01The two boards are protecting different people
This is the whole of it, and once it is clear the rest of the process stops feeling arbitrary. A university board is responsible for human participants in research conducted under the institution's name, and for the training and supervision of the person doing it. Its questions are about consent, risk, competence and whether a student is properly overseen.
A board at a hospital, clinic, school district or agency is responsible for something else: the people who walk through its doors and the staff who work there. Its questions are about the burden on patients and employees, about who will see the records, about what happens to operations while you are collecting data, and about what the organisation is exposed to if something goes wrong.
Neither is a formality and neither can speak for the other. A study can be perfectly sound as research and still be refused by a site because it would take twenty minutes a shift from nurses who have not got twenty minutes, and that refusal is not a criticism of the science.
02Which means two packets from one study
The study is the same and the emphasis is not, so writing one document and sending it to both is the reliable way to be sent back twice. Take the protocol as your base and prepare each submission for the reader who will actually be holding it.
The university version leads with the design, the consent process, the risks to participants and your supervision arrangements, and it will ask for evidence of your human subjects training. The site version leads with operational reality: who exactly will be approached and by whom, how much of anybody's working time it takes, where the data will physically live, who has access to identifiable records, whether any of the organisation's own systems are touched, and which member of their staff is sponsoring you internally.
That last item is worth naming early. Nearly every site board wants an employee who is accountable for the study inside the organisation, and finding that person is often the longest part of the whole process. Ask before you write anything.
03When one review can cover both
Not every dual situation requires two full reviews. Institutions increasingly agree that one board will act for the other, under a written arrangement usually called a reliance or authorisation agreement, and where that is available it removes a great deal of duplication.
It is worth asking about at the very beginning, because a reliance agreement takes weeks to put in place if one does not already exist, and it is only worth chasing if the answer is likely to be yes. Ask your university office these four questions in a single message:
- Does an agreement already exist between this university and the organisation, and if so, which board takes the lead.
- If none exists, how long does one usually take to arrange, and is it worth pursuing for a student project on this timescale.
- Does the site have a board of its own at all, or does it operate through a research office, a nursing research council or a departmental approval instead.
- What is needed from the site before the university will issue its own approval, and what is needed from the university before the site will consider the study.
04The letters, and the meeting dates behind them
Two documents cause most of the delay in dual review, and they point at each other. Your university generally wants a letter from the site confirming that you have permission to conduct the study there. The site generally wants to see your university's approval before issuing anything. Read as a rule this looks impossible, and in practice it is not, because most university offices will accept a letter of support or intent from the site rather than a full approval. Ask both offices in plain terms what stage of the other's process they need to see, and you will usually find the loop has a door in it.
Then look at how each board meets. Full board review happens on a calendar, sometimes monthly, with a submission cut off well before the meeting. Missing that cut off by two days does not cost two days; it costs until the next meeting. Put both boards' dates on the same page as your own milestones before you write a word, and count backwards from the term your data collection has to happen in.
One thing stays yours throughout. The protocol describes what you will do, and the person who signs the assurances is the person who will be answerable for them. A tender can carry the drafting, the formatting to each board's template and the answers to what comes back, which is what submission work and the wider research plot cover. Choosing the design that both boards will be reading is a separate decision, taken earlier, in qualitative or quantitative.
FAQQuestions to control
Do I need approval from my workplace as well as my university?
If you are collecting data at your workplace, almost certainly yes, in some form. Larger organisations run their own review boards; smaller ones use a research office, a nursing research council or a departmental sign off. Ask the organisation directly what its process is before you write the protocol, since the answer changes what has to be in it and how long it will take.
Which board should I approach first?
Ask both, in one message each, what they need to see from the other. Most universities will accept a letter of support or intent from the site rather than a completed site approval, which is what opens the sequence. Approaching either one blind is how students end up waiting on a document that was never actually required.
Can one board's approval cover both institutions?
Sometimes, through a written reliance or authorisation agreement in which one board acts for the other. Where one already exists between your university and the organisation, it saves a great deal of duplication. Where none exists, arranging one takes weeks, so ask early whether it is realistic for a student project on your timescale or whether two submissions is the faster route.
How long does dual review actually take?
Plan in months rather than weeks, and plan around meeting dates rather than working days. Boards that meet monthly have submission cut offs well before the meeting, so missing one by a day costs a full cycle. Add time for the site sponsor, for any reliance agreement, and for one round of questions from each board, which is normal rather than a sign of trouble.
