The Workbench · Craft
Not every pediatric study needs two parents' consent
This blog has already covered the eight elements informed consent has to carry under 21 CFR 50.25, and the narrow exceptions that let a physician proceed without obtaining it at all. Both posts treat consent as a question about one adult's signature. Once the subject is a child, Part 50 Subpart D asks a prior question first: which of four risk categories, defined at 21 CFR 50.51 through 50.54, does this investigation actually fall into — and the answer decides not just whether an IRB can approve the study at all, but how many parents have to say yes, and whether the child's own agreement counts as its own separate, checkable requirement under 50.55.
Four categories, not one pediatric standard
21 CFR 50.51 covers research presenting no greater than minimal risk to children. 50.52 covers research presenting greater than minimal risk, but with a prospect of direct benefit to the individual subject, justified against that benefit. 50.53 covers research presenting greater than minimal risk with no prospect of direct benefit to the individual subject, permitted only where the risk is a minor increase over minimal and the research is likely to yield generalizable knowledge about the subject's own disorder or condition. 50.54 is the residual category: research that doesn't satisfy any of the first three but presents a reasonable opportunity to understand, prevent, or alleviate a serious problem affecting children's health or welfare — and it can only proceed after referral to FDA, consultation with a panel of experts, and an opportunity for public review and comment. A device IDE built for pediatric subjects has to clear one of these four findings before anything else about the study's design matters.
The category is the IRB's own finding, not the sponsor's label
Which category applies isn't a determination a sponsor makes and hands to the IRB for a signature. The IRB itself has to reach the finding, against the specific criteria each section states, before it can approve the study at all. A protocol that doesn't fit any of 50.51 through 50.53 has no local approval route available — the only path left is 50.54's national-level review, a heavier and slower process by design, because the categories it skips past are the ones built for research an IRB can evaluate entirely on its own.
Permission scales with the category
50.55(e) ties the number of parents required directly to which category the study fell into. Research approved under 50.51 or 50.52 needs the permission of only one parent. Research approved under 50.53 or 50.54 needs the permission of both parents, unless one parent is deceased, unknown, incompetent, or not reasonably available, or unless only one parent has legal responsibility for the child's care and custody. A consent file that keeps a single “parental permission obtained” checkbox, without a field for which category triggered the requirement, can't show whether one signature was enough or two were actually owed.
Assent is a separate determination, not a smaller version of consent
50.55(a) requires the IRB to determine, separately from the permission question, whether the children involved are capable of providing assent — a judgment the section directs the IRB to base on the ages, maturity, and psychological state of the children in the study, not a fixed age cutoff. Where the IRB finds assent is required, 50.55(b) requires it to also decide how that assent gets documented. The IRB can waive the assent requirement, but only under conditions that mirror an informed-consent waiver: where the research holds out a prospect of direct benefit important to the child's well-being that's available only in the research context. Waiving assent and satisfying parental permission are two different dials, and a study can clear one without the other ever being asked.
Where this meets the file
A pediatric-study consent record needs three distinct fields a general Part 50 consent file never has to carry: the risk category the IRB actually found under 50.51 through 50.54, with its stated rationale; the resulting one-parent or two-parent permission requirement under 50.55(e); and the assent determination — required, documented how, or waived under 50.55(c)'s own narrow test. A pediatric-consent worksheet built around that structure is previewed in the launch catalog. If your program tracks a pediatric device study's consent requirements differently, the shelf takes that correction directly.
The Regulatory Toolkit launches soon — a free shelf of source-mapped templates, checklists and browser-only tools for regulatory teams. Get one email when it opens, or contribute a template.