The Emergency Severity Index (ESI) Version 5: Simulation of Predictive Validity and Triage Level Distribution

The Emergency Severity Index (ESI) was updated to the 5th version which emphasizes identification of abnormal vital signs in low acuity patients (ESI levels 3, 4, and 5) to prevent undertriage.

We conducted a simulation of the ESI 5th version (sESI-v5) with data of consecutive emergency department (ED) patients to (1) investigate changes in triage level distribution caused by vital signs assessment, (2) investigate the predictive validity (assessing resource utilization, length of stay, disposition, and mortality), and (3) identify individual cases with potential benefits from the ESI update.

In this prospective cohort, ESI 4th version (ESI-v4) was used for triage. sESI-v5 was applied retrospectively, and predictive validity (resource utilization, ED length of stay, disposition, and mortality) was compared. Patients with potential benefits from sESI-v5 (uptriage from ESI-v4 level 3, 4, or 5 to sESI-v5 level 2 with admission to intensive care unit [ICU] or early death) were identified.

In 6,230 adult ED patients, uptriage to ESI level 2 would have occurred in 636 patients (10.2%). The sESI-v5 showed a similar predictive validity to ESI-v4. We identified 30 (0.5%) patients with potential benefits from sESI-v5, as they were admitted to ICU or died within 30 days (29 were uptriaged from ESI-v4 level 3, 1 from ESI-v4 level 4, and none from ESI-v4 level 5).

Adopting ESI-v5 could increase the identification of high-risk patients but at the cost of significantly expanding the number of patients classified as ESI level 2. This shift may overwhelm ED resources and delay care for truly critical patients without clear evidence of improved outcomes. While ESI-v5 demonstrated similar predictive validity to ESI-v4, only a small subset of patients would have potentially benefited from the update.