A Doctor of Nursing Practice project is the culminating requirement of a practice doctorate: a study in which you identify a problem in a real clinical setting, synthesise the evidence around it, implement a change, and measure whether it worked. It is assessed on whether the translation into practice was methodologically sound, not on whether it produced new knowledge. That single criterion separates it from a dissertation and determines how the entire project must be built.
A research doctorate produces a dissertation. Its job is to generate knowledge that did not exist before, and it is judged on originality and methodological contribution.
A practice doctorate produces a scholarly project. Its job is to take knowledge that already exists and make it work somewhere specific, and it is judged on whether the implementation was faithful and the evaluation honest.
Three consequences follow, and candidates who miss them lose months.
A null result is not a failure. If you implemented a screening protocol faithfully and referral rates did not move, that is a complete project. The requirement was sound translation and honest measurement, not a positive finding. Candidates who treat the project as a dissertation chase significance they were never asked to produce.
Feasibility is a methodological requirement. In research, an ambitious design is a virtue. Here, an outcome you cannot measure with data your site will release, inside your timeline, is a design flaw. The realistic project beats the impressive one.
Scale is not the marker of quality. A tightly executed change on one unit with a clean measurement outranks a hospital-wide initiative you could not control or evaluate.
Programmes vary, but nearly all require the same structural elements.
A defined practice problem located in a specific setting, with evidence that it is actually a problem there. "Hand hygiene compliance is a national concern" is background. "Compliance on this unit audited at 58% against a target of 90%" is a practice problem.
A question in PICOT-D format. Population, intervention, comparison, outcome, and time, with the D marking the practice-change orientation. The question has to name what will change and what will be measured.
A named implementation framework, visibly used. The Iowa Model, the Johns Hopkins model, Stetler, ACE Star, and PARIHS each suit different situations, and we compare them in detail in our guide to evidence-based practice models.
A synthesis of the evidence, searched across CINAHL, PubMed, and the Cochrane Library, appraised with CASP or JBI, and summarised into evidence tables with levels of evidence assigned. The search has to be documented so a reader could repeat it.
An implementation plan, usually tested through Plan-Do-Study-Act cycles, with each cycle producing a measurement rather than one undifferentiated period of change.
An evaluation with a defined primary outcome, stated measurement points, and an analysis plan written before data collection rather than after.
How long it takes and where the time actually goes
Most programmes allocate between two and four semesters, and candidates consistently misallocate that time. Writing is rarely the bottleneck.
The time goes to institutional approvals, which can take longer than the implementation itself; to negotiating data access with a site that has no obligation to prioritise your project; and to the evidence synthesis, which is far more work than a coursework literature review because it must be reproducible.
The failure pattern is predictable. Months spent refining a topic that was never narrow enough, followed by a rush through the evidence chapter, followed by the discovery at implementation that the outcome data does not exist in retrievable form.
Most projects are quality improvement rather than research, which usually means a determination letter confirming that classification rather than full review. But the boundary is genuinely blurred, and the answer depends on your institution, not on a general rule.
The practical guidance is to establish the classification early and in writing, because it affects your timeline substantially, and because a project designed as quality improvement that later gets classified as research may need approvals you did not budget time for. Your institution makes the determination; your job is to ask early and design so the classification is defensible either way.
Strong topics share a shape. They target one practice, at one point of care, in a defined population, with an outcome already being recorded by the site for some other reason.
That last point does more work than any other. If the data already exists in the electronic record, in an audit, or in a mandatory report, your evaluation is feasible. If it requires a new collection process that clinical staff must sustain during a project you will not be present for, it usually is not.
Our guide to doctoral project ideas works through examples against exactly this test. If you are still shaping the question, writing a PICOT question covers the format, and the PICO builder will structure one free.
Where the project is a quality improvement study, SQUIRE 2.0 is the reporting guideline, and programmes increasingly name it. It asks for elements a framework will not prompt you to record, notably the rationale linking your intervention to the expected outcome and an account of how the intervention changed as you ran it. Recording those as you go is far cheaper than reconstructing them at write-up.
If you want the project built end to end with the framework, evidence synthesis, and evaluation designed together, that is what our doctoral scholarly project service does. Where the requirement is a capstone at bachelor's or master's level rather than a doctoral project, see capstone project support.
