A prospective, internal validation of an emergency patient triage tool for use in a low resource setting

Aim: Assess the performance of a simple triage disposition score based on mental status, mobility and either oxygen saturation or respiratory rate by three principal metrics: 24 h mortality, the need for hospital admission and the urgency ranking of patient presentations. Method: Prospective observational non-interventional study of consecutive patients presenting to the emergency and outpatient departments of a low-resource sub-Saharan hospital Results: Out of 14,585 consecutive patients arriving to hospital 1,804 (12.4%) were admitted and 39 died (0.3%) within 24 hours. No patients with normal mental status or a stable independent gait died within 24 h, and 95% of those who did had an oxygen saturation <94%. The c statistic of the score for death within 24 hours was >0.95 and not significantly changed if respiratory rate replaced oxygen saturation as a score component, or mental status was assessed subjectively or objectively. However, an objective measure of mental status significantly reduced the c statistic for hospital admission from 0.970 SE 0.003 to 0.956 SE 0.004, p 0.002. The score attributed a higher acuity rating than the South African Triage System urgency ranking of presentations to 11.1% of patients and a lower acuity rating to 1.3%. However, 53% of the patients given a higher acuity rating were subsequently admitted to hospital and 6.1% of them died. Conclusion: The score identified patients who subsequently required hospital admission and who were likely to die within 24 hours.