Modified Early Warning Score (MEWS) for clinical deterioration

Clinical background

MEWS is a track-and-trigger instrument that meets a concrete clinical need: to identify at the bedside those inpatients who are deteriorating before the deterioration becomes obvious. Two to five per cent of inpatients outside intensive care suffer clinical deterioration leading to death or transfer to intensive care, and in retrospect it has often been possible to see that abnormal vital signs preceded the event by several hours [2]. The score is not intended to predict a specific diagnosis, but to flag that the patient's physiological reserves are beginning to fail and that structured escalation is needed.

Calculating the Modified Early Warning Score

MEWS is the sum of points for five vital signs, each assessed against fixed thresholds:

$$\text{MEWS} = \text{RR}$${\text{points}} + \text{HR}{\text{points}} + \text{SBP}{\text{points}} + \text{Temp}{\text{points}} + \text{AVPU}_{\text{points}}

The five variables are respiratory rate, heart rate, systolic blood pressure, temperature and level of consciousness by the AVPU scale (Alert, responds to Voice, responds to Pain, Unresponsive). Each variable scores 0 to 3 points except temperature, which scores 0 or 2. The theoretical range is 0 to 14 points. All the variables can be measured at the bedside without laboratory tests or electronic equipment apart from a blood pressure cuff and a thermometer.

The derivation cohort consisted of 709 acute medical admissions to a 56-bed medical admissions unit at Wrexham Maelor Hospital, Wales, during March 2000 [1]. The outcomes were death, intensive care admission, HDU admission or cardiac arrest within 60 days. In this cohort, a score of 5 or more was associated with an increased risk of death (OR 5.4, 95% CI 2.8 to 10.7), intensive care admission (OR 10.9, 95% CI 2.2 to 55.6) and HDU admission (OR 3.3, 95% CI 1.2 to 9.2) [1].

Interpretation in practice

MEWS is not a diagnostic test but a trigger for action. The score governs the frequency of observations and the level of escalation, and the exact thresholds for what should be triggered at each level should be defined in local rapid response protocols. The commonly used bands rest on the derivation study and later validations:

A patient with a low score may still deteriorate. The trend is at least as important as the single value: a rising MEWS over time should trigger escalation even if the absolute value has not crossed the threshold.

Validation and performance

In the largest comparative study to date, comprising 1,487,263 admissions at 28 hospitals in California and Illinois, MEWS had an AUC of 0.83 (95% CI 0.83 to 0.84) and 0.84 (95% CI 0.84 to 0.85) for in-hospital death in the two states respectively [2]. This was lower than for the National Early Warning Score (NEWS), which reached an AUC of 0.87 and 0.86 in the same cohorts [2]. In an external validation of 383 emergency admissions at a hospital in Shenzhen, China, MEWS reached an AUC of 0.83 (95% CI 0.786 to 0.881) for in-hospital death, with an optimal threshold at 3.5 points giving a sensitivity of 87% and a specificity of 66% [3]. The negative predictive value at this threshold was 97%, supporting the usefulness of low scores for ruling out deterioration.

Performance falls when the instrument is used in patients who have already deteriorated. In a Korean study of 6,729 events in which a medical emergency team had already been called, the AUC for MEWS was only 0.58 (95% CI 0.56 to 0.59) for 28-day mortality [4]. In this population the score therefore had essentially no discriminatory ability, which is logical: the instrument was designed to capture early deterioration on general wards, not to risk-stratify patients who have already been the subject of escalation.

Limitations

MEWS applies to adult inpatients on general wards and acute admissions units. It is not validated for children, pregnant women or postoperative patients on obstetric or gynaecological wards. The score should not be used as the sole basis for a decision about intensive care admission, nor to assess prognosis in patients who have already triggered a rapid response team [4].

Several local variants of MEWS exist with different thresholds or additional variables, and it is essential to confirm which version has been validated locally. In many health systems, NEWS or NEWS2, which include oxygen saturation and oxygen therapy, have in practice replaced MEWS as the primary track-and-trigger instrument, since they have shown higher discrimination in direct comparisons [2].

A specific pitfall is that MEWS as a static snapshot does not capture the trend. A patient with a low but rising score may be heading for deterioration, and a patient with a high but falling score may have responded to treatment. Repeated measurements and awareness of the direction of travel are therefore essential. MEWS furthermore lacks variables for oxygen saturation and oxygen therapy, so patients whose respiratory deterioration is compensated by oxygen may be underrated.

References

1. Subbe CP, Kruger M, Rutherford P, Gemmell L. Validation of a modified Early Warning Score in medical admissions. QJM 2001;94(10):521-6. PMID: 11588210
2. Liu VX, Lu Y, Carey KA et al. Comparison of Early Warning Scoring Systems for Hospitalized Patients With and Without Infection at Risk for In-Hospital Mortality and Transfer to the Intensive Care Unit. JAMA Netw Open 2020;3(5):e205191. PMID: 32427324
3. Xie X, Huang W, Liu Q et al. Prognostic value of Modified Early Warning Score generated in a Chinese emergency department: a prospective cohort study. BMJ Open 2018;9(1):e024120. PMID: 30552276
4. Ahn JH, Jung YK, Lee JR et al. Predictive powers of the Modified Early Warning Score and the National Early Warning Score in general ward patients who activated the medical emergency team. PLoS One 2020;15(5):e0233078. PMID: 32407344