Table 1 Emergency department outcomes

 Deaths in ED (n = 4) [37, 42, 45, 52]

Patients with a MEWS > 4, had higher odds of death in ED death than patients with MEWS< 4 (p < 0.001) [45]. Other studies reported conflicting results: patients who died in ED had

• significantly higher MEWS (p < 0.001) and VIEWS (p < 0.001) scores [37]. or no significant differences in MEWS (p = 0.726) and REMS (p = 0.057) scores compared to patients discharged from ED [42].

• significantly higher MEWS (p = 0.003) and VIEWS (p = 0.002) scores [37], or no significant differences in MEWS (0.306) and REMS (0.402) scores compared to patients admitted to the wards [42].

• had no significant difference in MEWS (p = 0.130) and VIEWS (p = 0.058) scores [37] or MEWS (p = 0.810) and REMS (p = 0.626) scores compared to patients admitted to ICU [42].

Of 115 patients studied by Rocha et al. two died in ED: both had a MEWS score of ≥7 at 6 h of ED care. No patients with MEWS 1, 2–3 or 4–6 at 6 h of ED care died (p < 0.001) [52].

 Clinical deterioration in ED (n = 1) [46]

There was a significant correlation between MEWS and the risk of deterioration in the ED (AUC = 0.830, 95% CI 0.811–0.957, p < 0.001).

 Resolution of clinical instability/treatment effectiveness (n = 2) [20, 38]

Median duration of clinical instability was 39 min (IQR, 5–129 min) and clinical instability was resolved in 64.2% of cases [20]. For patients with dyspnoea, there was a significant decrease in 30-min EWS (compared to 15-min EWS), which along with significant positive changes (towards return to normal) in all vital signs except temperature, indicating ED treatment effectiveness [38].

 Early recognition of trauma severity (n = 1) [52]

Increasing MEWS during first 6 h of ED care had a significant relationship with transfer to the operating room, and ICU admission [52].

 ED length of stay (n = 1) [19]

Patients who fulfilled ED clinical instability criteria during ED care had longer median ED length of stay than patients whose ED care was unaffected by clinical deterioration (7.2 h vs 4.4 h, p < 0.001) [19].

 Clinician use of systems for recognising and responding to deteriorating ED patients (n = 3) [19, 20, 44]

System activation

A study of 204 patients with ED EWS activation showed that: 93.1% activations were by nurses and the median time between documenting physiological abnormalities and activation was 5 min (IQR: 0–20) [20]. Most patients who required an ED EWS activation, had one activation (91.7%), but 7.8% of patients had two activations and one patient (0.5%) had three activations during their ED care [20]. In a study of 472 ED episodes of care, 43.2% of ED patients had ≥1 escalation of care [44]: 56.8% of patients whose track & trigger score exceeded alert threshold had an escalation of care but 40.0% of patients whose track & trigger score did not exceed alert threshold still had escalation of care based presumably on clinician concern [44].

Deterioration triggers

Hypotension, tachycardia, bradycardia and tachypnoea were the most common reasons for ED EWS activation [20].

The most common episodes of documented physiological abnormalities were tachypnoea (34%) followed by tachycardia (29%) and hypotension (17%) [19].

Unreported clinical deterioration during ED care

Unreported deterioration decreased with each stage of ED RRS implementation but was not statistically (p = 0.141) [19]:

• Clinician discretion and no track and trigger charts (n = 150 patients): 86.7%

• ED CIC and escalation of care protocol, no track and trigger charts (n = 150 patients): 68.8%

• ED CIC and escalation of care protocol, new track and trigger charts (n = 150 patients): 55.3%

• ED CIC and escalation of care protocol, 12 months after track and trigger charts (n = 150 patients): 54.0%

Completion of documentation

A study of 2965 sets of vital signs from 472 ED episodes of care showed 85.8% of patients had documentation of ≥1 complete set of six vital signs: 87.6% sets of vital signs contained HR, RR, BP and SpO₂ and overall, 25.6% of vital signs were complete [44].

Track and trigger scoring

A total of 34.5% of observations (n = 2965 vital signs) contained a track & trigger score and 60.6% of patients had ≥1 track & trigger score documented in the ED [44]. However, 20.6% of track & trigger scores were incorrect, 79.1% of the incorrect track & trigger totals were underscored, and 93.4% of track & trigger score errors were from incorrect assignment of the score to an individual vital sign [44].