NIH Human Subjects Research Plan: What the Protection of Human Subjects Section Needs to Say

Why Human Subjects Review Trips Up More Researchers Than Expected

For an R01 investigator focused on the science, the human subjects section can feel like a compliance checkbox: fill in the IRB information, say you'll protect participants, move on. That's not how reviewers experience it.

Under NIH's Simplified Review Framework, the content of the human subjects section flows into the Approach score — Factor 2 — because reviewers are assessing whether your study design is feasible, whether risks are adequately addressed, and whether your inclusion plan is defensible. A weak human subjects plan can pull an otherwise strong Approach section from a 2 to a 3. That costs you points in ways that feel arbitrary when you're reading your summary statement and the scientific comments were mostly positive.

The other complication is structural. The human subjects documents are split across multiple fields and attachments in ASSIST, and they carry different requirements depending on whether your research meets NIH's definition of a clinical trial. Many investigators get that classification wrong — either over-complicating the application unnecessarily or, worse, under-reporting and triggering a Just-In-Time request to clarify research status before an award can go out. Neither outcome is one you want at that stage of the process.

What the Human Subjects Section Actually Contains

The "human subjects" umbrella in an NIH grant application is not a single document. It is several, and which ones you need depends on how you answer the screening questions in ASSIST's Human Subjects and Clinical Trials (HSCT) Information form.

Depending on your research designation, you'll be routed to some or all of these components:

Protection of Human Subjects attachment: your plan for participant protection, risk minimization, IRB status, and data and safety monitoring where applicable
Inclusion of Women and Minorities table: your justification for the composition of your study population by race/ethnicity and sex
Inclusion Across the Lifespan attachment or justification: whether children and older adults will be included or excluded and why
Clinical Trial Specific Information: additional questions required if your research meets NIH's four-part clinical trial definition

If your research doesn't involve human subjects — if you're working with fully de-identified datasets or animal models only — you select that designation at the top of the HSCT form and most of these sections disappear. But the designation matters. Reviewers see it, and if your research sounds like it involves human participants but you've marked it otherwise, that discrepancy generates questions and sometimes a request for clarification before the award can issue.

Writing the Protection of Human Subjects Plan

The core protection attachment follows a recognizable structure. Even when the FOA doesn't specify headings, reviewers expect to find answers to the same four questions in this order.

Risks to subjects

Name the risks concretely. "Minimal risk" alone isn't enough if the research involves blood draws, sensitive survey questions, or identifiable data. For online surveys, the primary risk is often breach of confidentiality rather than physical harm — write that explicitly. Reviewers flag vague risk statements, especially for studies involving clinical populations or sensitive topics.

Adequacy of protection against risks

Describe how you minimize each risk you named. Include your consent process, de-identification procedures, data security measures, and any specialized monitoring. If you have a Data and Safety Monitoring Board, name it here. If your study is low-risk enough not to require one, say so and explain why. The absence of a safety monitoring plan for higher-risk intervention studies is one of the most consistent notes in summary statements across study sections.

Potential benefits and importance of knowledge to be gained

State concretely what participants stand to gain — direct benefit, or none for purely observational research — and what knowledge the field gains if the study succeeds. This paragraph is often one sentence long in early drafts. It should be two or three. Reviewers use it to assess whether you've thought seriously about the risk-benefit ratio, not just the scientific rationale. A two-sentence paragraph here signals intentionality.

IRB review status

You don't need IRB approval in hand at the time of application. But you need to disclose the current review status — pending, approved, or exempt — and the name of the reviewing institution. For multi-site studies, note that each site will obtain its own IRB approval or that a central IRB arrangement is in place. Reviewers notice when this field is left blank or vague, and program officers will follow up before award.

Inclusion Requirements: Women, Minorities, and Age Across the Lifespan

NIH has required inclusion of women and minorities in research involving human subjects since 1994. The Inclusion Across the Lifespan policy broadened the framework further in 2019 to require justification for excluding any age group — not just children. These aren't suggestions. Missing or inadequate inclusion documentation is grounds for an administrative review hold before your application even reaches study section.

The Inclusion of Women and Minorities table asks you to list planned enrollment by race/ethnicity and sex, broken down by study arm. For a new application, these are estimates. For a renewal, they reflect actual enrollment to date plus remaining projected enrollment. What matters to reviewers is whether your enrollment reflects the population that stands to benefit from this research. A cardiovascular study proposing a predominantly male sample without explanation will draw a comment in 2026. That has been true for years and has only gotten more consistent.

The justification for exclusions needs to be specific. "Not applicable" is not a justification. If you're excluding children because the intervention is not age-appropriate for that group, write that in one sentence. If you're enrolling a single ethnic group because of a genetic question specific to that population, write that too. Reviewers are looking for evidence that you've thought about representativeness — not that your population is diverse for its own sake.

How Sex as a Biological Variable Intersects with Human Subjects Review

NIH's policy requiring consideration of sex as a biological variable applies to both vertebrate animal research and human subjects research. For human subjects work, the policy operates alongside the Inclusion of Women and Minorities requirement — but they address different questions.

The inclusion table is about who gets enrolled. The sex-as-biological-variable requirement is about whether sex is considered as an analytical variable in your primary analyses. For studies enrolling both men and women, reviewers expect to see a sentence somewhere — usually in the Approach section or integrated into the human subjects plan — that addresses whether primary outcomes will be analyzed separately by sex, whether the study is powered to detect sex differences, and why.

If your study isn't powered to detect sex differences, say so honestly. "This study is not powered to detect sex-stratified differences; exploratory analyses by sex will be conducted but interpreted cautiously" is a far stronger response than silence. Silence forces reviewers to wonder whether you forgot the requirement — and that is exactly the wrong question to be circulating during your score discussion.

The Exempt Research Trap

The NIH definition of exempt human subjects research tracks the Common Rule categories, and it is narrower than many investigators assume. De-identified secondary data from a private company is not automatically exempt. Survey research involving sensitive topics — substance use, mental health, immigration status — rarely meets the criteria for exemption even when no direct identifiers are collected. Educational tests have specific criteria that apply only in particular instructional contexts.

If you mark your research as exempt, you need to identify the specific Common Rule category and explain briefly why your research fits it. Reviewers do spot-check this. An application marked exempt under a category that clearly doesn't apply signals either a misunderstanding of the regulations or an attempt to sidestep oversight. Neither impression helps your Approach score — and a program officer may flag the designation before your application reaches study section.

For genuinely borderline cases, the safer path is to mark the research as involving human subjects, describe your IRB review process, and note that approval is pending. That is not a weakness. Reviewers understand that approval comes after submission. It is honest, and it is the correct answer for the review process.

Common Mistakes That Generate Reviewer Concerns

Marking "no human subjects" when data came from people

Even if your dataset is de-identified, if it contains identifiable information or was collected without appropriate waivers, the designation may be wrong. When in doubt, talk to your IRB before submission — not after a Just-In-Time request arrives asking you to clarify.

No safety monitoring plan when the study design requires one

Any study with greater-than-minimal risk interventions is expected to describe how safety will be monitored. Missing this generates a reviewer note almost automatically. A sentence saying you'll form a monitoring committee — or that independent safety review will be provided through a specific mechanism — is usually sufficient for lower-risk interventional studies.

Underestimating the inclusion justification

A single sentence saying "women and minorities will be included" doesn't satisfy the requirement. Reviewers need enrollment numbers, a justification for those numbers, and acknowledgment of any groups being excluded. The inclusion table alone isn't enough — the text needs to explain the reasoning behind it.

Letting the human subjects section contradict the Approach

Reviewers read both together. If the Approach describes 200 participants with blood draws and the human subjects attachment lists minimal risk with no mention of venipuncture, that inconsistency draws a comment. Keep the two sections consistent and cross-reference them where the logic connects — your assigned reviewer will notice when they don't.

Ignoring ASSIST page and character limits specific to each attachment

Not every attachment in ASSIST shares the same limit as the Research Strategy sections. Some human subjects attachments have their own page or character constraints. Read the ASSIST instructions for your specific FOA before finalizing these sections — limits that apply to parent R01 announcements can differ from program-specific FOAs.

Frequently Asked Questions

Does the NIH clinical trial definition apply to my behavioral intervention study?

It may. NIH's clinical trial definition is broader than the FDA's and includes behavioral interventions that prospectively assign participants to conditions and measure health-related outcomes. If you're randomizing participants to different behavioral conditions and measuring any biomedical or behavioral outcome, your study likely meets the definition. NIH has a clinical trial decision tool you can walk through before submitting — use it, and document the conclusion in your files in case questions arise later.

Do I need a DSMB for a minimal-risk study?

Not necessarily, but you do need to address safety monitoring regardless. For minimal-risk studies, a brief description of the PI's responsibility for monitoring and reporting adverse events is often sufficient. What you can't do is leave the question unanswered. Reviewers are instructed to assess whether safety monitoring is appropriate for the level of risk — a missing answer reads as an oversight, not as a decision.

Can I submit without IRB approval in hand?

Yes. NIH doesn't require IRB approval at the time of application for most research categories. You report the current status — pending, approved, or exempt — and the name of the reviewing institution. Approval must be in place before NIH will issue the Notice of Award, so projects with longer IRB timelines should start that process early regardless of where the application stands in review.

What if my study enrolls a single racial or ethnic group?

You can do this, but the scientific justification needs to appear in the inclusion section itself — not only in the Significance section of the Research Strategy. If there's a scientific rationale for studying a specific population, write it clearly in the inclusion attachment. Reviewers aren't trying to require diverse enrollment when the science doesn't call for it — they're looking for evidence that the choice was intentional and thought through.

NIH Human Subjects Research Plan: What the Protection of Human Subjects Section Needs to Actually Say

Table of Contents