Objective To assess the safety and efficiency of triaging low urgent self-referred patients at the emergency department (ED) to a general practitioner (GP) based on the Manchester triage system (MTS).

Methods All self-referred patients in the evening, night and weekends were included in this prospective observational study. Patients were triaged by an ED nurse according to the MTS and allocated to a GP or the ED according to a predefined care scheme. For patients treated by the GP, assessments were made of safety as measured by hospitalisation and return to the ED within 2 weeks, and efficiency as measured by referral to the ED.

Results In 80% of cases allocation of the self-referrals to the ED or GP was according to a predefined scheme. Of the 3129 low urgent self-referred patients triaged to the GP, 2840 (90.8%) were sent home, 202 (6.5%) were directly referred to the ED, 36 (1.2%) were hospitalised. Within a random sample of low urgent patients sent home by the GP (222 of 2840), 8 (3.6%) returned to the ED within 2 weeks. Against the agreed MTS scheme, the ED also directly treated 664 low urgent patients, mainly for extremity problems (n=512). Despite the care agreements, 227 urgent patients were treated by the GP, with a referral rate to the ED of 18.1%, a hospitalisation rate of 4.0% and a 4.5% return rate to the ED within 2 weeks.

Conclusions Low urgent self-referrals, with the exception of extremity problems, were shown to be treated efficiently and safely by a GP. A selected group of more urgent patients also seem to be handled adequately by the GP. Triage of low urgent patients with extremity problems and reasons for nurses not following a predefined triage allocation scheme need further elaboration.