Exempt, expedited, full board — or no human participants at all: what sets an Aspen project’s review level
You do not pick your level of review. Aspen says so in the plainest sentence in its handbook: neither the candidate nor the chair nor any committee member holds authority to classify a project, and the call belongs to the IRB Chair or a designee. What you control is the design — and the level falls out of the design the way a weight falls out of a scale. Four features decide it, all four are legible in your own documents today, and none of them cares what the form claims.
Isabel Rourke, PhD · 2026-08-23
In short. The board answers four questions in sequence: is it research, are human participants involved, is it exempt, is it eligible for expedited handling? Anything left goes to the full committee. Aspen reports most files read as exempt, some expedited, very few by the whole board.
Who assigns the level, and when?
The IRB Chair, or someone that Chair designates — after the application arrives, not before. This matters more than it sounds, because it changes what a well-written application is for. It is not an argument for a category. It is a complete and accurate account of what you propose to do, written so that the person applying the test can apply it without having to guess or write back.
Aspen describes the application as designed to gather exactly the information needed to determine the level required. Read that as an instruction: every question on the form exists because some branch of that decision depends on the answer. A vague answer does not buy you a lighter reading; it buys a request for more information, which is a cycle you have paid for with nothing.
What are the four outcomes, in the order the board reaches them?
Aspen has published its own account of the sequence, and it runs as a series of gates rather than a menu.
| Outcome | The question that produces it | Who reads it |
|---|---|---|
| Not research, or no human participants | Is this a systematic investigation aimed past this setting? Is there a living person about whom you obtain data through interaction, or identifiable private information? | An administrator, who issues a letter of determination |
| Exempt | Minimal risk, and does the activity sit inside one of the defined categories — surveys, interviews, observation, educational settings, secondary use of records? | The Chair or a designee, sometimes with a limited review of privacy protections |
| Expedited | Still minimal risk, but outside the exempt categories — does it match one of the listed procedures? | The Chair or a designated reviewer, alone |
| Full board | More than minimal risk, a vulnerable population, sensitive subject matter, or simply minimal-risk work that fits neither list. | A convened meeting of the committee |
The first row is the one people forget exists, and it is worth wanting. A written determination that your activity is not human-subjects research is not a technicality dodged — it is a document that answers, permanently, the question of what authority you collected under. Whether your project belongs in that row is the subject of the piece on quality improvement and research.
What is minimal risk, and what is exempt actually exempt from?
Minimal risk has a definition, and Aspen repeats the federal one: the probability and magnitude of harm or discomfort expected are no greater than what people ordinarily meet in daily life, or during routine physical or psychological examinations. It is a comparison, not a feeling. Interviewing nurses about a workflow clears it easily. Interviewing them about errors they personally made does not clear it as easily, because the harm in play is employment and reputation rather than physical.
And exempt does not mean exempt from oversight. Aspen is careful about this: exempt status releases a project from some federal requirements only. State law still applies. Institutional policy still applies. The obligations of ethical conduct still apply, in full. An exempt project that recruits before its letter arrives is in exactly as much trouble as any other.
What lifts a file from exempt to expedited?
Usually identifiability, occasionally the procedure itself. The exempt categories for interviews, surveys and observation turn on how the record is kept: if responses cannot readily be traced back to a person, or if disclosure could not damage someone’s standing, employment, reputation or expose them to liability, the file sits comfortably. Keep identifiers attached and the same activity needs the board to run a limited review of your privacy and confidentiality provisions before exempt status holds.
The expedited list, meanwhile, is a list of procedures rather than a level of comfort. Three entries carry nearly all nursing work: research using records or specimens that were gathered for treatment or other non-research reasons; the collection of voice, video or image recordings made for the study; and research built on focus group, survey, interview, oral history, program evaluation or quality assurance methods. Being on that list is not a promise — the activity must still be minimal risk in its actual circumstances.
The practical lesson is that identifiability is a design decision you make early and cheaply, or discover late and expensively. Whether you need a name attached to a response, and for how long, is a question worth settling before the instrument is built rather than after the board asks.
What puts a file in front of the whole committee?
Three things, and one of them catches Aspen nursing projects constantly.
- Sensitive subject matter. Aspen names the territory: sexual conduct and history; substance use, licit or otherwise; mental health, including self-harm and compulsive behaviours; and traumatic experience, combat among it. An instrument that asks about any of these is asking for a convened reading.
- A population that needs extra protection. Aspen’s list is long and deliberately so, and it includes groups that ordinary nursing work meets every day: people you supervise or teach, older adults past sixty-five, people who are pregnant, people with a life-threatening illness such as cancer or HIV, residents of a facility, people with cognitive impairment or a debilitating psychiatric condition, veterans, people in crisis, immigrants and refugees, and anyone not fluent in the language the work is conducted in.
- An activity Aspen simply does not permit. The handbook states that Aspen does not allow work requiring a participant to introduce a substance into their body by any route — oral, topical, injected or otherwise. That is not a level of review; it is a wall, and it is better met on paper than after a site has been recruited.
Look again at that middle bullet, because it is the trap. A project surveying the nurses on your own unit involves people you supervise. A project improving discharge teaching on a geriatric ward involves adults past sixty-five. Neither fact ruins a file — but each one has to be seen, named in the application, and answered with the safeguards that make the work sound. Aspen asks you to demonstrate that your procedures are ethically sound, that risk is as low as it can be made, and that the contribution justifies asking these particular people to take part.
What does a convened reading involve?
Members read the application and the proposal in advance, then discuss it at a meeting attended by a majority of the board, which must include at least one member whose main interests lie outside the sciences. Approval takes a majority of those in the room. Aspen notes that the candidate may be invited to answer questions during the meeting — which is a good reason for the file to say what you would say.
Can a design be steered away from a convened reading?
Often, and Aspen openly encourages it. The handbook suggests working with your chair to pivot ideas that would require a convened reading, and gives its own worked example: rather than interviewing children who have lived through trauma, interview the counsellors and teachers who work with them. The question survives; the risk profile changes completely.
That is legitimate design, not evasion, and the distinction is worth keeping clear. Steering means choosing a population or a method that answers your question at lower risk. Concealment means running the same risky design while describing it as something tamer — and that is the file that comes back, because the board reads the instrument, the recruitment text and the consent, and those three documents will always tell on a form that has been softened. If the design genuinely warrants a convened reading, the right move is to write the file so well that the reading is straightforward. The findings boards write most covers what that looks like item by item.
What to do next
Before you fill in a single form, have someone read your design for these four features: what you touch, whom you touch, how identifiable the record is, and how sensitive the subject matter is. That is precisely what the free application review does. Send the proposal and whatever drafts exist to our contact desk; three consultants read it as Aspen’s board would, and you receive a written report giving the level the design earns, the features that earn it, and any finding standing between you and that outcome — each with its evidence, a severity mark and the correction. No cost, no obligation, and if the design already sits where you hoped, the report tells you that and stops. Where it does not, we can carry the rest: the category settled from the design, every document built to match, the submission prepared and filed with you, and each reply answered for as many rounds as the board chooses to write. The work stays yours; the classification is the board’s alone. How it works sets out the stages, and our FAQ answers the fifteen questions that arrive most often.