NR-707A Week 4 The Deviation and Decision Record: How to Write It

What NR-707A Week 4 asks for

A documentation review at the halfway point of one implementation found that a screening step designed to run at admission was being completed at the first medication pass instead, roughly ninety minutes later, in almost every case. Nobody had decided this. The admission window was already congested and the medication pass was the first moment a nurse was reliably at the workstation. Two responses were available: enforce the original timing, or move the designed step to where the work actually fits and adjust the protocol. Both are defensible. Neither is defensible without a written record of what was observed, what was considered, and what governed the choice.

The boundary again, because this stage tempts people to blur it. Practicum hours, hour logs, encounter counts, site forms and preceptor or mentor evaluations are your own record of your own work, never drafted, reconstructed or estimated with assistance. Support extends only to the written and analytic layer: how a deviation is described, how the decision reasoning is structured, how the section is organized for a doctoral reader. Anything drawn from real encounters is de-identified before it is written, and deviations are described in terms of process and role rather than named individuals.

The concept worth getting right in your prose is the difference between fidelity to form and fidelity to function. Some elements of an evidence-based change are the active ingredient and must be preserved; others are packaging that can and should be adapted to a setting. A doctoral submission that treats every departure from the written protocol as a failure has misunderstood implementation science, and one that treats every departure as a legitimate adaptation has abandoned the evidence. Say which core elements are non-negotiable, and why, before you evaluate any specific deviation.

Expect a section reporting mid-implementation status, sometimes framed around a plan-do-study-act cycle or an equivalent structure from your project's chosen framework, often with the decisions taken listed explicitly. Write it while the reasoning is fresh. Decisions reconstructed at the end of a block always read as though the author knew then what they only learned later.

The NR-707A Week 4 method, step by step

Six moves for writing deviations and the decisions they force.

1. List the core elements of the change before you judge anything

   Name the components that carry the evidence and must be preserved, and the components that are local packaging. This list is the standard against which every deviation is classified, and writing it once serves the rest of the block.

2. Describe each deviation as observed behavior

   What was designed, what is happening instead, in how many of how many cases, since when, and on which shifts or units. Deviations written as complaints cannot be analyzed; deviations written as observations can.

3. Find the reason from the people doing the work

   Ask the staff performing the step why the alternative is easier or safer from where they stand. Most deviations are rational responses to a constraint the designer did not see, and the constraint is the finding.

4. Classify: drift, adaptation, or signal of a design fault

   Drift erodes an active ingredient and needs correcting. Adaptation preserves function while changing form and can be adopted. A design fault means the protocol asked for something the setting cannot deliver, and that is the most valuable of the three.

5. Write the decision with its reasoning and its date

   What was decided, by whom, on what evidence available at that moment, and what was explicitly not chosen. Recording the rejected option is what makes the record analytic rather than declarative.

6. State what the decision does to your measurement

   If the protocol changed mid-block, your measures may now be counting something slightly different, and any chart covering both periods needs the change annotated. Saying this yourself pre-empts the sharpest question a committee can ask.

A layout that turns deviations into a defensible decision record

Our frame for a mid-implementation deviation and decision section, sized for roughly 1,300 to 1,600 words. It is our own outline rather than anything the university issues, and your week's rubric outranks it wherever the two disagree.

Evidence craft for mid-implementation writing

Ground core elements in the evidence, not in the proposal. Saying a component is essential because your protocol says so is circular. Cite the source that establishes it as an active ingredient, with a year, and where the evidence is silent about a component, say that too, because silence is itself a licence to adapt.

Write decisions in the past tense with an actor. The project team agreed on the fifteenth to move the screening step, rather than it was decided to move the step. Doctoral readers are assessing your judgment, and the passive voice removes the person whose judgment is being assessed.

Preserve the information horizon. Write what was known at the time of the decision and mark later knowledge as later. A record in which every decision looks prescient is a record that has been edited into uselessness.

Use your framework's vocabulary consistently. If your project runs on improvement cycles, name the cycle, its aim, its prediction and its result. If it runs on an implementation framework, use its constructs for the barriers you found. Mixing vocabularies from three models is the most visible sign of borrowed structure.

Keep deviations de-identified and process-level. A shift, a role, a unit and a count. Never an individual whose practice could be recognized by a colleague reading the document, and never a patient case described closely enough to be located.

Five mistakes that cost points in this week's territory

• Reporting no deviations. Four weeks into a live change in a real setting, an account with nothing to correct reads as unobserved rather than perfect.
• Adaptation without a core-element test. Changing the protocol because it was inconvenient, with no argument about the active ingredient, abandons the evidence base.
• Decisions with no owner or date. An unattributed decision cannot be evaluated, and an undated one cannot be aligned with your data.
• Retrospective reasoning. Writing a choice as though later information was available at the time destroys the credibility of the whole record.
• Silent protocol change. Altering what is delivered without annotating the measures leaves a discontinuity in your data that a committee will find.

Before you submit

• Core elements are declared with an evidence source for each
• Every deviation carries counts, dates and the units or shifts involved
• Each cause is traced to a constraint described by the people performing the work
• Each deviation is classified as drift, adaptation or design fault, with reasoning
• Every decision names an actor, a date, the evidence used and the rejected alternative
• Measurement consequences and chart annotations are stated explicitly

Questions students ask about this stage

Keep going