Workstream 2: feasibility study

Aim

Assess the feasibility of intervention delivery and the proposed RCT methods to inform refinement.

Pilot intervention delivery and assess data quality/availability

A summary follows, with detail elsewhere.24

A single-arm, mixed-methods feasibility study (six practices; two East Midlands CCGs; May 2017–January 2018) tested two patient recruitment approaches (postal invitations; GP prompts) and the feasibility of intervention delivery and data collection methods.

Three participant groups were involved: patients, practice staff and SSME provider staff. Quantitative data were collected from primary care electronic medical records (extracted data; 2877 patients) and self-completed patient questionnaires (self-reported data; 423 participants). Health economics data were collected from providers and a questionnaire and interview with practice managers and CCG staff. An integrated ethnographic study included observations, interviews and document analysis.

Key findings were:

• HbA1c (RCT primary outcome) was over 90% complete in extracted data.
• More patient participants completed the questionnaire via postal invite (85%) than GP prompt (15%).
• Ninety-one per cent of patient participants consented to linkage of self-reported and extracted data.
• Ethnographic and interview methods were appropriate.
• Intervention costing was possible based on the resource use data collected.

It was therefore concluded that the RCT was feasible with design improvements, including the following:

• There was limited engagement and capacity constraints from most feasibility practices and staff, so it was planned that the RCT intervention implementation would focus on providers, while maintaining relationships with practices and CCGs. However, while the RCT included providers, much of the embedding activity remained at practice level because the primary outcome was extracted from practices, so the focus needed to be on study practices. This may not have happened with a provider-level focus, as providers worked across study and non-study practices. Furthermore, randomisation was at practice level, not provider level. With more time between the feasibility study and RCT, the RCT could have been redesigned to facilitate this change of intervention focus.
• The sample size calculation was updated based on feasibility study recruitment figures.
• SSME information and preferences and biomedical data were removed from the self-report questionnaire due to poor reporting.
• A preassigned participant number was added to the patient questionnaire to reduce the quantity of missing NHS numbers for data linkage.
• CCG staff and practice managers were removed as sources for resource use data because this proved unfeasible. This came from the embedder and provider staff in the RCT.
• The ethnographic study identified challenges around the intervention and feasibility study, including organisational factors (e.g. safety and burden of data extraction, and confusion among stakeholders in duality of project manager and embedder role), intervention features (e.g. uncertainty on how to assess the toolkit and misunderstandings around language used to describe the intervention) and characteristics of local research teams (e.g. logistical and capacity issues within the qualitative team and duality of project manager and embedder role meant that study issues detracted from intervention activities).24 Learnings from these challenges were taken into improving the design of the intervention and the RCT.
• Not all findings could be readily incorporated into the domains included in NPT, so we adjusted those constructs (see Appendix 1, Table 5; see also Davies et al.24).

Refine intervention

See Modifications.

Assess progress of research programme

The following pre-agreed stop/go criteria were met:

1. Completion of evidence synthesis (WS1) and feasibility study (WS2).
2. Tailored Embedding Package for practices and CCGs tested and ready for the trial.
3. Improvements made in the recording of structured education uptake to the satisfactory level for analysis within the RCT.
4. Stakeholder and PPI groups established.
5. Core team for phase 2 WS3 and WS4 appointed.
6. Trial protocol completed and ethical approval gained.
7. Study submitted for adoption on the Clinical Research Network portfolio.
8. Interim statement on intellectual property.

Changes from workstream 2 original plan

There were no major changes. The feasibility study started late (see Challenges), which shortened the planned gap before the RCT started.