NR-702B Week 5 Specifying the Practice Change: How to Write It

What NR-702B Week 5 asks for

What separates a specified intervention from a described one? A specified intervention can be handed to a colleague who was not in any of the meetings. Take an occupational health clinic adding structured hypertension follow-up after workplace screening events. The described version says that employees with elevated readings will receive enhanced follow-up. The specified version says that any employee with two readings above a defined threshold at a screening event is entered on a callback list the same day by the screening nurse, contacted within five working days by the clinic nurse using a two-question template, and offered an appointment slot held for that purpose on two afternoons a week, with the callback logged in the existing clinic record. The second version can be delivered, audited and improved. The first cannot.

The second demand is that adaptation be visible and reasoned. Almost no intervention transfers unchanged, and the doctoral skill is deciding what must be preserved and what may bend. If the evidence suggests that follow-up within a week is what carries the effect, that element is core and your local version must protect it. If the studies used a pharmacist and your site has none, delivery by another role is an adaptation that must be named as yours, with a sentence about the risk it introduces. Silent adaptation is the failure mode: a project that quietly drops the active ingredient and then reports that the evidence-based intervention did not work.

Third, the deeper hour load in this variant makes specification genuinely testable before implementation. There is room to walk the proposed workflow with the people who would run it, to find out that the callback list has nowhere to live in the record system, and to redesign before anything is written into a final plan. Writing that reflects that kind of checking reads very differently from writing produced at a desk, and readers can tell.

The boundary that governs every page in this manual. Clinical hours, hour logs, encounter counts, preceptor evaluations, site documentation and signatures belong to the student and are never drafted, reconstructed or estimated with help. No tutor contacts a site, performs clinical work, or documents clinical activity on a student's behalf. Written support covers only the scholarly layer: making a specification complete, checking that adaptations are justified, and holding doctoral register. Describe roles rather than named individuals, and de-identify every patient detail that reaches a page.

The NR-702B Week 5 method, step by step

Seven moves for specifying a change well enough that somebody else could run it.

1. List the components the evidence actually contained

   From your matrix's delivery column, not from the intervention's name. Most interventions are bundles, and knowing the components is what lets you decide which ones you are keeping.

2. Mark each component as core or flexible

   Core components are the ones the evidence associates with effect. Flexible components are delivery details that varied across studies without apparent consequence. State the basis for each classification, because this is a judgment a reader will test.

3. Write the workflow as a sequence with actors

   Trigger, actor, action, tool, record, and what happens when the step is missed. Six lines usually covers it. If a step has no actor, it will not happen, and if it has no record, it cannot be measured.

4. Name every adaptation and justify it separately

   Different role, lower frequency, shorter script, paper instead of an electronic field. For each, say what forced the change and what risk it introduces to the core mechanism you identified.

5. Cost the change in the site's own units

   Seconds per encounter, callbacks per week, one build request, thirty minutes of training per staff member. A specification with no cost estimate cannot be approved by anyone who has to fund the time.

6. Define fidelity before implementation begins

   What proportion of eligible cases would need to receive the step for the change to count as delivered. Choosing this in advance is what allows a later evaluation to distinguish a weak intervention from a weak rollout.

7. Walk it with someone who would run it, then revise

   The person who owns the workflow will find the failure in ninety seconds: the field does not exist, the afternoon slot is already committed, the trigger happens when nobody is in the room. Revise the specification and say in the writing that it was checked.

A layout and word budget for an intervention specification

Our frame for this deliverable, sized for roughly 1,300 to 1,600 words. It is our own outline rather than anything the university publishes, and your week's rubric outranks it wherever they disagree. If a workflow diagram is included, the sequence section shortens and the justification sections stay as they are.

Evidence craft for specification writing

Use a reporting framework for intervention description and name it. Conventions exist for describing what an intervention is, who delivers it, how, where, how often and how much, and following one produces a complete specification without your having to invent the checklist. Cite the framework you used.

Tie every core component to a source. The claim that a component drives effect is an evidence claim and needs a citation, or an explicit statement that it is your inference from the pattern in the matrix. This is the sentence a committee is most likely to probe, so write it deliberately.

Quantify the intervention as well as the outcome. Dose matters: how many contacts, how long, over what period. Studies frequently differ in effect because they differ in dose, and a specification that omits dose cannot be compared with the literature it came from.

Write exception handling explicitly. What happens when the eligible patient is missed, when the role is absent, when the system is down. Real workflows fail regularly, and a plan that only describes the successful path will be delivered inconsistently and then evaluated as though it had not been.

Describe approvals without predicting them. Where a change requires a policy update, a standing order, a template build or a review determination, say what will be submitted and to whom. Do not assert the outcome of a review or imply that an approval is a formality, since those decisions belong to the bodies that make them.

Five mistakes that cost points in this week's territory

• An intervention named rather than specified. Structured follow-up is a label; a trigger, an actor, a tool and a record is an intervention.
• Silent adaptation. Changing who delivers the change or how often, without flagging it, breaks the link between your plan and the evidence you cited for it.
• No dose. A specification with no frequency or duration cannot be compared with the studies that support it or delivered consistently by anyone else.
• Only the happy path. Workflows without exception handling get delivered inconsistently, and the inconsistency is then misread as an ineffective intervention.
• Costs described in adjectives. Minimal added time is not a unit, and no manager can weigh it against the other demands on the same ninety seconds.

Before you submit

• Components are separated into core and flexible with reasons attached
• Every core component is tied to a source or marked as your inference
• The workflow names a trigger, an actor, a tool and a record for each step
• Dose is specified as frequency and duration
• Each adaptation states the constraint that forced it and the risk it creates
• Costs appear in minutes, builds, materials and training counts
• A fidelity threshold is defined before implementation
• Exception handling is written for the steps most likely to fail

Questions students ask about this stage

Keep going