03 · What You Need to Know
Disagreement and noncompliance are different things
A researcher can disagree with a reviewer and still participate properly in ethics review.
OHRP's current guidance for U.S. IRBs recommends written procedures covering how required modifications or clarifications are communicated, how investigators' responses are subsequently reviewed, and how reasons for disapproval and opportunities for investigators to respond are handled. For covered research, the review process therefore explicitly anticipates communication between investigators and the IRB rather than treating every committee comment as the end of discussion.
Other ethics systems may organize this exchange differently. The general distinction remains important.
Reasoned disagreement
You explain your concern through the authorized review process and ask the committee to consider clarification, evidence, or an alternative.
Noncompliance
You conduct research in a way that is not authorized because you personally decided the committee's requirement should not apply.
The first can be part of responsible scholarly exchange. The second can create a serious research-governance problem.
Before disagreeing, make sure you understood the comment correctly
Some apparent disagreements disappear once the comment is unpacked.
"Please address how participant distress will be managed" may be a request for additional explanation rather than a demand to remove the sensitive questions. "Please revise the recruitment process to address possible coercion" may allow several solutions rather than prescribing one.
Read the decision letter carefully. Separate:
- questions seeking clarification;
- requests for additional information;
- suggested wording;
- specific modifications required for approval;
- conditions attached to approval;
- reasons the committee cannot yet approve the study.
If the wording remains unclear, use the committee's permitted communication route to ask what concern the requested change is intended to resolve.
Ask what ethical criterion is at issue
A productive disagreement begins with the committee's concern rather than your preferred solution.
Is the issue risk minimization? Voluntariness? Privacy? Confidentiality? Fair participant selection? Adequacy of consent? Scientific justification? Additional protection for a particular population?
Once the concern is clear, you can determine whether you actually disagree with the ethical objective or only with the particular method proposed to achieve it.
Often it is the latter.
You can agree with the concern and disagree with the solution
This is one of the most useful distinctions in ethics correspondence.
You might agree that direct recruitment by a supervisor could create perceived pressure but disagree that the entire employee population must therefore be excluded. You might agree that identifiable longitudinal data create confidentiality risk but disagree that complete anonymity is feasible because the design requires repeated observations to be linked.
In those situations, state both parts clearly:
Concern
"We agree that the proposed recruitment relationship could create perceived pressure to participate."
Problem with the requested change
"Excluding this population would prevent the study from addressing the research question for which it was designed."
Alternative
"We therefore propose independent recruitment and a procedure preventing supervisors from knowing who accepts or declines the invitation."
Request
"We respectfully ask whether this alternative would adequately address the committee's concern."
The exact wording will vary, but the reasoning is stronger than simply writing "We disagree."
Use evidence when the disagreement depends on facts
If you believe the requested change rests on an incorrect factual assumption, provide evidence relevant to that assumption.
This might include official guidance, methodological literature, validation documentation for an instrument, evidence about the frequency or severity of a procedure's risks, technical information about data security, or a concise explanation of the study design.
Do not overwhelm the committee with citations simply to make the response look authoritative. Evidence should resolve the disputed point.
Explain methodological objections as methodological consequences
If your objection is that the requested change would weaken the research, specify how.
Would it invalidate an instrument? Prevent measurement of the primary outcome? Introduce systematic selection bias? Destroy a longitudinal linkage? Make groups incomparable? Change the intervention being tested?
The deeper problem of responding when a requested ethics change may harm the methodology deserves this level of precision because "we prefer our original method" is rarely enough.
Do not treat disagreement as a negotiation over participant protection
Researchers sometimes approach ethics comments as though the objective is to persuade reviewers to remove inconvenient requirements.
A better question is whether the committee's ethical concern can be addressed adequately in another way.
If the requested change is necessary to satisfy an applicable ethical or regulatory requirement and no acceptable alternative exists, methodological preference does not automatically override that requirement. The study may need to change, be reconsidered more fundamentally, or not proceed in its proposed form.
The committee should have reasons for significant adverse decisions
Within the U.S. HHS and FDA regulatory context, if an IRB disapproves research, written notification must include the reasons for the decision and provide the investigator an opportunity to respond in person or in writing. OHRP guidance also states that IRB meeting minutes must document the basis for requiring changes or disapproving research.
This is useful as an example of a review system that recognizes the importance of reasons and investigator response. Do not assume that the same procedural entitlement exists in identical form in every jurisdiction. Check the rules governing your committee.
A response to comments is not necessarily an appeal
If a committee asks you to revise the protocol and you explain why you prefer an alternative, you are usually still participating in the ordinary review process.
An appeal generally refers to a more formal challenge to a decision under procedures established by the institution or ethics system. Whether such a procedure exists, which decisions can be appealed, who considers the appeal, and what outcomes are possible vary considerably.
Do not escalate an ordinary clarification into a formal dispute unnecessarily. Conversely, if ordinary dialogue has concluded and the committee has issued a final decision, determine whether a formal appeal process is available.
Approval with conditions does not mean every issue remains negotiable after you start
Under OHRP guidance, an IRB can approve covered research with conditions in certain circumstances when it can make the required approval determinations assuming specified conditions are satisfied. It can also require modifications and defer the study for further review when the outstanding issues prevent approval.
The distinction matters because the researcher's freedom to proceed depends on the actual decision. Do not assume that "conditional" means "start now and discuss the remaining disagreement later."
Before beginning research, determine what the committee's conditional decision actually permits.
You cannot solve disagreement by making an unapproved change yourself
Suppose the committee approves your proposed alternative, but after recruitment begins you decide the original procedure would work better after all.
That is no longer the same disagreement. It is a proposed change to approved research.
For research subject to the U.S. Common Rule, proposed changes generally require prior IRB review and approval before implementation except where necessary to eliminate apparent immediate hazards. OHRP's compliance materials document failures to obtain prior review of protocol changes as regulatory problems.
Watch Out
The fact that you previously argued for a procedure does not authorize you to use it. What matters operationally is the version of the research currently approved under the applicable review process.
Professional tone matters because the objective is resolution
A response can be firm without becoming adversarial.
Avoid statements such as "The committee clearly misunderstands qualitative research" or "This requirement makes no scientific sense." Even when frustration is understandable, those formulations do not explain the methodological issue or help reviewers resolve it.
State what you understand the concern to be, what evidence or reasoning leads you to a different conclusion, and what action you propose.
The ethics committee does not need ceremonial agreement. It needs a response it can evaluate.
04 · A Practical Example
Disagree with the proposed safeguard without dismissing the concern
Hypothetical Example
Interviewing nurses about workplace safety
A researcher proposes confidential interviews with nurses about workplace safety incidents. The ethics committee is concerned that participants could identify colleagues or managers and requests that all questions about specific incidents be removed.
The researcher believes that removing discussion of actual incidents would make it impossible to answer the research question, which concerns how safety problems are recognized and handled in practice.
Unhelpful response
"We disagree. Specific incidents are essential to qualitative research, so we will retain the questions."
Identify the concern
The committee appears concerned about identifiable third-party information and possible occupational consequences if interview material is disclosed.
Explain the disagreement
The researcher explains that concrete accounts of safety incidents are necessary to examine the processes central to the research question and that removing them would change the phenomenon being studied.
Propose safeguards
The researcher proposes instructing participants not to use names where possible, removing or generalizing identifying details during transcription and reporting, restricting raw-data access, and avoiding quotations whose contextual details create an unacceptable identification risk.
Request reconsideration
The researcher asks whether these protections would address the committee's concern while allowing the incident-focused questions to remain.
The researcher has disagreed with the requested change, but not with the need to protect confidentiality. That distinction gives the committee something substantive to reconsider.
