The reasons packets come back
It is a short list, and it is almost never about the merit of your research. Boards read for internal consistency first, because inconsistency is the easiest thing to spot and the most likely to indicate a real problem.
Read your own packet as a stranger would, in one sitting, and you will find most of these yourself.
- The protocol and the consent form disagree about time, procedure or risk
- Recruitment material promises something the data plan cannot deliver
- Anonymity is claimed while an identifier is quietly being collected
- Consent is written several grade levels above the population reading it
- Data storage, access and destruction are vague or simply missing
- A site permission letter is promised rather than attached
The lane you file in decides your timeline
Three routes exist and they run at wildly different speeds, so aiming at the wrong one is expensive in a way that has nothing to do with the quality of your study. Nobody helpfully moves your submission across; it comes back and you begin again several weeks poorer.
That question gets closed before drafting starts. Occasionally a modest adjustment to the design keeps a minimal-risk study on the quicker route while leaving the science untouched, and now and then the adjustment is something as small as which person distributes a flyer. Learning that up front is worth a great deal more than learning it from a meeting you did not make.
How to answer a revision letter
Receiving one is routine and carries no judgement whatsoever. It arrives as a numbered list, and how you reply determines whether you clear the very next sitting or the one after that.
Take the points in their order, keep their numbers, and for each one say what has changed and where, pasting the new wording in so nobody opens an attachment to find it. The mistake that costs a cycle is defending the original. Someone requesting a plainer statement of risk wants that statement, not a case for why the first version sufficed. Replies leave here within the week, since meeting calendars are indifferent to how busy you are.
Timing it around the calendar
This is where most delay actually comes from. Boards meet on fixed dates, and a packet arriving the day after a submission deadline sits untouched until the following cycle even if it is perfect. Two days can cost a month.
So the work gets planned backwards from your board's dates rather than submitted whenever a draft happens to be finished. Full protocol writing, instrument permissions and the analysis planning that should happen before you collect anything all sit in IRB and research help, and it runs alongside your chapters rather than after them.
Send the packet
Whatever exists, plus your board's template and any letter it has already sent you. Quoted in two hours.
Fix the contradictions
Documents reconciled with each other, category settled, and the whole thing readable in one sitting.
Answer and clear
Revisions handled point by point inside the week, planned around your board's actual meeting dates.
Questions people actually ask
Is a long list of comments a bad sign?
Usually the opposite. A dozen small clarifications means reviewers read carefully and found your design workable. Three comments aimed at your whole approach is the version worth worrying about. Either way the reply keeps their numbering, states what moved and where, quotes the new text, and leaves within the week.
Can I speed up approval?
Through timing far more than through writing. Meetings happen on fixed dates, so landing inside a submission window rather than just past one is the largest single gain available. After that comes the filing route, which is worth settling before a word of the protocol exists rather than after a rejection explains it to you.
What if my study involves patients or minors?
Then you're almost certainly outside the fast lanes and planning has to assume a full board review. That's manageable, it just needs a realistic timeline and a packet that anticipates the questions those populations always raise: additional protections, assent as well as consent, and a genuinely careful account of risk and how it's minimised.
Do you write the protocol or just review it?
Either. Some people send a complete draft and want the contradictions found before submission, which is the cheapest useful order. Others send a topic and a template and want the packet built. Both are normal, and if a review finds the draft is basically sound you'll be told that rather than sold the larger job.
Who does this work?
Doctorate-holding specialists who have sat on review committees themselves, picked for your discipline, and one of them carries you from the earliest draft to the letter that clears you. Continuity is worth more here than in most work, since three separate documents in your packet have to go on agreeing with each other under scrutiny.
