Primary care streaming in emergency departments and perceived mechanisms for effectiveness
To manage increasing demand on UK emergency and unscheduled health care services, policy has promoted the streaming of patients presenting to emergency departments (EDs) with non-urgent problems to primary care services. Currently little is known about the different contexts in which primary care streaming is operated, its effectiveness in directing patients to the most appropriate clinician and the impact on flow and waiting times.
Using realist methods we aimed to explore the effectiveness of primary care streaming. We visited 13 hospitals and conducted non-participant observations and semi-structured interviews with staff and patients. Data consisted of field notes and audio-recorded interviews. Transcripts were coded in NVivo v12 and analysed to produce context, mechanism and outcome configurations to help explain how streaming works in different contexts and what mechanisms are perceived to lead to effectiveness outcomes.
Primary care streaming was perceived to help improve flow and waiting times in some settings because GPs saw patients with minor illness and made quick assessments without the need for investigations, enabling ED doctors to focus on more acutely unwell patients. In other settings where there was low demand from patients with primary care problems, there was less perceived improvement on flow. In more integrated services, with a shared governance structure and where there were good working relationships between ED and primary care clinicians, streaming was perceived to be more effective in terms of patients being streamed to the most appropriate clinician, GPs often saw a wider range of patients More experienced nurses working in a streaming or triage role were perceived to be effective in streaming patients to the most appropriate clinician because they had more skills and knowledge to support clinical decision-making. Nurses who had knowledge of primary care and of the skillset of primary care clinicians working in the ED were also perceived as more effective. Some higher band nurses also improved waiting times and flow because they could prescribe, order X-rays, discharge patients or redirect them to their own GP. Clear guidance and training was perceived as essential in streaming patients to primary care clinicians and a lack of guidance and miss-matched governance where the services were provided and managed separately was found less effective. Less experienced nurses with less knowledge, skills and confidence to support streaming decisions were perceived as less effective in streaming patients to the most appropriate clinician and took longer in undertaking streaming assessments, contributing to delays in waiting times and flow.
Effectiveness of streaming relies on experienced nurses, using clear guidance based on the local service, support and training. An NHS Trust managed service, a shared governance structure and a good working relationship between the ED and primary care clinicians improved effectiveness.