FDA accepts Latin American device data under 21 CFR 812.28 and 814.15 when it meets GCP, documentation and validation conditions and applies to the US intended-use population. A sponsor checklist.
October 27, 2026
7
min read
By
Julio G. Martinez-Clark, CEO, bioaccess®
In short
The Food and Drug Omnibus Reform Act of 2022 (FDORA) requires Diversity Action Plans for Phase 3 drug studies and for device studies under an IDE or supporting a 510(k), De Novo or PMA — but only for studies that begin enrollment 180 days after FDA publishes final guidance. As of October 2026 that guidance is still a draft. Latin America brings together Indigenous, African-descent, European-descent, mestizo and Asian-descent communities across ICH-GCP countries, so sponsors can plan for representative enrollment from the first-in-human phase onward.
FDA accepts clinical data from Latin America for IDEs, 510(k)s, De Novo requests and PMAs when the study is well designed, well conducted and run under good clinical practice (21 CFR 812.28). Accepted data still has to persuade FDA that the results apply to the people who will use the device in the United States. That second question — whether the data represent the US intended-use population — is where most planning succeeds or fails.
The rule: 21 CFR 812.28
Since February 2019, 21 CFR 812.28 has set one framework for data from clinical investigations conducted outside the United States. It covers an IDE and every major device marketing pathway: PMA and HDE (sections 515 and 520(m)), 510(k) premarket notification, and De Novo classification (section 513(f)(2)). For background, see our guide to FDA acceptance of Latin American clinical data.
FDA will accept such data if the investigation is well designed and well conducted and three conditions are met:
- A GCP statement. The sponsor states that the study followed good clinical practice, which the regulation defines to include independent ethics committee review and approval before the study starts, continuing ethics review, and freely given, documented informed consent.
- Supporting information. For significant-risk devices, the sponsor submits the full set described in 812.28(b): investigator qualifications, research facilities, the protocol, a summary of results, ethics committee details and decisions, the consent form, and how GCP was ensured. Non-significant-risk and exempt studies submit or keep on file a reduced set.
- Validation. FDA must be able to validate the data, through an on-site inspection if needed.
A study that cannot meet these conditions does not automatically disappear: the regulation lets FDA consider it as supportive information or grant a waiver. Sponsors should not plan around either.
The higher bar: 21 CFR 814.15
For a PMA, 21 CFR 814.15 first points back to 812.28. It then sets a stricter test for a PMA based solely on foreign data. FDA may approve it only if:
- the foreign data are applicable to the U.S. population and U.S. medical practice;
- the studies were performed by clinical investigators of recognized competence; and
- the data can be considered valid without an FDA inspection, or FDA can validate them through inspection or other appropriate means.
The regulation also encourages a presubmission meeting with FDA when an applicant plans to rely only on foreign data. In practice, that meeting is where a Latin American strategy gets confirmed or redirected.
"Applicable to the US population" is a clinical question
Sponsors often treat applicability as paperwork. It isn't. FDA reviewers ask whether differences between the study setting and the US could change how the device performs. The usual factors:
- Patient characteristics: disease severity, comorbidities, body size, anatomy and, where relevant, skin type.
- Standard of care: comparators, concomitant therapy and follow-up practices that differ from US practice.
- Operators and settings: clinician training and facility type compared with where the device will be used in the US.
- Endpoints and measurement: whether assessments, imaging and core-lab methods match what FDA expects.
None of these rules out Latin American data. Each one has to be addressed in the protocol and in the submission narrative, before the first patient is enrolled.
Representativeness and the intended-use population
FDA's question is about generalizability: does the evidence reflect the people the device is intended for? Congress put this into law in the Food and Drug Omnibus Reform Act of 2022 (FDORA), which requires Diversity Action Plans for device studies under an IDE or supporting a 510(k), De Novo or PMA. Those plans must state enrollment goals by race, ethnicity, sex and age group, the rationale for the goals, and how the sponsor will meet them.
Timing matters. The requirement takes effect only for studies that begin enrollment 180 days after FDA publishes final guidance. As of October 2026, FDA's June 2024 guidance is still a draft marked "Not for implementation"; the page was taken down in January 2025 and restored under a court order. The requirement is in statute but not yet active. Sponsors who design for the intended-use population now won't have to redesign later — and will be better placed to answer the applicability question that 812.28 and 814.15 already raise today.
Why "Hispanic" in the US is not the same as "Latin American"
A tempting argument says that because US Hispanic patients are underrepresented in device trials, Latin American sites fill the gap automatically. That's too simple, and FDA reviewers know it.
- "Hispanic or Latino" is an ethnicity category, not an ancestry profile. In the US it covers people of Mexican, Puerto Rican, Cuban, Dominican, and Central and South American origin, each with different mixes of Indigenous American, European and African ancestry.
- Ancestry varies widely across Latin America. Countries, and regions within them, differ in ancestral composition; a cohort from one city does not stand in for all US Hispanic patients.
- Environment and care differ too. Diet, access to care, disease stage at presentation and prior treatment can differ between a patient in Bogotá and a Colombian-American patient in Miami.
The defensible position is narrower and stronger: Latin American sites can broaden the ancestry, body-type and clinical diversity of a device dataset. Then show, with collected demographic and clinical data, how the study population maps to the US intended-use population, rather than assuming it does.
What this means for early feasibility and first-in-human studies
Many US medtech startups start in Latin America with an early feasibility or first-in-human study, then move to a US IDE pivotal study. The 812.28 framework applies to data from those early studies whenever they are submitted to support an IDE or marketing application. Running the FIH study to the 812.28 standard from day one keeps the data usable; collecting demographics with the same categories you will use in the US study keeps it comparable.
Sponsor checklist
Before relying on Latin American data in a US device submission:
- Confirm the pathway (IDE, 510(k), De Novo, PMA, HDE) and whether foreign data will be supportive or primary.
- Request a Pre-Submission on your foreign-data strategy, especially for a PMA based solely on non-US data (814.15(c)).
- Write the GCP statement into your plan from the start: ethics committee approval, continuing review, documented informed consent (812.28(a)(1)).
- Plan the 812.28(b) supporting information early: investigator qualifications, facilities, protocol, results summary, ethics committee details, consent form, and how GCP was ensured.
- Be inspection-ready: source documents, monitoring records and data access that let FDA validate the data (812.28(a)(3)).
- Write an applicability rationale: patient characteristics, standard of care, operators, settings and endpoints compared with US practice.
- Define the US intended-use population and set enrollment goals by race, ethnicity, sex and age group, even before DAPs take effect.
- Collect demographics using FDA categories, consistent with your planned US sites, so data can be pooled and compared.
- Avoid assumed ratios: there is no FDA-set US/non-US split. Base site mix on indication, epidemiology and FDA feedback.
- Track the DAP final guidance: studies that begin enrollment 180 days after it is published will need a plan.
The bottom line: Latin American data counts toward a US device submission when it is GCP-compliant, documented to the 812.28 standard, open to FDA validation and — for a PMA based solely on foreign data — applicable to the US population and US medical practice. Designing for representativeness from the first patient turns a regional dataset into US-grade evidence.
Frequently asked questions
Does FDA accept clinical data from Latin America for a device submission?
Yes. Under 21 CFR 812.28, FDA accepts data from clinical investigations conducted outside the US to support an IDE, 510(k), De Novo, PMA or HDE if the investigation is well designed and well conducted, a GCP statement and the required supporting information are submitted, and FDA is able to validate the data.
Can a PMA be based only on Latin American data?
It can. Under 21 CFR 814.15(b), a PMA based solely on foreign data may be approved if the data are applicable to the US population and US medical practice, the investigators are of recognized competence, and the data can be validated. FDA encourages a presubmission meeting first.
Is there a required ratio of US to non-US subjects?
No FDA regulation sets a US/non-US ratio. The right site mix depends on the device, the indication, epidemiology and FDA feedback, ideally from a Pre-Submission.
Are Diversity Action Plans required now?
FDORA 2022 requires them for device studies under an IDE or supporting a 510(k), De Novo or PMA, but only for studies that begin enrollment 180 days after FDA publishes final guidance. As of October 2026 the June 2024 guidance is still a draft.
bioaccess® helps medtech sponsors plan and run early feasibility, first-in-human and pivotal studies in Latin America to the standards FDA applies to foreign data. If you are weighing a Latin American site strategy, we are glad to share what we have learned.
Sources: 21 CFR 812.28 and 21 CFR 814.15 (eCFR); FDA, "Diversity Action Plans to Improve Enrollment of Participants from Underrepresented Populations in Clinical Studies," draft guidance, June 2024; FDA, Report to Congress: Diversity Action Plans Summary FY2023–FY2024; AgencyIQ, January 2025.
Global Trial Accelerators
Regulatory updates & clinical trial insights for Latin America.
Stay ahead with clinical trial insights for Latin America — delivered to your inbox.