Implementing evidence means putting a supported practice into everyday use and evaluating how that use works in the intended setting. It requires more than distributing an article or completing an education session. The National Cancer Institute describes implementation science as the study of methods that promote adoption and integration of evidence-based practices into routine care and public health.
A local implementation project is not automatically implementation research. Its purpose and methods determine the appropriate classification and review; see quality improvement versus research.
Specify the practice and the problem
Write down what should change, who should receive or deliver it, and what evidence supports the change. Separate essential components from details that may need local adaptation. Identify the outcome the practice is intended to improve and the process that shows whether it is actually being delivered.
Before choosing a training session as the solution, investigate the obstacle. Staff may lack information, but they may also lack time, equipment, a clear responsibility, or a workable documentation route. Different obstacles suggest different actions.
An illustrative implementation plan
Imagine a unit has approved a structured handoff approach. The team could observe the current handoff process, identify where interruptions occur, and agree on how the new approach fits each shift. It could arrange practice opportunities, make the agreed prompts accessible, and define who will answer questions during early use.
The evaluation could distinguish whether handoffs follow the agreed structure, whether critical information is available, and whether the change creates delays or duplication. This invented example does not claim that a particular handoff format improves outcomes.
If the process is not used as intended, a disappointing result cannot be interpreted exactly as a test of the fully delivered approach. If it is used consistently, that still does not prove the desired clinical benefit occurred.
Include people who do and receive the work
Ask staff and affected patients or caregivers about practical fit, communication needs, and foreseeable burdens. Implementation involves organizations and communities as well as individual behavior. A plan developed only around one shift's workflow may miss problems encountered elsewhere.
Document adaptations so that later reviewers know what was delivered. An adapted version may be sensible, but it should not be described as identical to the version studied.
Monitor and sustain
Choose a manageable set of measures with clear definitions, denominators, and review dates. Include the intended outcome, actual delivery, and important unintended consequences. State who can act on the findings and what would trigger revision or a pause.
Plan beyond launch: responsibility for materials, onboarding new staff, maintaining supplies, reviewing evidence updates, and checking whether the practice remains useful. A one-time completion count cannot show sustained use.
For the earlier decision about fit, see research generalizability.
For a practical way to begin on one unit, see small tests of change.
Sources
- National Cancer Institute: About Implementation Science (accessed 2026-09-17).
- AHRQ Methods Guide: Assessing the Applicability of Studies When Comparing Medical Interventions (accessed 2026-09-17).
- AHRQ: The Evidence for Evidence-Based Practice Implementation (2008) (accessed 2026-09-17).