Table 5 Evidence table: Development and validation studies – Patient groups differentiated by (suspected) condition

Albright et al. (2014) [29], USA

Risk of bias: Low

850 pregnant & post partum women with suspected SIRS/sepsis

MEWS (≥5)

REMS (≥6)

ICU Admission within 48 h prediction

MEWS: Sensitivity 100.0%, Specificity 77.6%

REMS: Sensitivity 77.8%, Specificity 93.3%

Cildir et al. (2013) [38], Turkey

Risk of bias: Low

230 diagnosed with community acquired sepsis.

CCI (>5)

MEWS (≤5)

28-day mortality

CCI: AUROC 0.65 (p = 0.001)

MEWS: AUROC 0.61 (p = 0.008)

28-day mortality (n = 64 with sepsis)

CCI: AUROC 0.65 (p = 0.18)

MEWS:AUROC 0.57 (p = 0.48)

28-day mortality (n = 166 with severe sepsis)

CCI: AUROC 0.62 (p = 0.006)

MEWS: AUROC 0.60 (p = 0.04)

Considine et al. (2015) [39], Australia

Risk of bias: Low

600 adult with presenting with SOB, chest pain or abdominal pain

ED CIC

Episodes of unreported clinical deterioration

T0 (Clinical decision making) (86.7%);

T1 (Escalation of care protocol) (68.8%);

T2 (Escalation of care protocol, single parameter TTS chart) (55.3%);

T3 (Escalation of care protocol, single parameter TTS chart (year 2012)) (54.0%);

(p = 0.14).

Corfield et al. (2014) [75] (and related conference abstract Corfield et al. (2012) [40], Scotland

Risk of bias: Low

2003 with sepsis (suspected or confirmed within 2 days of attendance and 2 or more of sepsis criteria)

NEWS (≥9 versus 0–4)

ICU (within 2 days)

OR 5.76 (95% CI 3.22–10.31; p = 0.00)

Mortality (30 days)

OR 5.64 (95% CI 3.70–8.60; p = 0.00)

Combined (ICU and/or mortality)

9–20: OR 5.78 (95% CI 4.02–8.31; p = 0.00)

Cut-off point with highest Youden’s Index: NEWS 9

Geier et al. (2013) [32] Germany

Risk of bias: Low

151 with suspected sepsis

ESI

MEWS

CCI Score

In-hospital mortality

ESI: Sensitivity 0.73, Specificity 0.0

MEWS: Sensitivity 0.43, Specificity 0.74

CCI: Sensitivity 0.82, Specificity 0.64

Howell et al. (2007) [45], USA

Risk of bias: Low

2132 with suspected infection

mREMS

28-day in-hospital survival

AUROC 0.80 (95% CI 0.75–0.85)

Jo et al. (2013) [46], Korea

Risk of bias: Low

299 patients with blunt trauma, Injury severity score ≥ 9

VIEWS-L

In-hospital mortality

AUROC: 0.83 (95% CI 0.77–0.91)

Jo et al. (2016) [47], Korea

Risk of bias: Low

553 with pneumonia

NEWS-L score (≥3.1)

NEWS (≥5)

In-hospital mortality

NEWS-L: AUROC 0.73 (0.66–0.80)

NEWS: AUROC 0.70 (0.63–0.77)

Jones et al. (2005) [48], USA

Risk of bias: Low

91 with initial ED vital signs consistent with shock

SAPS II

MPM0 II

LODS

In-hospital mortality

SAPS II: AUROC 0.72 (95% CI 0.57–0.87)

MPM0 II: AUROC 0.69 (95% CI 0.54–0.84)

LODS: AUROC 0.60 (95% CI 0.45–0.76)

Nguyen et al. (2012) [59], USA

Risk of bias: Unclear

541 with severe sepsis

PIRO

APACHE II

In-hospital mortality

PIRO: AUROC 0.71 (95% CI 0.66–0.75)

APACHE II: AUROC 0.71 (95% CI 0.66–0.76)

Vorwerk et al. (2009) [51], UK

Risk of bias: Low

307 with sepsis

MEWS (≥5) Blood lactate (≥4 mmol/l)

28-day mortality

MEWS: AUROC 0.72 (95% CI 0.67 to 0.77)

Lactate: AUROC 0.62 (0.54 to 0.70)

Williams et al. (2016) [52], Australia

Risk of bias: Low

8871 with presumed infection

SAPS II)

SOFA

APACHE II

30-day mortality

APACHE II: AUROC 0.90 (0.88–0.91)

SAPS II: AUROC 0.90 (0.89–0.92)

SOFA: AUROC 0.86 (0.84–0.88)