Intervention Design Process

What is an intervention design process?

An intervention design process is a structured way to develop an action, program, or service intended to change an outcome. For a behavioral intervention, it includes specifying the behavior and population, investigating relevant barriers, and deciding what to deliver and evaluate.

Stages of intervention design

The 2021 MRC–NIHR framework distinguishes development or identification, feasibility, evaluation, and implementation. Work can revisit phases rather than follow a fixed linear sequence. Context, program theory, stakeholder perspectives, uncertainty, refinement, and resource implications remain relevant throughout.

How to design an intervention: from idea to test

Consider a service with missed appointments. The team should investigate whether timing, transport, cost, or information is the problem before selecting reminders. It should specify what counts as success and how it will distinguish the intervention’s effect from other changes. A reminder cannot supply a ride or pay the fare.

Working through the appointment example

1. Specify the behavior: Adults booked for a first appointment attend the correct location by the scheduled start time.
2. Investigate the experience: Ask attendees, people who missed appointments, and staff what happens between booking and arrival. Compare their accounts with service records.
3. Form a barrier hypothesis: Suppose the evidence suggests that some people misunderstand the location. That is a testable explanation, distinct from assuming everyone forgets.
4. Design a response: Try a booking message that gives the correct entrance and a way to ask for directions. Explain how this should address the identified barrier.
5. Check feasibility: In a small pilot, check delivery, comprehension, accessibility, and the work needed from staff. Refine the message if people still cannot find the entrance.
6. Evaluate outcomes: Compare attendance and timely arrival with an appropriate comparison group over a defined period. Also record cancellations, rescheduling, and staff workload so a change in one measure is not mistaken for overall success.

This is a hypothetical application of the framework. If interviews instead identify unaffordable transport, the team should revisit the intervention rather than continue refining directions.

Testing feasibility and effectiveness

A plausible explanation and a workable prototype are different from evidence of effectiveness. Evaluation must address the outcome and setting of interest. An intervention can be feasible without producing the intended effect, and an observed effect may not transfer unchanged to another population.