Objective: To investigate whether nurses performing triage are able to predict the need for admission of patients attending the emergency department (ED) with sufficient accuracy to facilitate hospital bed management.
Methods: A prospective observational study was performed in which nurses performing triage, in a large urban UK hospital, were asked to predict whether patients would ultimately be admitted or discharged from the ED.
Results: 3144 patients attended the ED during the trial period, of which 296 were excluded from the study. The positive predictive value of the nurse performing triage's prediction for the whole study cohort was 54.23.
Conclusion: Predicting admission at triage is not sufficiently accurate to inform hospital in-patient bed management systems. The decision to admit can only be determined after a comprehensive clinical work up and patients cannot be accurately 'signposted' during the triage process.