Clinical Research Administration

Clinical research administration is the operational and regulatory layer that sits
between a research idea involving human participants and a trial that can legally enroll,
run, and report. It spans ethics review and informed consent, Good Clinical Practice,
FDA regulatory pathways for drugs and devices, public registration and results reporting,
clinical data management, safety monitoring, and the contracts, materials-handling, and
billing rules that keep a trial compliant end to end. Unlike a purely scientific protocol,
almost every step here is also a documented regulatory event — which is exactly the layer
a research administrator, IRB coordinator, or clinical trials office exists to manage.

What counts as a “clinical trial” — and why the definition keeps shifting

The single most consequential administrative question in this cluster is also the
least settled one: whether a given study meets the definition of a “clinical trial,”
because that determination cascades into funding-announcement eligibility, registration
duties, training requirements, and IRB review pathway. Three different bodies define it
differently, and the differences are not cosmetic.

NIH’s definition reduces
to a four-question test — human participants, prospectively assigned to an intervention,
in a study designed to evaluate that intervention’s effect on a health-related biomedical
or behavioral outcome — and it is deliberately phase-agnostic: a Phase 1
pharmacokinetics-only study can still meet it. ICMJE applies a similarly broad,
phase-agnostic definition for its own registration
policy. FDAAA 801’s “applicable clinical trial” (ACT) category, by contrast,
explicitly excludes Phase 1 studies of drugs and biologics — so an NIH-funded Phase 1 study
can trigger NIH’s own registration and reporting policy while never becoming an FDAAA
“applicable clinical trial” at all. A fourth, unrelated definition governs Medicare
coverage: CMS National Coverage Determination 310.1 sets out “qualifying clinical trial”
criteria (therapeutic intent, a Medicare benefit category, and a deeming criterion such as
NIH, VA, DOD, CDC, AHRQ, or CMS funding, or conduct under an FDA IND) that determine
whether Medicare will pay the routine patient-care costs of a beneficiary enrolled in a
trial — a billing-compliance question, not a grants-administration one, and independent of
whether NIH or FDAAA also call the same study a trial.

The definitional landscape shifted materially in 2026. NIH previously treated any study
meeting both its “basic research” definition and all four clinical-trial questions as a
“Basic Experimental Study Involving Humans” (BESH), administered as a clinical-trial
subtype. NOT-OD-26-032 (29 January 2026) ended that classification, with NOT-OD-26-067
(6 April 2026) supplying implementation detail, effective for applications due on or after
25 May 2026. BESH studies no longer carry NIH’s clinical-trial-specific administrative
burden (funding-opportunity gating, ClinicalTrials.gov registration, the fuller
human-subjects form) — but human-subjects protections and NIH’s other generally applicable
policies, including the Data Management and Sharing Policy, still apply. Already-funded
BESH awards continue under their original terms.

Human-subjects protection: ethics review, informed consent, and the Common Rule

No clinical study involving human participants proceeds without independent ethical
review. In the US this runs through an Institutional Review Board (IRB),
constituted under 45 CFR 46.107; in the UK and much of the EU, the equivalent body is a
Research Ethics Committee (REC).
The governing US framework is the Common Rule (45 CFR 46),
the federal human-subjects policy that, following its 2018 revision, some twenty federal
departments and agencies have adopted or committed to follow. Its foundational ethical
principles — respect for persons, beneficence, and justice — trace to the 1979
Belmont Report,
which the Common Rule operationalizes into specific regulatory requirements.

Informed consent is regulated in two
separate pieces that are commonly conflated: 45 CFR 46.116 governs the consent process
itself (voluntariness, comprehension, and specific required disclosure elements), while
46.117 governs documentation of that consent — which is why an IRB can, under defined
conditions, waive the signed-form requirement without waiving the underlying consent
process. A separate four-part test at 46.116(f) governs waiving the consent process
itself. The 2018 revision also created broad consent
as a new regulatory option — a single, prospective consent covering future secondary
research use of identifiable data or biospecimens — where it did not exist as a formal
category before. Three subparts of 45 CFR 46 layer additional protections onto specific
populations: Subpart B (pregnant women, human fetuses, neonates), Subpart C (prisoners),
and Subpart D (children, requiring parental permission plus, where an IRB finds the child
capable, the child’s own assent — an affirmative agreement, not mere non-objection).

Not every human-subjects study needs full-board review. A study can be
exempt human subjects research
if an IRB — never the investigator alone — formally determines it falls within one of the
eight categories at 45 CFR 46.104(d); several of those categories still require a “limited
IRB review” of privacy and confidentiality safeguards before the exemption applies. For
multi-site, non-exempt human-subjects research, NIH’s single-IRB (sIRB) policy (issued 21
June 2016, effective 25 January 2018) requires domestic sites using the same protocol to
rely on one IRB of record rather than each obtaining separate local approval, subject to
narrow exceptions. Once an ethics decision is made, publication-facing documentation
completes the loop: an ethics
approval statement in the eventual manuscript names the approving body, the
approval/protocol reference number, and the approval date (or documents an applicable
exemption). See the practical mechanics in
Informed Consent in Research and
IRB/REC Approval Process.

Research involving identifiable health information carries a second, independent
compliance layer in the US: the HIPAA
Privacy Rule (45 CFR Parts 160 and 164, Subparts A and E), which requires a specific
authorization, a waiver, or another permitted basis before protected health information
can be used or disclosed for research — a separate legal analysis from Common Rule
consent, even though many institutions bundle both into a single participant-facing form.

Good Clinical Practice: the ICH E6 framework and training requirements

ICH GCP is the
International Council for Harmonisation’s E6 guideline: an international ethical and
scientific quality standard for the design, conduct, recording, and reporting of clinical
trials, first issued in 1996 (E6(R1)), integrated with an addendum in 2016 (E6(R2)), and
most recently revised as E6(R3), finalized by ICH on 6 January 2025. The EMA set 23 July
2025 as the EU effective date for the Principles and Annex 1; FDA issued its own final
E6(R3) guidance on 8 September 2025. The R3 revision places more explicit emphasis on data
governance and risk-based quality management than earlier versions. Compliance with ICH
GCP is the standard regulators and sponsors point to as evidence that a trial protected
participants’ rights, safety, and wellbeing, and that its data are credible.

NIH separately mandates GCP training itself, distinct from certifying compliance with
the guideline: NOT-OD-16-148 (issued 16 September 2016, effective 1 January 2017) requires
NIH-funded investigators and staff responsible for the conduct, management, or oversight
of an NIH-funded clinical trial to complete GCP training, refreshed at least every three
years, regardless of trial phase or intervention type. This is a genuinely separate
obligation from general human-subjects-protections training (commonly delivered through
the CITI Program) that many institutions also require —
the two cover different content and are typically both required, not interchangeable.

FDA regulatory pathways: investigational drugs, devices, and manufacturing

A trial studying an unapproved drug, biologic, or a new use of an approved one
generally requires an Investigational New Drug (IND) application under
21 CFR
Part 312 before the investigational product can be shipped or administered across
state lines for a clinical investigation. The sponsor — the entity that initiates and
takes responsibility for the investigation, defined at 21 CFR 312.3, not necessarily a
company: a physician-investigator can hold both roles as a “sponsor-investigator” —
assembles the IND from preclinical toxicology data, an investigator’s brochure, and the
proposed clinical protocol, and the FDA has 30 days to place the IND on clinical hold
before the trial may proceed. Investigational medical devices follow a parallel but
distinct pathway under 21
CFR Part 812: a device study is classified as significant risk (SR, requiring full IDE
and IRB-plus-FDA approval), nonsignificant risk (NSR, an abbreviated pathway relying on
IRB determination alone), or exempt. Drug substance used in a trial is itself subject to
manufacturing controls — see Good
Manufacturing Practice (GMP) for how cGMP requirements scale by phase, including the
narrower Phase 1 exemption under 21 CFR 210.2(c).

Registration, results reporting, and public transparency

Once a trial is underway, transparency obligations run on their own separate clock from
ethics approval. Clinical
Trial Registration and Reporting Compliance covers the full mechanics; in outline,
ICMJE requires prospective
registration — submitted to the registry at or before first participant consent, with
ICMJE treating the submission timestamp, not the public posting date, as operative — as a
condition of eventual publication in an ICMJE-following journal. FDAAA 801 imposes a
separate legal mandate for “applicable clinical trials”: registration within 21 days of
first enrollment and results reporting no later than 12 months after the primary
completion date, backed by FDA civil penalties. NIH’s own Policy on Dissemination of
NIH-Funded Clinical Trial Information (effective the same date as FDAAA’s implementing
rule, 18 January 2017) is broader still — it reaches every NIH-defined clinical trial,
including Phase 1 and behavioral studies FDAAA excludes — and is enforced through grant
terms rather than FDA’s penalty authority. All of this converges on
ClinicalTrials.gov, operated by
the National Library of Medicine in collaboration with FDA, one of the primary registries
recognized under the WHO International Clinical Trials Registry Platform (ICTRP) network
alongside the EU’s CTIS, ISRCTN, and others. Downstream of registration and reporting sits
publication quality: CONSORT 2010, a 25-item
checklist and participant-flow diagram, is the reporting standard against which
randomized-trial manuscripts are checked, closing the loop between what was registered,
what was run, and what was published — see the wider publication landscape at the
Scholarly Publishing cluster.

Clinical data management: the CDISC standards stack

Clinical Data Management is where trial
data acquisition, cleaning, coding, and submission-ready formatting happen, and it runs on
its own standards stack, maintained by CDISC: CDASH structures data at collection, SDTM is
the tabulation format FDA expects for submission, and ADaM is the analysis-ready
derivative used for statistical reporting (SEND is the nonclinical/toxicology equivalent).
FDA’s Study Data Technical Conformance Guide specifies SDTM, ADaM, SEND, and an
accompanying define.xml metadata file as the required submission formats to CDER and CBER.
Adverse-event and concomitant-medication terms are coded against controlled dictionaries —
MedDRA, maintained by the ICH’s Maintenance and Support Services Organization, and
WHODrug, maintained by the Uppsala Monitoring Centre — so that safety signals are
comparable across sites, sponsors, and regulators. The Society for Clinical Data
Management’s Good Clinical Data Management Practices (GCDMP) is the field’s own practice
reference, organized by area of responsibility into “Minimum Standards” and “Best
Practices.” Underpinning all of it is data-integrity discipline commonly summarized as
ALCOA+ (Attributable, Legible, Contemporaneous, Original, Accurate, plus Complete,
Consistent, and Enduring), applied by convention across GxP domains including GCP data,
and 21 CFR
Part 11, FDA’s electronic-records/electronic-signatures regulation, which sets the
audit-trail and access-control baseline for any Electronic Data Capture (EDC) system or
broader Clinical Data Management System (CDMS) handling submission-bound data — a
different layer again from a Clinical Trial Management System (CTMS), which tracks trial
operations (site contracts, monitoring visits, enrollment) rather than the clinical data
itself, though vendor marketing routinely blurs the three terms together.

Safety monitoring and adverse-event oversight

Every NIH-funded clinical trial requires some form of data and safety monitoring,
scaled to risk, under NIH’s foundational 1998 Policy for Data and Safety Monitoring
(NOT-98-084). For multi-site trials involving more than minimal risk, that typically means
an independent Data Safety Monitoring Board (DSMB) operating under a charter that fixes
what will be monitored, how, and on what schedule — reviewing accumulating safety and, in
some designs, efficacy data at prespecified intervals, and holding the authority to
recommend early stopping. This oversight layer sits alongside, not instead of, a sponsor’s
ordinary safety-reporting obligations to FDA and to the IRB across the life of an IND.

Contracts, materials, and billing compliance

A clinical trial agreement is, structurally, a specialized instance of a
sponsored research agreement —
the contract between the research-performing institution and an industry, foundation, or
government sponsor that sets scope, deliverables, IP terms, publication rights, payment
schedule, and reporting obligations, with clinical trials adding indemnification,
subject-injury, and data/sample-ownership terms specific to human-subjects work. Where a
trial involves sharing biological materials across institutions, a
Material Transfer Agreement
(MTA) governs permitted use, ownership of derivatives, and onward-transfer
restrictions — see the practical process in
Material Transfer Agreements: The
Practical Process. Specimens collected for a trial are frequently retained afterward in
a biorepository, a
tissue bank for solid human tissue specifically,
or, more generally, a sample repository —
each raising its own consent-for-future-use, custody, and (per CDISC/DataCite practice)
identifier questions distinct from the trial’s primary dataset.

On the billing side, the CMS “qualifying clinical trial” test under NCD 310.1 (see
above) determines only whether Medicare will pay the routine patient-care costs of a
covered beneficiary enrolled in a qualifying trial — it says nothing about which costs
count as “routine” versus investigational in the first place. That determination is made
institution by institution through a coverage analysis performed before enrollment opens,
cross-walking every protocol-required service against what a study’s insurers would cover
outside the trial, to route each charge correctly and avoid billing insurers for
costs the sponsor is contractually responsible for (or vice versa). Administratively, this
work typically sits with the same research-administration infrastructure — often a
departmental or central
sponsored programs office, split across
pre-award and post-award roles —
that manages the rest of an institution’s sponsored research portfolio, plus
trial-specific staff (principal investigators, coordinators, and a clinical trials office)
layered on top.

Where this cluster connects to research administration and international rules

Clinical research administration does not run in isolation from the rest of a research
office. The trial’s principal
investigator (PI) holds the same continuing scientific-direction and accountability
role a PI holds on any sponsored award, often supported by one or more
co-investigators with defined portions of
the work; the same credentialing infrastructure that recognizes research-administration
expertise generally — see the
Certified
Research Administrator (CRA) credential — applies whether or not a given office
specializes in clinical trials. Where a trial involves participants’ personal data and an
EU or UK nexus, GDPR compliance runs alongside, not instead of, Common Rule and HIPAA
obligations — see
GDPR and Data Protection
Compliance in Research. Multi-site and multinational trials involving controlled
technology, software, or materials can also intersect with
export
control (EAR/ITAR) requirements, particularly for device and diagnostic studies. And
because clinical research is one of biomedicine’s most heavily standards-governed
sub-fields, it connects directly to CASRAI’s broader work on research data management (see
the Research Data Management cluster for data availability,
repository certification, and FAIR-compliance questions that apply equally to trial
datasets), research integrity and export controls (Research
Integrity & Compliance), and publication standards (Scholarly
Publishing). CASRAI’s dedicated medical-research hub,
including its clinical-trials specialty
page, covers the adjacent biomedical-publishing and reporting-standards angle (ICMJE,
CONSORT/SPIRIT extensions, journal-specific requirements) from an authorship and
publication-craft perspective rather than the operational-administration perspective this
cluster takes.

Frequently asked questions

Is a Phase 1 trial a “clinical trial”?

It depends which definition is being applied, and the difference is not academic. Under
NIH’s and ICMJE’s phase-agnostic definitions, yes — a Phase 1 pharmacokinetics or
safety/dose-finding study can meet both tests and trigger NIH’s registration, GCP-training,
and single-IRB requirements. Under FDAAA 801’s narrower “applicable clinical trial”
category, Phase 1 drug and biologic studies are explicitly excluded, so the same study may
have no FDAAA registration/results-reporting obligation at all. Check both, not just one.

Does IRB approval also satisfy HIPAA?

No. Common Rule review (via the IRB) and the HIPAA Privacy Rule are separate legal
frameworks answering different questions — whether the research itself is ethically
sound and consented to, versus whether a specific use or disclosure of protected health
information has a valid legal basis (authorization, a waiver, or another permitted
ground). Many institutions combine both into a single participant-facing consent/
authorization form for convenience, but the underlying determinations, and the
documentation an auditor will ask for, remain distinct.

What’s the difference between NIH’s single-IRB (sIRB) policy and a general reliance
agreement?

NIH’s sIRB policy is a specific mandate: domestic sites participating in NIH-funded,
multi-site, non-exempt human-subjects research under the same protocol must rely on one
IRB of record, subject to narrow exceptions. “Reliance” more broadly is the general
mechanism — typically formalized through an IRB Authorization Agreement — by which one
institution defers ethics review to another institution’s IRB, and it is used far beyond
NIH-funded work, including industry-sponsored multi-site trials that adopt a central or
commercial IRB voluntarily rather than by federal mandate.

Who is the “sponsor” of a clinical trial?

Under FDA’s IND regulations (21 CFR 312.3), the sponsor is the entity — a company,
institution, or individual — that initiates a clinical investigation and takes
responsibility and legal accountability for it, including for submitting the IND. A
sponsor is not automatically the funder: a “sponsor-investigator” is a single person who
both initiates and personally conducts a trial, holding both roles at once, common in
investigator-initiated studies. This is a distinct concept from a sponsored research
agreement’s contracting party, though the two frequently overlap.

Is GCP training the same as human-subjects-protections training?

No. NIH’s GCP training mandate (NOT-OD-16-148) is a separate requirement from
general human-subjects-protections/research-ethics training (often delivered via the CITI
Program) that most institutions independently require of anyone engaged in human-subjects
research. The two cover different material — GCP focuses on trial conduct, data
integrity, and ICH E6 principles; human-subjects training focuses on the ethical framework
and regulatory categories under the Common Rule — and completing one does not substitute
for the other.