Footprints in Primary Care: Using mixed methods to assess fidelity to a complex intervention for frequently attending patients

Background
At present the NHS is struggling to meet the demands on the service.
The idea for this study originated in a local primary care practice
who felt that improvements could be made regarding how it
was managing and caring for its most frequent attenders. The idea
was developed into a RCGP award-winning intervention consisting of
several components including matching eligible patients with a
named GP, and training GPs to unpack background psychosocial issues
in a contained way during the consultation using the ‘BATHE’
technique.
Aim
The aim of our feasibility study was to explore the main uncertainties
to designing a full trial to evaluate effectiveness and cost-effectiveness
of this intervention in a pilot cluster RCT involving six
practices (4 intervention, 2 usual care). Two of our key objectives
were to optimise the content and delivery of staff training to support
the intervention; and to assess the extent of implementation fidelity.
Methods
As part of the feasibility work, qualitative interviews were held with
the stakeholders to build a clear description of the intervention, how
it was implemented and expected to work. To ensure a high quality
evaluation of implementation fidelity our study design included the
collection a diverse range of implementation data (quantitative and
qualitative) at multiple time points. Observations of implementer
training sessions and of appointment-making between patients and
reception staff, were conducted. A varied sample of face-to-face and
telephone consultations of GPs using the BATHE technique with patients
were video and audio recorded and transcribed in all intervention
practices. Conversation analytic methods were applied to assess
fidelity to BATHE and the nature and extent of patient response in
the consultation recordings Routine monitoring data was collected
from practice records to determine how often study patients were
being matched with their named GPs and patient consultation records
were audited for presence of a BATHE code. All quantitative
data was analysed descriptively to determine intervention dose and
reach.
Results
Analysis of electronic medical records data throughout the 12 month
intervention period enabled us to monitor dose, reach, provide motivational
feedback and document the effects of subsequent implementer
trainings. Observations of appointment-making in all four
intervention practices and video-recordings of all 12 implementer
trainings elicited practical barriers and facilitators that could be addressed,
as well as success stories. Conversation analyses of 20 consultation
recordings enabled a dynamic assessment of the delivery
and receipt of BATHE in situ, that revealed common pitfalls in delivery;
specified and added new dimensions to the underpinning theoretical
assumptions of the intervention; and provided valuable real-world
examples for future training.
Discussion
The findings were used to provide tailored top-up trainings, to clarify
and help address misunderstandings and problems in implementation
and to encourage implementer engagement via whole practice
and individual level feedback. Mixed methods were valuable at different
time-points in enabling a full exploration of what might determine
the success or failure of a future trial; to optimise training and
implementation fidelity; and to understand how and why future participants
might resist or engage with the intervention.