Quick answer
A SNF data project can improve local care and also constitute research. Document its purpose, interventions, participants, identifiability, funding and institutional commitments before selecting the review route. Publication intent alone does not decide the question. Obtain the appropriate institutional determination and separately resolve privacy permissions, vendor use and any consent requirements.
A vendor may propose to learn from resident data while helping the facility improve a workflow. An academic partner may add random assignment or future model development to a local improvement project. Those additions can change the oversight question even when the original project remains useful to the SNF.
OHRP explains that some quality-improvement activities also serve a research purpose and that intent to publish is not decisive. Its FAQ predates the revised Common Rule and explicitly warns that older citations must be reconciled with current requirements. OHRP quality-improvement FAQs.
Describe the project before choosing its label
Prepare a short protocol containing the operational problem, planned intervention, comparison groups, assignment method, data fields, identifiable recipients, future uses, funding and resident contact. Identify who can change the protocol and who controls any exported data.
The current definition of research focuses on a systematic investigation designed to develop or contribute to generalizable knowledge. Human-subject involvement includes specified interactions or interventions with living individuals and uses of identifiable private information. These are separate questions. 45 CFR 46.102(e) and (l).
Route three fact patterns
| Synthetic project | Facts that matter | Determination to request |
|---|---|---|
| Local review of completion of an established documentation process | Routine practice; existing records; purpose limited to local operations | Whether this is nonresearch improvement under the actual protocol |
| Vendor compares two new prompts by assigning units to different workflows | Prospective assignment; evaluation of generalizable effects; resident information | Research/human-subject assessment and applicable review before intervention |
| University analyzes existing identifiable records | Secondary use; recipient access; funding; analysis purpose | Applicability and any specific exemption or other review route |
These are routing examples, not final classifications. A local project may acquire a new research component; a research project may qualify for an exemption under precise conditions. Keep the determination and the version of the protocol it covers together.
Common Rule coverage depends on federal conduct, support or regulatory applicability, with institutional commitments and other requirements also relevant. Lack of a federal grant does not by itself resolve every oversight obligation. 45 CFR 46.101.
Current exemption categories have conditions, including specific provisions for secondary research. Obtain the determination through the institution's authorized process; do not assume that “retrospective” or “exempt” means unrestricted. 45 CFR 46.104.
Keep privacy and research decisions separate
A research determination does not itself authorize disclosure of PHI. The privacy team must determine the permitted data use, applicable authorization or other pathway, recipient responsibilities and any special record restrictions. Conversely, a privacy arrangement does not establish that human-subject review is unnecessary.
Specify whether a supplier may use data solely to operate the contracted service, evaluate this project, train a general model or retain it after the work ends. Those are materially different purposes. Avoid a contract that describes one purpose while the technical export supports another.
Make approval boundaries executable
Before launch, create a decision record listing the protocol version, authorized reviewer, determination, conditions, approved fields, recipients and change triggers. A new intervention, additional facility, identifiable export or expanded reuse should be checked against that record before implementation.
Test the platform with an approved field list. If the project needs ten fields and an export includes 40, identify and remove the unnecessary fields before release rather than assuming the broad data set is covered. Verify that project access can end without disrupting the underlying clinical record.
The buyer's output is an authorized project boundary with reproducible data controls. Ease's skilled nursing offering is planned; participation in any future data pilot would require its own scope, review and agreements.
Questions & answers
Does publishing a SNF improvement project automatically require IRB approval?
No. OHRP says publication intent alone does not determine whether an activity is research. Its actual purpose and design, human-subject involvement, applicability and exemption conditions must be assessed.
Can a vendor label a resident-data pilot as quality improvement to avoid research review?
The label does not settle the classification. Ask for the protocol, planned uses, interventions, data recipients and funding, then obtain the appropriate institutional determination before starting the dependent activities.
Primary sources
- OHRP Quality Improvement FAQs
- 45 CFR 46.102: Current research and human-subject definitions
- 45 CFR 46.101: Applicability
- 45 CFR 46.104: Exempt research
Use the cited source and its applicable version when making a facility-specific decision. These guides explain operational evaluation methods; they are not individualized clinical, coding, or legal advice.