Triage Ability of Emergency Medical Services Providers and Patient Disposition: A Prospective Study

Abstract Study objective: To determine the ability of emergency medical services (EMS) providers to subjectively triage patients with respect to hospital admission and to determine patient characteristics associated with increased likelihood of admission. Methods: A prospective, cross-sectional study of a consecutive sample of patients arriving by ambulance during the month of February 1997 at an urban, university hospital, Emergency Department. Emergency medical services providers completed a questionnaire asking them to predict admission to the hospital and requested patient demographic information. Predictions were compared to actual patient disposition. Results: A total of 887 patients were included in the study, and 315 were admitted to the hospital (36%). With respect to admission, emergency medical services providers had an accuracy rate of 79%, with a sensitivity of 72% and specificity of 83% (kappa = 0.56). Blunt traumatic injury and altered mental status were the most common medical reasons for admission. Variables significantly associated with high admission rates were patients with age > 50 years, chest pain or cardiac complaints, shortness of breath or respiratory complaints, Medicare insurance, and Hispanic ethnicity. The emergency medical services providers most accurately predicted admission for patients presenting with labor (kappa = 1.0), shortness of breath / respiratory complaints (kappa = 0.84), and chest pain (kappa = 0.77). Conclusion: Emergency medical services providers can predict final patient disposition with reasonable accuracy, especially for patients presenting with labor, shortness of breath, or chest pain. Certain patient characteristics are associated with a higher rate of actual admission.