response team trigger, at least one nursing intervention was recorded in 63.7 percent of the audits (Considine et al., 2023). Medical management, such as antibiotics, fluids and imaging tests, depends on the assessment you are requesting.

Nursing management: escalate and communicate

With 11 points, the patient is in the 7 or more range. The recommended response is to notify the medical team immediately and request emergency assessment by a team with competence in critical care, with continuous monitoring (Holland and Kellett, 2023). In many hospitals, this means activating the rapid response team.

Rapid response teams

A systematic review found moderate evidence that implementing rapid response teams is associated with lower hospital mortality and fewer cardiac arrests outside intensive care (Hall et al., 2020). The effect on transfers to intensive care was inconclusive (Hall et al., 2020). The team only helps, however, if it is called. The arm that triggers the alarm, which is almost always nursing, is as important as the arm that responds.

SBAR: saying what matters in one minute

SBAR organizes communication into situation, background, assessment and recommendation. A systematic review found moderate evidence of improved patient safety with the implementation of SBAR, especially when used to structure telephone communication (Müller et al., 2018). For our patient, the call might sound like this:

• Situation: "I am the nurse from surgery, bed 12. The patient has a NEWS2 of 11, with new confusion."
• Background: "Second day after colectomy. At midnight he was oriented and had a score of 2."
• Assessment: "Respiratory rate 24, saturation 93 on room air, systolic 102, pulse 112, temperature 38.4. Abdomen more distended and low urine output. I think it may be sepsis."
• Recommendation: "I need someone from the rapid response team to come now. I have already applied oxygen, the head of the bed is raised and the access is patent."

The recommendation is the part most often omitted. Saying what you need, and in how long, prevents the call from ending with "I will come by later". The post about interprofessional clinical reasoning discusses how different professions build the decision together.

Nurse concern as a trigger

Not every deterioration shows up in the score. In a retrospective cohort study with 4634 rapid response team activations, 1263 were motivated only by nurse concern, with no change in vital signs (Gentil et al., 2025). The group activated only by concern needed fewer transfers to intensive care than the group activated by altered vital signs, 40 versus 50 percent, although the observational design does not allow a claim of cause (Gentil et al., 2025). The practical message is that recorded and communicated concern is a clinical datum.

Why the alert fails

Having a score is not enough. In the Australian study already cited, when there was a rapid response team or pre-emergency trigger, activation of the rapid response system occurred in only 26.8 percent of the audits (Considine et al., 2023). The authors concluded that there were gaps in escalation according to policy, although nursing performed several interventions within its own scope (Considine et al., 2023).

The barriers that nursing describes

A qualitative study with nurses from clinical wards in Canada identified eleven themes of barriers and facilitators to calling the rapid response team, united by a central idea: the complexity that the nurse imposes on the decision to call (Fitzgerald et al., 2026). Among the facilitators were consulting more experienced colleagues and the confidence that grows with experience (Fitzgerald et al., 2026). The article's title sums up the most common fear: not wanting to raise a false alarm (Fitzgerald et al., 2026).

Alarm fatigue and false security

The score also fails in both directions. With a high threshold, many patients who will die fall below it. The review by Holland and Kellett estimates that a threshold of 7 points would trigger an alert in about 4 percent of observations and detect just over half of deaths within 24 hours, while a threshold of 3 points would detect almost 9 in 10 deaths, at the cost of alerts in more than a quarter of observations (Holland and Kellett, 2023). With a low threshold, specificity drops and alarms pile up (Fernando et al., 2019). That is why the score organizes clinical judgment but does not replace it.

There is also the context. In an English health system that adopted the score in the emergency department, ambulances, community and primary care, most records outside the hospital had scores of 0 to 2, and fewer than a fifth reached 5 or more (Scott et al., 2019). The same number can carry different weight depending on the place and the patient, and nursing staff need to read the score in light of who is in the bed.

Step-by-step reasoning in the case

First, the measurement: the nurse counts the breathing for 60 seconds, without telling the patient, and confirms 24 per minute.

Second, the sum: with saturation of 93 on room air, systolic of 102, pulse of 112, temperature of 38.4 and new confusion, the NEWS2 reaches 11.

Third, the trend: at midnight, the score was 2. The change matters as much as the value.

Fourth, the extra data: abdomen more distended, low urine output, normal blood glucose, no extra dose of opioid.

Fifth, the first measures: head of bed raised, oxygen per protocol, patent venous access, continuous monitoring.

Sixth, the call: SBAR to the rapid response team, with a clear recommendation for immediate assessment.

Seventh, the reassessment: the nurse repeats the vital signs while the team arrives and records the time of each step.

Common mistakes

The first mistake is judging sign by sign. Each one seems acceptable, but the sum is high.

The second mistake is estimating the respiratory rate. Counting for 60 seconds changes the number.

The third mistake is attributing the confusion to age or the time of day. New confusion scores high and calls for investigation.

The fourth mistake is using the saturation scale 2 without a recorded indication.

The fifth mistake is waiting for the next round. The 7 or more range calls for an emergency assessment.

The sixth mistake is calling without a recommendation. Saying what you need shortens the path to the bedside.

The seventh mistake is ignoring your own concern because the score is low. The post on diagnostic pause shows how to stop and review a picture before closing the conclusion.

How NEXO helps train this reasoning

To use the case of the day as training, try three habits. Before the first guess, ask which vital signs you would add up and which range the patient would be in. With each clue, ask whether it changes the urgency or only the cause. After the result, note which small clue you let slip by.

Three exercises with cases

Add up the signs. For each case with vital signs, calculate the score and say what the response would be by range.

Write the SBAR. In four sentences, summarize the case as if you were calling the on-call team.

Look for the trend. Picture the same signs six hours earlier and ask what would have changed in the management.

The post on simulation of patient care shows how to train patient assessment, and the post on case-based learning with clinical cases shows how cases help cement that habit.

The limits on using the game are described on the NEXO Economy page. If you research the teaching of clinical reasoning, the page for Researchers explains how to get in touch. Professionals who write or review cases find information on Curator and Creators, and institutions can learn about the proposal on the partnerships page.

Frequently asked questions

Which parameters go into NEWS2?

Respiratory rate, oxygen saturation, use of supplemental oxygen, temperature, systolic blood pressure, pulse and level of consciousness on the ACVPU scale (Kostakis et al., 2021).

What should be done with a NEWS2 of 7 or more?

Continuous monitoring, immediate alert to the medical team and emergency assessment by a team skilled in critical care (Holland and Kellett, 2023).

Does a single parameter scoring 3 already call for escalation?

Yes. According to the response table, 3 points in a single parameter fall in the same band as 5 or more, with reassessment every hour and an urgent alert to the medical team (Holland and Kellett, 2023).

When should the saturation scale 2 be used?

Only in patients with hypercapnic respiratory failure or at risk of it, with a target of 88 to 92 (Pimentel et al., 2019).

Can I call the rapid response team just because I am worried?

Yes. Nurse worry is used as a trigger in several hospitals, and one study linked this trigger to fewer transfers to intensive care (Gentil et al., 2025).

Final checklist

• I counted the respiratory rate for 60 seconds without telling the patient.
• I added up the seven parameters and chose the right saturation scale.
• I compared the score and the mental state with the previous assessment.
• I measured blood glucose, urine output and pain, and looked at the wound and drains.
• I applied the response for the score band.
• I made the call in SBAR, with a clear recommendation.
• I took the first measures within my scope.
• I reassessed and recorded the time of each step.
• I took my own concern into account even when the score was low.

Want to practice reasoning with progressive clues? Play the case of the day.

References

1. Churpek MM, Yuen TC, Winslow C, et al. Multicenter Comparison of Machine Learning Methods and Conventional Regression for Predicting Clinical Deterioration on the Wards. Crit Care Med. 2016. PMID 26771782.
2. Considine J, Hutchinson AM, Mitchell I, et al. Vital sign assessment and nursing interventions in medical and surgical patients with rapid response system triggers. J Clin Nurs. 2023. PMID 37365897.
3. Fernando SM, Fox-Robichaud AE, Rochwerg B, et al. Prognostic accuracy of the Hamilton Early Warning Score (HEWS) and the National Early Warning Score 2 (NEWS2) among hospitalized patients assessed by a rapid response team. Crit Care. 2019. PMID 30791952.
4. Fitzgerald L, Sears K, Wilson R, et al. "We Don't Want to Cry Wolf": A Qualitative Study About Nurses' Experiences Activating Rapid Response Teams. Can J Nurs Res. 2026. PMID 41364454.
5. Gentil LLS, Nascimento MS, Jaures M, et al. Nurse worry as a trigger for rapid response team activation improving outcomes: a retrospective cohort study in non-critical units. BMC Nurs. 2025. PMID 39806453.
6. Hall KK, Lim A, Gale B. The Use of Rapid Response Teams to Reduce Failure to Rescue Events: A Systematic Review. J Patient Saf. 2020. PMID 32809994.
7. Holland M, Kellett J. The United Kingdom's National Early Warning Score: should everyone use it? A narrative review. Intern Emerg Med. 2023. PMID 36602553.
8. Kallioinen N, Hill A, Christofidis MJ, et al. Quantitative systematic review: Sources of inaccuracy in manually measured adult respiratory rate data. J Adv Nurs. 2021. PMID 33038030.
9. Kostakis I, Smith GB, Prytherch D, et al. The performance of the National Early Warning Score and National Early Warning Score 2 in hospitalised patients infected by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Resuscitation. 2021. PMID 33176170.
10. Mitsunaga T, Hasegawa I, Uzura M, et al. Comparison of the National Early Warning Score (NEWS) and the Modified Early Warning Score (MEWS) for predicting admission and in-hospital mortality in elderly patients in the pre-hospital setting and in the emergency department. PeerJ. 2019. PMID 31143553.
11. Müller M, Jürgens J, Redaèlli M, et al. Impact of the communication and patient hand-off tool SBAR on patient safety: a systematic review. BMJ Open. 2018. PMID 30139905.
12. Pimentel MAF, Redfern OC, Gerry S, 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. PMID 30287355.
13. Scott LJ, Redmond NM, Garrett J, et al. Distributions of the National Early Warning Score (NEWS) across a healthcare system following a large-scale roll-out. Emerg Med J. 2019. PMID 30842204.