Study title

Evaluating effectiveness, safety, patient experience and system implications of different models of using GPs in or alongside Emergency Departments. (GPs in EDs)

Principal / Lead Investigator


Type of study

Realist evaluation using mixed methods

Who is the study sponsor?

Cardiff University


Hospital emergency departments (EDs) are under increasing, sometimes extreme, pressure. Patients are experiencing long hours of waiting to be seen. We have seen media reports of emergency services in crisis. This situation has arisen partly because people attend EDs with problems that GPs could deal with. Finding better ways to assess and treat patients coming to EDs could have a major impact on the experience and care of the millions of people attending EDs and on all NHS services by providing evidence of how best to manage resources.

Many EDs now work closely with GPs in three main ways: 

  • Assessing patients and sending them to a GP clinic next door to the ED;
  • Placing GPs in the ED to make a brief assessment of patients and decide what type of healthcare provider they need to see
  • Placing GPs in the ED as extra staff to provide treatment.           

It is not known how well these arrangements provide care or if they make a difference to how quickly patients receive treatment and referrals for tests and other services. Our study will look at what works, for whom and in what circumstances. 

In this study we will look at: 

  1. What are the different ways of working, how are they are organised and how common are they?
  2. How do they work in practice? Do they do what they were meant to do?
  3. Hospital admissions, re-attendance, waiting times, patient safety and staff and patient satisfaction
  4. What are the key factors that lead to a better service? 
Public / patient involvement

Two public collaborators recruited from the Involving People network were co-applicants for the project funding.  They attended and contributed to all meetings, read and commented on drafts and prepared the lay summary. During the study they will have a key role in planning the stakeholder events and supporting public attendees. 

How could this research potentially benefit patients?

Currently, the effectiveness of each GP-ED model is unclear in terms of its impact on service provision, patient experience and the effectiveness and safety of clinical care delivery. This study plans to address that research gap.

Does the study involve a methodological innovation?

The mixed methods design will use ‘realist evaluation’ to examine the effectiveness of the models in terms of what works, for whom and in what circumstances. It will be conducted in 3 main Phases. 

Phase 1: We will undertake a national Survey to characterise which GP-ED models are in use and to what extent in England and Wales, and the aims of the services they provide. To complement this, we will undertake a rapid realist (literature) review of the contexts in which and mechanisms by which different models achieve their outcomes. We will then generate initial working theories on the GP-ED models taxonomy, to explore and clarify in a Stakeholder conference. The Stakeholder conference will also confirm a set of ‘marker conditions’ (presenting conditions thought to be managed differently by GPs and ED clinicians i.e. investigation and admission rates) for subsequent in-depth analysis, including the effectiveness and safety of patient care provided, in Phase 2. 

Phase 2: Using a mixed methods design, twelve study sites will be purposively sampled. Quantitative data analysis will consist of Interrupted Time Series analysis of NHS England / NWIS (Wales) and Emergency Care Dataset analysis, plus description of case mix and outcomes, and economic / resource use evaluation. This will be integrated with qualitative data collection (‘marker conditions’, field observations, key informant interviews, incident report analysis) using techniques such as pattern matching for a mixed methods synthesis. Three sites will be selected for each of the three main model types (GPs working geographically adjacent to EDs; GPs serving in a triage and screening capacity; and GPs fully integrated into ED service provision) and three that do not use GPs in or close to the service. 

Phase 3: Segmented regression analysis of quantitative interrupted time series data, economic modelling of resource and outcome data, combined with field observation and interview data will inform the realist evaluation to compare how the different contexts and mechanisms lead to the intended and unintended outcomes across the different GP-ED models and identify what works for whom and in what circumstances. Findings will be provided to Survey participants and a second Stakeholder conference for feedback on credibility, transferability, and discussion. 


NIHR NETSCC grant income  NIHR HS&DR                

Total grant value

TBC (subject to cost amendments)

Start date

To be confirmed

End date 

To be confirmed

Further info (e.g. related web link)

To follow 

Outputs generated (Reports / Publications / Impact)

To follow