Predicting 2-day mortality in the pre-hospital setting: A retrospective cohort study comparing the New Zealand Early Warning Score (NZEWS) and paramedic triage Status Codes

Objective To investigate whether the New Zealand Early Warning Score (NZEWS) has greater diagnostic accuracy than current clinical triage Status Codes for predicting 2-day mortality in the Aotearoa New Zealand pre-hospital setting. Methods A retrospective cohort study used de-identified data from the Aotearoa New Zealand Paramedic Care Collection. Records from patients attended by the Hato Hone St John ambulance service between 1 July 2022 and 31 December 2022 were screened for eligibility. Included patients were adults (≥15 years) transported to hospital with a complete set of initial vital signs and an initial clinical triage Status Code. NZEWS was applied retrospectively and compared with clinical triage Status Codes for predicting 2-day mortality using area under the receiver operating characteristic curve (AUROC), sensitivity, specificity, positive predictive value and negative predictive value. The optimal NZEWS threshold was identified using the best balance of sensitivity and specificity. Secondary analyses used Chi-square and Mann–Whitney U tests to describe demographic differences, while logistic regression examined associations between demographic and contextual factors and 2-day mortality. Results The study included 125,278 patients. Both NZEWS and clinical triage Status Codes were strongly associated with 2-day mortality. NZEWS demonstrated superior diagnostic accuracy (AUROC 0.80, 95% CI 0.79–0.81) compared with clinical triage Status Codes (AUROC 0.74, 95% CI 0.73–0.76; DeLong test: Z = 9.66, p < 0.001). The optimal NZEWS threshold for predicting 2-day mortality was lower than previously reported high-risk thresholds (score of 6), identifying more at-risk patients but increasing over-triage. Older age, Māori and Pacific ethnicity and rurality were independently associated with increased 2-day mortality. Conclusion NZEWS outperformed clinical triage Status Codes in predicting 2-day mortality, identifying more at-risk patients at the cost of increased over-triage. NZEWS may support earlier recognition of high-risk patients in the pre-hospital setting, although further research is needed to refine thresholds and guide implementation.