01The packet has two audiences
A board reads for risk: who could be harmed, how, and what stands between them and it. Your protocol, your recruitment wording and your instruments are all read with that one question in hand.
The consent document has an entirely different reader, and it is the only academic writing most people ever do for somebody with no reason to be interested. That reader has to manage five things after a single pass.
- Say what the study will ask of them, counted in minutes and in visits.
- Say what could go wrong for them, in ordinary words rather than in categories.
- Say what happens to whatever they tell you, and who besides you will ever see it.
- Say how to stop, at any point, without giving a reason and without losing anything.
- Find somebody to contact who is not you, because a board insists on an independent name.
02So it gets drafted first, not last
Most people write the consent form at the end, as an appendix, out of the protocol. That order produces a document explaining the study to a colleague and confusing everybody else, which is where a large share of returns begin.
Written first it does something more useful. Explaining a procedure to a stranger in plain sentences exposes every place the design is still vague, and a step you cannot describe simply is usually a step you have not finished deciding. The protocol is then written to match a document that already reads clearly, which is also how the two stay in agreement.
03What holds a packet up, and none of it is science
Boards rarely return work on the merits of the idea. They return it because something underneath is missing, or because two documents in the same envelope say different things.
The usual causes are dull and entirely avoidable. A training certificate that lapsed. A site letter naming a department that has since been renamed. An instrument used without written permission from whoever owns it. Recruitment wording offering a payment the protocol never mentions. A consent form that lost a paragraph during reformatting. Everything is drafted on your board's current templates, checked against the others, and filed together. Which queue a design falls into, and why that outranks everything else, is covered under IRB and research.
04Answering what comes back, and the boundary
Almost nothing clears on a first pass, and what arrives is a set of instructions rather than a verdict. Each numbered point gets its own numbered reply, the documents are amended to match every answer given, and the changes are marked so nobody has to hunt for them. A long list usually means somebody read closely.
What stays yours is everything involving a real person. Nobody here consents a participant, speaks to one, or collects anything at all. A protocol is written to describe your study rather than to dress it up, and where data has already been gathered without approval you are told plainly that no paperwork repairs that. The chapters built on top of the packet run through dissertation work.
Send the design and the forms
What you intend to do, to whom, and where, plus your board's current templates if you have them. A tender who has served on a panel answers inside two hours.
Start at the consent document
It is drafted first, in language a participant can read once and act on. Anything vague in the design surfaces here, while changing it is still cheap.
File once, then answer
Protocol, recruitment, instruments and appendices built to agree with it, submitted together, every revision point answered in the board's own numbering.
FAQQuestions to control
What reading level should a consent form be written at?
Lower than feels natural to somebody writing a doctorate. Many boards name a target in their guidance, and where yours does not, plain sentences and everyday words are the safe default. Technical terms are explained the first time they appear rather than assumed. A form your participant understands on one reading is also the form least likely to come back.
How long does a submission usually take to clear?
It depends on the route your design falls into and on your board's own calendar. Work using records that already exist can clear quickly at many institutions. Anything going to a full panel waits on that panel's next sitting, and the clock covers the days before the meeting as well as those after it. You are given a realistic figure for your own board first.
My university and my hospital both want a submission. What changes?
Two packets, built together rather than one after the other, since each board has its own forms and its own view of who may be approached. Approval from one is not approval from the other, and a delay at either stops the study. Which goes in first is decided at the start, because some boards will not open a file until the other has ruled.
How much of the packet can you actually write?
Every document: the protocol, the consent and assent forms, recruitment wording, the data handling plan, and whichever appendices your board asks for. What a tender will not do is decide the design for you. Those choices come back as options with what each one costs in weeks attached, and you pick.
Can you help after a submission has already been rejected?
Yes, and plenty of packets arrive here that way. The first job is reading the letter against what you actually sent, because a rejection often turns on one structural problem repeated in several places rather than on the number of points listed. Where the objection is to the design itself, you are told what changing it would cost in weeks before anything gets rewritten.
