Nursing AI: Barriers, Literacy, and Decision Authority

Two newly published nursing measurement studies raise a useful question for DNP project planning: What, exactly, would an AI-related score tell you? A November 2 study examines educators’ barriers to adoption. An October 30 paper evaluates an Arabic digital and AI literacy questionnaire. An earlier Oncology Nursing Society position statement adds a different question about professional authority.

These sources address different decisions. A barrier survey can inform planning; a literacy questionnaire can identify perceived learning needs; a professional position can inform governance discussions. Combining them into a single label such as “AI ready” would hide those distinctions.

The applications below are DNP Blueprint editorial suggestions. They are not validated checklists, institutional authorization, or evidence that an AI tool improves patient outcomes.

1. New research: Identify barriers before choosing an intervention

What happened

On November 2, Scientific Reports shared an accepted, peer-reviewed early version of the BAI-NE development study. The 15-item instrument was evaluated with a convenience sample of 470 nurse educators from seven nursing faculties. It addresses technical, training, ethical, and organizational barriers. The authors describe the evidence as preliminary and specify interpretation by subscale rather than a combined total. The early version remains subject to editing. This is measurement research, not a trial of an AI education program. Read the publisher’s article.

Why it matters for a DNP project

Consider a hypothetical faculty-development project. If the team starts by scheduling a workshop, it has already assumed that knowledge is the main obstacle. A needs assessment could instead reveal questions requiring a different response, such as uncertainty about an institutional process.

Define who will use the assessment and what decision it will support before collecting responses. A faculty leader planning education and a clinical manager considering a care workflow may need different evidence. An instrument’s title alone cannot establish suitability for either decision.

A useful proposal can explain why the chosen construct fits the project’s question and who can act on the result. That explanation gives a survey a purpose beyond producing a baseline number.

What to measure or check

Write a short decision statement: “We need to understand this issue among these respondents so this role can decide the next step.”

Then check the instrument’s intended population, language, version, scoring instructions, and conditions of use. A study of educators does not automatically support interpretation among staff nurses or patients. Do not invent a readiness cutoff or combine dimensions merely to simplify reporting.

For a small exploratory assessment, document invitations, participation, incomplete responses, and the agreed action following review. Report the decision the assessment informed. Collecting scores does not show that an intervention worked.

2. New research: Keep perceived literacy separate from demonstrated performance

What happened

JMIR Nursing published the Arabic NURS-DIAL evaluation on October 30. It involved 196 nursing students at one Saudi university and a final 52-item, five-factor questionnaire. The authors limit current use to exploratory research and formative assessment in similar settings; validated cutoff scores are unavailable. Objective competence comparisons and test-retest reliability were not assessed. Only 106 participants completed all 52 items, making incomplete responses relevant to interpretation. The paper offers preliminary measurement evidence, not certification of safe AI use. Read the original study.

Why it matters for a DNP project

In an educational project, first decide whether you want to understand learners’ perceptions, their priorities for teaching, or their ability to complete a task. Those are different outcomes.

A hypothetical curriculum pilot could use a formative needs assessment to help select content. If it also aims to evaluate performance, the project would need a separate, clearly defined assessment. Faculty might review an invented teaching scenario and agree what an adequate response would contain before learners attempt it.

Keep the interpretation proportionate. A positive reaction to a lesson is useful feedback, but it does not establish that learners can transfer a skill to an unfamiliar situation. A project can report both findings without presenting them as interchangeable.

What to measure or check

Record the exact questionnaire version and language. Check what each dimension represents and which direction its response options run. Ask the instrument’s coordinating team about use or adaptation when needed.

Plan missing-response handling before analysis and report completion with a denominator. Do not silently replace blanks, change scoring, or create a new overall score. Describe any departures from the published approach.

If you add a locally developed task, identify it as such. Document faculty review, scoring decisions, and feasibility limitations. A small educational pilot should not claim to validate a complex instrument simply because its participants completed a survey.

3. Earlier guidance: Make nursing authority practical

What happened

The ONS AI position statement was approved in September 2026. It is earlier background guidance, not a November announcement. It calls for oncology nurses to participate in AI governance and retain the practical ability to challenge and escalate unsafe recommendations. It also emphasizes education, adequate time, and organizational processes. This is a professional position, not experimental evidence of clinical benefit. Read the official statement.

Why it matters for a DNP project

A discussion of staff knowledge can miss a more concrete question: Does a person who identifies a problem know what happens next?

With an authorized faculty advisor or site mentor, a project could examine how an existing escalation process is described. Use a fictional tabletop scenario and the site’s current policies. Ask participants to explain whom they would contact, what information they would provide, and how they would learn that a concern had been addressed.

This exercise evaluates a process description. It does not authorize a new technology or test an AI system on patients. Findings from a specialty position also need interpretation within the project’s own setting and scope.

What to measure or check

For each scenario, record whether participants identify an accountable role, a usable contact route, and a way to confirm follow-up. Retain disagreements for review instead of counting every completed form as a successful escalation.

Distinguish a missing instruction from a lack of training or a problem with access. These observations may suggest different improvement actions.

If the process changes, explain what was revised and how it was reviewed. Improved agreement in a tabletop exercise is a limited process finding. Patient safety or clinical effectiveness would require different evidence.

Choose one decision for this week

Bring one question to your advisor or site team: Are we assessing an adoption barrier, exploring perceived learning needs, observing a defined performance, or reviewing a governance process?

Choose the smallest useful deliverable and state its intended interpretation. Keep the source date, population, version, and limitations alongside any measure you select. A well-bounded conclusion is more useful than a broad readiness claim that the project cannot support.

Sources

  1. Mohamed, Mohamud, and Sorour: Development and psychometric validation of the barriers to AI integration in nursing education scale (BAI-NE): a multi-institutional study (Journal)
  2. Alnomasy, Alrashedi, and Pangket: Preliminary Psychometric Evaluation of the Arabic Version of the NURS-DIAL Questionnaire Among Nursing Students in Saudi Arabia: Cross-Sectional Study (Journal)
  3. Oncology Nursing Society: Artificial Intelligence in Oncology Nursing. Board approval: September 2026. (Institution)

All AI Watch posts