Turning Delphi findings into an implementation plan
A final list of agreed recommendations is a useful research output. Planning delivery also requires knowing what blocks an action, who can address it, and how a proposed solution will be tested.
Feasibility findings can explain where delivery needs support
An October 2026 secondary analysis of an Australian fear-of-cancer-recurrence pathway Delphi found feasibility ratings ranging from 49.4% to 90.1%. Researchers analyzed comments using the updated Consolidated Framework for Implementation Research and mapped determinants to strategies from the Expert Recommendations for Implementing Change taxonomy. Read the study.
The example shows how quantitative ratings and qualitative explanations can inform implementation planning. The workflow below is Calibrum’s practical extension for study teams. It does not mean a strategy has already been tested or that consensus establishes effectiveness in practice.
Ask what makes an action difficult—and under which conditions
Keep the recommendation identifier, wording, rating dimension, setting, and time horizon together. Ask about feasibility separately from importance or expected benefit, then invite comments on the conditions needed for delivery.
Original example prompts
“What would prevent this recommendation from being implemented in your setting within 12 months?”
“Which resources, approvals, training, or changes to the proposed action would be needed?”
“Are there settings or groups for which this action would be difficult to deliver?”
A comment about staffing and a comment about an unclear responsibility require different responses. Retain their links to the original item and context. Tell participants how comments will be used and avoid including identifiable details in published examples.
Review comments before turning themes into recommendations
Agree whether the project needs formal qualitative analysis or a practical descriptive summary. For formal coding, record the framework or codebook, reviewers, treatment of ambiguous comments, and how differences were resolved. An informal summary should be labeled accordingly.
Keep barriers, enablers, proposed solutions, and unresolved questions distinct. Preserve important minority concerns even when few people mention them. A comment count is not automatically the number or proportion of panelists holding that view: participants may submit several comments or leave none.
Review interpretations against the source material. If AI assists with an agreed analysis task, retain human review, document its role, and ensure the data use fits the project’s privacy requirements. An automatically generated theme should not become a final research conclusion without review.
Connect each proposed action to the finding that prompted it
Build an action record that retains the original Delphi result alongside the team’s proposed response. The following examples illustrate the structure; they are not results from the cited study.
| Finding | Proposed action | Check before adoption |
|---|---|---|
| R12: respondents describe limited training time. | Pilot shorter training modules with protected time. | Responsible lead, staffing approval, completion rate, and staff feedback. |
| R18: responsibility for a handoff is unclear. | Define an owner and test a documented handoff. | Local authority, workflow fit, missed handoffs, and workload. |
| R23: access differs across locations. | Explore an alternative delivery route with affected groups. | Resources, accessibility, uptake, and effects across locations. |
For every action, record the source item and round, rating profile, reviewed theme, proposed response, owner, dependencies, intended timing, evaluation measure, and decision status. Separate participant suggestions from actions developed later by the research or implementation team.
Preserve what the panel actually agreed
An action that becomes feasible after new funding, narrower eligibility, or a different delivery model has changed its conditions. Preserve the original recommendation and findings rather than silently replacing them with the revised version.
Record who approved any adaptation, why it was made, and whether further review or panel rating is required under the protocol. A steering-group decision after the Delphi should be attributed to that group. It is not a new panel consensus unless the panel has assessed it under an appropriate process.
Treat the plan as a proposal to evaluate
Agree what a pilot needs to establish: acceptability, adoption, delivery quality, resource use, reach, or another relevant outcome. Specify the setting, review date, and who will decide whether to continue, adapt, or stop. Local feasibility can change as staffing and resources change.
Surveylet can collect separate ratings and item-linked comments and provide exports for analysis. Define the qualitative review, action mapping, ownership decisions, and evaluation plan as part of the project’s analysis and implementation work.
Discuss a tailored implementation section through Delphi Analysis & Final Report. Agree the data, expertise, deliverables, and scope before assuming a formal framework analysis or action plan is included.
Source and scope
Designing for delivery: a Delphi study of feasibility-informed implementation strategies for a fear of cancer recurrence clinical pathway. Supportive Care in Cancer; October 9, 2026.
The paper provides a worked research example of framework-based analysis. Calibrum’s prompts and action-record examples are original planning suggestions and should be adapted to the study’s setting and approved methods.
Plan how the findings will be used
Define the questions, comments, and reporting outputs needed to help your team move from an agreed recommendation to a reviewable implementation proposal.