Evidence-Based Practice Models in Nursing: Iowa, Johns Hopkins, Stetler, ACE Star, and PARIHS

An evidence-based practice model is a structured sequence for moving research findings into clinical practice, and nursing programmes expect you to name one and show your project moving through it. The models that dominate American nursing are the Iowa Model, the Johns Hopkins Nursing Evidence-Based Practice model, the Stetler model, the ACE Star model, and PARIHS. They are not interchangeable, and choosing on familiarity rather than fit is the most common reason a project proposal is returned.

Why the choice changes the project, not just the paperwork

A model determines what counts as a starting point, who the agent of change is, how evidence gets weighed, and what evidence of success looks like. Pick a model built around organisational decision-making for a project where you are the only person changing anything, and every stage will fit badly.

If you already know which model your programme mandates, the next step is the project itself, and we build evidence-based practice projects under Johns Hopkins or Iowa.

The failure mode faculty see most often is a model named once in the opening chapter and never seen again. If your framework is doing real work, a reader should be able to point at the section where each stage happened. That is the test to apply as you write.

The Iowa Model

The Iowa Model begins with a trigger: either a problem, such as an audit result or a sentinel event, or new knowledge, such as a fresh guideline. It then moves through a sequence of explicit decision points, asking whether the topic is a priority for the organisation, whether there is a sufficient research base, and whether a change is worth piloting before wider adoption.

Its distinguishing feature is those decision points. The model expects you to stop and justify continuing, and it treats piloting as a required step rather than an optional one. That makes it a strong fit for unit-level practice changes where you have organisational backing and an appetite for a small trial before full rollout.

It fits badly when there is no organisational decision to make, or where you cannot pilot because the change is all-or-nothing.

The Johns Hopkins Nursing Evidence-Based Practice model

The Johns Hopkins model runs on three phases, commonly remembered as practice question, evidence, and translation. Its real contribution is the appraisal apparatus: evidence is rated for both level and quality, so a low-level study executed well and a high-level study executed poorly are not treated as equivalent.

That makes it the strongest choice when your evidence base is mixed and you will have to defend how you weighed it. If your topic has two randomised trials, several cohort studies, and a body of expert consensus pulling in different directions, this model gives you a defensible way to reconcile them.

It carries more appraisal overhead than the alternatives, which is wasted effort if your evidence base is small and uniform.

The Stetler model

The Stetler model is deliberately practitioner-oriented. It moves through preparation, validation, comparative evaluation and decision-making, translation and application, and evaluation, and it treats the individual clinician as the primary unit of change rather than the organisation.

It is the honest choice when you are changing your own practice, or the practice of a small team you directly influence, without an organisational mandate. It is also useful when the realistic outcome of your project is a considered decision not to adopt a change, because the model treats rejecting evidence after evaluation as a legitimate result.

The ACE Star model

The ACE Star model frames knowledge transformation as five points: discovery, evidence summary, translation into guidelines, integration into practice, and evaluation. Its value is conceptual clarity about what happens to knowledge as it moves toward the bedside, and it is particularly useful when the contribution of your project is the translation step itself, for instance turning a guideline into a workable protocol for your setting.

It offers less procedural guidance than the Iowa or Johns Hopkins models, so it suits candidates who need a clear conceptual spine rather than a checklist. Where that spine is a nursing theory, or a concept that must be defined before it can be measured, see our nursing theory and concept analysis support.

PARIHS

PARIHS treats successful implementation as a function of three interacting elements: the evidence, the context in which it will land, and the facilitation available to support the change. Its central claim is that strong evidence in a hostile context with no facilitation will fail, and that this is a predictable failure rather than bad luck.

That makes it the right framework when you already know your barrier is organisational rather than informational. If everyone agrees on what the evidence says and the change still is not happening, PARIHS gives you the vocabulary to analyse why and to design facilitation into the project rather than hoping for goodwill.

Choosing between them

Check your programme handbook before deciding. Some departments mandate a specific model, and a well-argued case for a different one still loses to a rubric that names one.

Reporting the work

Models describe how to implement. They do not tell you how to report. If your project is a quality improvement study, SQUIRE 2.0 is the reporting guideline, and it asks for things a model will not prompt you to record, including the local problem, the rationale linking intervention to expected outcome, and details of how the intervention changed during the project. Building to it from the start is far cheaper than retrofitting the document later.

Where the change is tested through Plan-Do-Study-Act cycles, design each cycle to produce a measurement you can report, rather than one long implementation period with a before and after.

Applying this to your own project

If you are selecting a model for a doctoral scholarly project, the choice interacts with your question format and your evaluation design, and we work through all three together on our doctoral project build service. If you are still narrowing a practice problem, our doctoral project ideas guide covers what makes one feasible, and writing a PICOT question covers turning it into something searchable. For a capstone at bachelor's or master's level, where a model is often required but implementation is not, see capstone project support.

Where the model meets the evidence

Whichever model you choose, it will send you to the literature, and the quality of that search decides how defensible the rest of the project is. Nursing evidence sits partly outside the databases a general search covers, which is why the search has to be built deliberately rather than typed into one box. If your programme asks you to appraise sources one at a time first, our guide to writing an annotated bibliography shows the three-part structure. Our systematic literature search service builds and documents reproducible searches, and where the project needs a formal appraisal of study quality, risk of bias and critical appraisal covers the tools programmes expect.