Does your GCU DPI project need IRB review, or a QI determination?
Most Direct Practice Improvement projects at Grand Canyon University are routed as quality improvement rather than as human subjects research — but that routing is not something a candidate declares and moves on from. GCU decides it through a form signed with your chair and a survey embedded in the application portal, and a QI analyst can send the whole submission down the other path. Knowing which answers open which door is the difference between one filing and three.
The short answer
It depends on what the QI Survey inside GCU's iRIS portal returns. Complete the QI Self Certification Form with your chair first; the portal then opens either a quality-improvement path or a full research application, and a QI analyst confirms which one fits.
What does GCU actually ask a DPI candidate to file?
GCU builds the determination into the submission system rather than issuing a separate letter. Its guidance is specific on three points. First, projects deemed to be quality improvement do not require an IRB application, review or IRB designation — a QI survey exists inside the GCU iRIS application so applicants can obtain approval on that path. Second, all doctoral applicants are expected to complete the GCU QI Self Certification Form with their chair before starting an iRIS application; the candidate and the chair both sign it, and it is uploaded with the application under additional documents. Third, GCU notes that Doctor of Nursing Practice applications typically fall under the QI category because of their required DPI projects.
Then it branches. Where the QI Survey answers indicate quality improvement, the QI application questions appear automatically; where they indicate research, the system presents an IRB application instead. That is not advisory — it changes which form you are filling in.
And the branch is reviewable. GCU states that a QI analyst decides whether an application falls under human subjects research or is non-research quality improvement, and may request revisions to move a submission from one to the other. So the survey answers are a claim; the analyst's reading is the determination.
What makes an activity "research" in the first place?
Federal regulation gives one test, at 45 CFR 46.102(l), and it repays slow reading. Research is a systematic investigation — one that takes in development, testing and evaluation — whose design aims at knowledge meant to hold beyond the case at hand. Two conditions, both required. Systematic alone is not enough; a competent improvement cycle is systematic by construction. The converting element is the second condition: an intent that the findings speak past the setting they came from.
The same section names four things deemed not research at all — scholarly and journalistic work, public health surveillance under a public health authority, criminal justice collection and analysis, and certain authorized national security operations. None ordinarily reaches a DPI project, but they demonstrate that the regulation draws hard edges rather than pulling everything methodical inside.
What does OHRP say about quality improvement specifically?
The Office for Human Research Protections takes this one head-on. Asked whether the human subjects regulations reach activities that go no further than putting a practice in place to improve care quality, plus gathering patient or provider data on how that rollout went for clinical, practical or administrative use, OHRP answers no. Such work misses the definition of research, so those regulations impose neither IRB review nor research consent on it.
OHRP's own examples are recognisably DPI-shaped. A clinic adopting a widely accepted assessment inside routine care, then auditing charts to see whether the right patients are getting it. Affiliated hospitals adopting a procedure known to cut prescription errors, then pulling chart data to check adherence and whether errors fell as expected. Each time: a known practice goes in, and data come back to show whether the change took locally.
One caveat belongs on the record. OHRP issued that guidance before the revised Common Rule took effect, so its citations point at the older paragraph numbers, and OHRP flags the document as non-binding recommendation rather than regulation. The reasoning holds; the numbering moved.
Which signals point which way?
| Signal | Points toward quality improvement | Points toward research |
|---|---|---|
| Purpose | Improve care in this setting | Produce knowledge meant to hold elsewhere |
| The intervention | An established practice or guideline already accepted | An untested change, or one being compared against alternatives |
| Assignment | Everyone in scope receives the change | Participants allocated to arms or conditions for study purposes |
| The data | Routinely collected operational or clinical data | Data gathered from people specifically for the project |
| Identifiers | Aggregate or de-identified before analysis | Identifiable private information retained and analysed |
| Population | Staff and patients already in the workflow | Recruited participants, or anyone the regulations protect specially |
A single row does not settle a project. GCU's QI Survey and the QI analyst's reading do; this table is how to anticipate them.
Even if it is research, are there human subjects?
A second gate sits behind the first, and it often settles a chart-based project. The regulation's human subject, defined at 45 CFR 46.102(e)(1), is a living person whom a researcher, doing research, either engages through intervention or interaction and then draws usable information from — or about whom that researcher handles identifiable private information. And identifiable, at 46.102(e)(5), means an identity the investigator can readily work out, or one already travelling attached to the record.
OHRP reads it the same way: analysis of data the investigators cannot readily link back to people, where nobody was interacted with to obtain it, involves no human subject. Which is why a project built on retrospective operational data can be methodical, ambitious about generalizing, and still fall outside the regulations — not for being unimportant, but because there is no identifiable person on the far side of it.
Whether that reasoning fits your file is exactly what the analyst or the board is being asked to confirm. It is not a finding a candidate reaches privately and then acts on.
Does wanting to publish turn a DPI project into research?
Not by itself, and this is the most persistent misreading in practice-improvement work. Publication is a route to dissemination; the regulatory test is about design intent. Work built to fix a local problem does not become research because someone later writes it up, and work built to produce generalizable findings does not escape the definition because nobody submits it anywhere.
GCU applies the same logic elsewhere, which is a useful tell for how the university thinks. In its guidance on classroom-based and archival data, GCU excludes material gathered solely for instructional improvement or reflective teaching from site authorization and IRB approval — yet states that where there is an intention to use it for research, including analysis for publication or conference presentation, IRB approval must be obtained. Intent at design time is the hinge in both settings.
So answer the QI Survey about the project you are actually designing, and say plainly if dissemination is part of the plan. Routing that rests on answers which do not describe the work is routing that gets undone later.
What happens if the analyst disagrees with your answers?
You revise inside the same system rather than starting a new submission. GCU's published route is to open a submission response, read the reviewer's comments on the specific questions flagged, choose to revise the existing submission, change the answers identified — and then continue through the whole form, because changing a routing answer opens sections that were not there before. Converting a submission from a QI application to an IRB application, or the reverse, means answering a materially different set of questions, not editing a few fields.
That is worth planning for. If a project sits near the line, the sensible preparation is to have the exhibits an IRB application would need drafted alongside the QI path, so a re-route costs a filing rather than a rebuild. What those exhibits are is set out in the exhibit file a GCU IRB application carries.
What should you have in hand before you open iRIS?
- The QI Self Certification Form, worked through with your chair and signed by both of you, ready to upload.
- A project description that matches your approved proposal exactly — routing answers and proposal must describe one piece of work. Why that holds at every desk: AQR locks the proposal before the GCU IRB reads a word.
- CITI completion reports, registered under your GCU email so the records reach the submission packet unaided.
- Site authorization wherever the project touches people, records or space, on campus or off — the two are separate GCU processes.
- A clear account of identifiers: collected, recorded, linked, destroyed — worded identically everywhere it appears.
- A dissemination intention you can say out loud, because the analyst reads the project you described, not the one you meant.
What to do next
Where a DPI project sits near the line, the move worth making is to have it read against both tests — the federal definition and GCU's own survey logic — before you answer anything in the portal. That reading costs nothing here, and it ends in a written finding: the route the project fits, the evidence behind that, and the exhibits the file needs either way.
After that, the whole GCU sequence can move to this desk: the determination and document plan, each exhibit drafted against your approved proposal, the filing in GCU's current portal, and each answer the board or the analyst asks for until the letter reaches you. The project remains yours throughout. The determination remains GCU's.
Sources
- GCU Technical Support — iRIS QI Survey and Revision Process — QI needs no IRB application; the form signed with the chair; automatic branching; the analyst's determination; revision in both directions
- GCU Office of Research — Institutional Review Board — CITI requirement; forms and templates; DNP contact route
- GCU — Steps to Prepare IRB Application — site authorization on and off campus; chair review; iRIS
- GCU — Classroom-Based Research and Archival Data — instructional-improvement data excluded; research or publication intent requires IRB approval; data use agreement
- 45 CFR 46.102 — what counts as research, what is carved out of it, who is a human subject, and when information is identifiable
- OHRP — Quality Improvement Activities FAQs — local practice rollout with operational data is not research; unlinkable analysis involves no human subject; predates the 2018 Requirements, non-binding
gcuirb.com is an independent consulting practice. There is no affiliation with Grand Canyon University and no endorsement by it, and nothing here is clinical, legal or regulatory advice. Where this page diverges from GCU's current doctoral handbook or IRB portal, GCU's text governs.