A Realist Evaluation of the Clinical- and Cost-effectiveness of Paramedics Working in General Practice
Problem
General Practice (GP) services are responding to demand by increasing the non-medical workforce supporting service delivery. Paramedics are increasingly working in GP, as their generalist skillset for undifferentiated problems may be well-suited. Paramedics carry out a range of tasks including home visits, routine and same-day appointments, and telephone triage. There is significant variation in the types of patients that paramedics manage, their models of working and their contractual engagement. To date, no research investigates the variation in paramedic models in GP and the associated impacts on patient safety, clinical- or cost-effectiveness. Implementation guidance struggles to reflect this variation, making it difficult to make contextually-informed decisions around how to successfully implement the paramedic role.
Approach
We describe a mixed-methods, Realist Evaluation of different models of paramedics in GP, to understand how they: achieve good clinical outcomes for patients; provide safe care; improve patient experience; relieve GP workload; influence the workload of other staff; make efficient use of healthcare resources.34 case study sites (practices with or without paramedics) were recruited across England using a sampling frame to ensure variation in site demographics (size, geography, urbanity, deprivation). Informed by Initial programme theories (IPTs) developed in previous work, sites were classified into models based on key domains of variation such as level of paramedic integration with the GP team and types of task undertaken.Semi-structured realist interviews were conducted with patient participants (or their adult carers), paramedics, GPs, and other general practice staff, to elicit information about how paramedic and non-paramedic models work, for whom and under which circumstances. Interview data was focussed on understanding the mechanisms through which the intervention, in a given context, results in intended and unintended outcomes.Quantitative data were collected from a sample of adult patients via questionnaires, at both paramedic and control sites. Domains captured assessed patient reported experience, safety and outcomes of the consultation. Additionally, routine clinical data were extracted from general practice systems including: re-consultations, investigations, medications and referrals during a 30 day (care episode) after the initial consultation (index visit).
Findings
At the time of abstract submission, the study has closed to recruitment. Qualitative, quantitative and health economic data analysis is underway. Complete baseline questionnaire data has been collected from 722 participants, 69 interviews have been conducted and routine data has been extracted for over 18,000 patients across paramedic and non-paramedic sites. Qualitative and quantitative data triangulation will be completed to present a comprehensive set of programme theories to support evaluation of the impact of different models of paramedic care on individual and system wide costs and outcomes.
Consequences
The final programme theories (constructed as statements) will be used to inform the initiation and implementation of paramedics in general practice according to variation in local need and circumstances.