ANVISA device clinical trial submission guide for Brazil

What ANVISA actually authorizes (and what it doesn’t)

ANVISA, Brazil’s Agência Nacional de Vigilância Sanitária, is the federal health authority responsible for authorizing clinical investigations of medical devices conducted on Brazilian territory. Its jurisdiction covers the investigation itself: the protocol, the device, the risk controls, the investigator qualifications, and the conditions under which the device may be imported for study purposes. What ANVISA does not grant is marketing authorization or commercial registration. Getting a clinical investigation approved and getting a device registered for sale in Brazil are two entirely separate regulatory acts, governed by different processes and different rules.

This distinction matters more than most early-stage sponsors expect. An ANVISA “anuência” (authorization) for a clinical investigation tells you the agency reviewed your Dossiê de Investigação Clínica de Dispositivos Médicos (DICD) and found sufficient grounds to permit the study. It says nothing about the device’s commercial status and cannot substitute for ANVISA registration if the sponsor later intends to commercialize the device in Brazil.

The governing regulation for device clinical investigations is Resolução da Diretoria Colegiada (RDC) nº 837, issued December 13, 2023. RDC 837/2023 replaced previous frameworks and sets the current rules for how clinical investigations must be designed, documented, submitted, and conducted. Every sponsor working in Brazil needs this document as a primary reference, not a secondary one.

Under RDC 837/2023, whether ANVISA authorization is required depends on a combination of device risk class and study purpose. Higher-risk devices (generally Class III and IV under the Brazilian classification system) intended to support regulatory outcomes in Brazil require formal DICD submission and ANVISA anuência before any human use. Lower-risk studies or studies using already-registered devices may qualify for exemption from ANVISA authorization in some circumstances. But exemptions from ANVISA review do not exempt sponsors from the ethical approval pathway, which runs in parallel through Brazil’s CEP/CONEP system and is non-negotiable regardless of device risk class.

Deciding whether you need a formal DICD submission

Use this decision logic before you file anything:

1. Is the device registered in Brazil for the intended indication? If yes and the study uses standard care parameters, a reduced pathway may apply. Confirm against RDC 837/2023 exemption criteria.
2. Is the study intended to generate evidence for ANVISA registration, a de novo assessment, or Brazilian regulatory outcomes? If yes, full DICD submission is required.
3. Is the device unregistered in Brazil and implantable or life-sustaining? Full DICD and ANVISA anuência are required regardless of registration status abroad.
4. Is this a first-in-human (FIH) or early feasibility study (EFS)? Almost without exception, FIH/EFS programs for novel devices require full DICD submission. The rationale is straightforward: ANVISA needs to see the preclinical safety package, risk controls, and monitoring plan before human exposure begins.

Even if you conclude that ANVISA authorization isn’t required for a specific study, Brazilian law (Resolução CNS nº 466/2012, December 12, 2012) mandates ethical review for any research involving human subjects. The CEP/CONEP pathway runs regardless of ANVISA’s involvement.

For FIH/EFS sponsors specifically: don’t underestimate this sequencing challenge. Getting ANVISA anuência and CEP/CONEP approval to land at roughly the same time takes deliberate planning. A gap between the two creates an “activation dead” period where you have one approval but can’t legally start.

DICD dossier checklist: what goes into your submission

ANVISA’s official “Guia de Investigação Clínica de Dispositivos Médicos” (Guia 30, published on gov.br) provides the structured framework for organizing your DICD. Follow it precisely. Reviewers expect documents in the structure the guide specifies.

Core documents every DICD must include:

• Clinical investigation protocol (versioned, dated, signed by principal investigator and sponsor representative)
• Investigator’s Brochure or equivalent investigator information document, summarizing the device’s preclinical and prior clinical evidence
• Informed consent materials (TCLE – Termo de Consentimento Livre e Esclarecido), in Portuguese, covering the specific risks and procedures described in the protocol
• Device technical file: includes device description, specifications, classification justification, applicable standards, and manufacturing information
• Risk management documentation, aligned to ISO 14971 or equivalent, showing identified hazards and mitigations
• Preclinical evidence package: bench, animal, or simulation data proportionate to the device’s risk profile
• Adverse event reporting plan and monitoring strategy
• Investigator qualifications (CVs, certifications relevant to the device type and procedures involved)
• Site qualification documentation for each clinical site

Evidence alignment check (do this before you file):

The most common reason DICD submissions receive pendencies is internal inconsistency, not missing documents. Build a cross-walk table:

• Protocol endpoints → Does the IB present data that supports those specific endpoints?
• Risk controls in the risk management file → Are the same risks reflected in the consent form (TCLE)?
• Monitoring frequency in the protocol → Is it consistent with the adverse event reporting plan?
• Device version in the technical file → Does it match the device version referenced in the protocol?

If any cell in that cross-walk is blank or inconsistent, fix it before filing.

Document control warning: Every document must carry the same version number and date across both the ANVISA submission and the ethics submission to CEP/CONEP. Version drift between the two dossiers is a documented source of pendencies at both ends.

Local representation requirements

Foreign sponsors cannot submit directly to ANVISA without an in-country presence or a designated local representative. Under RDC 837/2023 and ANVISA’s broader regulatory framework, the sponsor of record for a device clinical investigation in Brazil must either be a Brazilian legal entity or must appoint a qualified local representative (sometimes referred to in the sponsor/ORPC context) with legal authority to sign submissions, receive official communications, and respond to pendencies.

Action steps:

1. Before preparing the dossier, confirm who will serve as the sponsor of record in Brazil and ensure they have an active CNPJ (Brazilian tax identifier).
2. Confirm the designated individual or entity who will manage ANVISA portal access and sign the petitions.
3. Document the authority chain clearly. ANVISA may issue pendencies requiring responses within fixed windows; whoever receives those communications must have decision authority to act quickly.

Submitting the DICD: portal, petition code, and fees

ANVISA clinical investigation submissions are handled through the electronic petitioning system on the gov.br/ANVISA portal. The specific petition subject for the DICD authorization is classified under “Produtos para a saúde,” with the petition type described as “Anuência em Processo do Dossiê de Investigação Clínica de Dispositivos Médicos (DICD),” subject code 80104 in ANVISA’s Documentos de Instrução system.

The submission workflow proceeds in this sequence:

1. Prepare the complete DICD dossier (protocol-ready, versioned, cross-walked)
2. Register or confirm the legal entity on ANVISA’s portal
3. Open the petition under subject code 80104, attach all required documents per the Documentos de Instrução checklist
4. Pay the applicable Taxa de Fiscalização de Vigilância Sanitária (TFVS) fee via GRU (Guia de Recolhimento da União). The exact fee is calculated based on company size and economic class. ANVISA publishes the fee schedule; confirm the current amount directly with ANVISA or through the petitioning portal at the time of filing. Do not assume a fee amount from third-party sources, as the schedule can change.
5. Monitor the petition for validation/triage outcome (ANVISA confirms whether the dossier is technically acceptable for review)
6. Respond to any pendencies within the specified window
7. Receive the formal anuência (authorization) document
8. Activate the study only after both ANVISA anuência and CEP/CONEP approval are in hand

Practical tips for faster technical acceptance: use descriptive, standardized file names (protocol_v1.0_date.pdf, not “final_final2.pdf”), stay within the portal’s file size limits per attachment, and submit in Portuguese where required. Dossiers that arrive without a matching Documentos de Instrução checklist or with mismatched file naming conventions frequently bounce at validation before they reach technical review.

What to expect on timeline

No fixed date can be promised, but the milestone model below reflects what sponsors typically encounter:

Milestone	Typical range
Dossier preparation (complete, cross-walked)	8–16 weeks
ANVISA validation/triage	2–4 weeks after filing
Technical review period	30–90 days (may be extended if pendencies are issued)
Pendency response cycle	30–60 days per cycle
CEP site review (local ethics committee)	30–60 days
CONEP review (if triggered)	Additional 60–90 days
First patient enrolled after all approvals	Subject to site activation and import clearance

The CEP/CONEP and ANVISA tracks run in parallel but must both resolve before the study starts. Submit to CEP at the same time you’re finalizing the ANVISA DICD, or as close to simultaneously as your dossier state allows. Waiting for one before starting the other wastes months.

For FIH/EFS programs: sponsors running structured 12-month-to-first-patient timelines (similar to the FIH-12 model bioaccess® uses for programs in other Latin American jurisdictions) need to build pendency cycles into the timeline, not treat them as exceptions. A single pendency cycle can add 6–8 weeks if response is delayed.

Import authorization for investigational devices is also tied to the anuência. The device cannot legally enter Brazil for clinical use until the authorization is granted. Import planning (logistical route, customs documentation, cold chain if applicable, device labeling for investigational use) should begin in parallel with the submission process, not after approval arrives.

CEP and CONEP: the ethics pathway in practice

Brazil’s ethics oversight for human research research runs through a two-tier system. Local Research Ethics Committees (CEPs, Comitês de Ética em Pesquisa) review studies at the institution level. The National Research Ethics Commission (CONEP, Comissão Nacional de Ética em Pesquisa) provides a second layer of review for specific categories of research.

CONEP review is triggered for certain study types including, but not limited to: studies involving novel devices with no prior human use (FIH studies), multi-center studies, and research involving specific vulnerable populations. If your study qualifies, budget additional time. CONEP review adds to the overall timeline and runs through Plataforma Brasil, the national platform for research protocol management.

What you submit to CEP/CONEP:

• The complete protocol (same version as in the ANVISA DICD)
• TCLE (informed consent form, in Portuguese, per Resolução CNS nº 466/2012 formatting requirements)
• Investigator qualifications and site documentation
• Sponsor authorization documents
• Device information consistent with what’s in the DICD technical file

What you receive: a Parecer Consubstanciado (formal opinion document) from CEP, and if required, a supplementary opinion from CONEP. These are the artifacts you need to activate the study at site level.

The single most common ethics pendency for device studies is a consent form that doesn’t accurately reflect the risks described in the protocol or the device’s risk management file. A reviewer comparing the TCLE to the protocol who finds discrepant risk language will issue a pendency. Prevent it with the cross-walk table described in the dossier checklist section above.

Top DICD rejection traps and how to prevent them

Missing or thin device technical documentation. ANVISA reviewers need to see a device description that goes beyond marketing language. Include materials, dimensions, intended use, operating principles, applicable standards, and classification rationale. FIH sponsors sometimes submit an abbreviated technical file appropriate for a pre-submission meeting but not for a formal DICD. That’s a structural error.

Investigator’s Brochure not aligned to the protocol. If the IB summarizes preclinical data from an earlier device iteration and the protocol uses a modified version, the discrepancy will generate a pendency. The IB must describe the exact device version in the protocol.

Unclear stopping rules and monitoring plan. ANVISA’s review of FIH dossiers focuses closely on how the sponsor will identify and respond to safety signals. A protocol that defines primary endpoints but doesn’t specify data safety monitoring board membership, stopping criteria, or adverse event escalation paths will not pass technical review without pendencies.

Protocol endpoints not matching the safety rationale. If your primary endpoint is a feasibility measure (e.g., delivery success rate) but your device risk management identifies a serious potential harm, ANVISA reviewers will expect to see safety endpoints that address that harm. They won’t accept a protocol where the formal evidence plan ignores a documented risk.

Consent form inconsistencies. The TCLE must describe risks in language that matches the risk management file and the protocol. Minor differences in risk terminology between documents cause pendencies at both CEP and ANVISA.

Fast approval prep checklist:

• Complete internal QC gate: have someone outside the authoring team read the entire dossier for internal consistency before filing
• Build and review the Protocol/IB/Risk Management/Consent cross-walk table
• Run a regulatory/ethics reconciliation meeting: confirm the version submitted to ANVISA matches what goes to CEP on the same day
• Confirm file naming and format compliance against ANVISA’s Documentos de Instrução for subject code 80104

When ANVISA issues pendencies: Don’t panic, but don’t delay either. Triage the pendency notice immediately to determine whether it’s a document request (provide a revised document), a clarification request (provide a written response), or a substantive concern (may require protocol amendment). Track every response with a version number and maintain a response log. Send responses through the portal as structured submissions, not informal emails.

How your Brazil ANVISA data can support later FDA work

Running a device clinical investigation under RDC 837/2023 doesn’t generate FDA acceptance by default. But a well-structured Brazil study can produce evidence that supports later U.S. submissions when the study is conducted under Good Clinical Practice (GCP) standards, the protocol is designed with U.S. regulatory endpoints in mind, and the clinical study report (CSR) meets the documentation standards FDA expects.

21 CFR 812.28 addresses IDE applications that incorporate foreign clinical data. The principle: foreign clinical data supporting investigational applications must come from studies conducted in accordance with GCP and must be submitted in a format that allows FDA to assess the data’s reliability. A Brazil DICD conducted under ISO 14155 (the international GCP standard for device investigations) and documented with a full, audit-ready CSR meets this foundational standard.

bioaccess® explicitly structures its FIH programs to produce FDA-aligned evidence packages from studies conducted in Latin America, anchoring regulatory strategy to U.S. pathways while executing in jurisdictions with faster trial activation timelines. The evidence produced from a Brazil or other Latin American investigation can feed directly into FDA pre-submission strategy, IDE applications, and 510(k) or PMA evidence packages, provided the study design, endpoints, and documentation meet the bar.

Data package translation checklist for sponsors:

• Clinical Study Report (CSR) structured to ICH E3 or equivalent device format
• Device description and version confirmation matching final commercial specifications
• Risk management update incorporating clinical findings
• Safety signal summary with adverse event narratives
• Protocol deviation log with impact assessments
• Monitoring artifacts and audit trail documentation

One warning worth repeating: the ANVISA anuência for clinical investigation is not a marketing authorization. It doesn’t accelerate or substitute for ANVISA device registration. It also doesn’t constitute FDA clearance, approval, or any equivalent determination. Sponsors who conflate investigational authorization with commercial authorization create serious downstream problems for their regulatory strategies.

How COFEPRIS, ANMAT, and Panama MINSA compare

For sponsors asking about process differences across Latin American jurisdictions, here’s a focused comparison of the key variables:

COFEPRIS (Mexico): Mexico’s health authority uses a “permiso” (permission) for clinical investigation with medical devices. The submission is structured similarly to Brazil’s in that it requires a technical dossier, protocol, and ethics approval, but the ethics pathway runs through Mexico’s COFEPRIS-recognized IRBs or CEIs (Comités de Ética en Investigación), not a CONEP-equivalent national tier. Timeline expectations differ; sponsors commonly report faster validation cycles in Mexico for well-prepared dossiers.

ANMAT (Argentina): Argentina’s Administración Nacional de Medicamentos, Alimentos y Tecnología Médica governs device clinical investigation authorizations. The process involves submission of a clinical trial authorization request (similar in structure to Brazil’s DICD concept) with device technical documentation, protocol, and ethics approval from a recognized CEIC (Comité de Ética Independiente de Investigación Clínica). Multi-center studies require coordinated ethics review across sites. ANMAT’s review has its own timeline characteristics separate from ANVISA’s.

Panama MINSA: Panama is frequently selected for FIH/EFS programs because its authorization timelines are among the shortest in the region. Panama MINSA’s regulatory pathway for clinical research falls under “Regulación de Investigación para la Salud,” with clinical trial protocols registered through the RESEGIS platform. Bioethics review is coordinated through the national committee ecosystem under CNBI/SENACYT. The speed advantage is real for sponsors who prepare complete submissions; incomplete submissions face the same pendency dynamics as any other jurisdiction.

Across all four jurisdictions, the pattern is consistent: complete, internally consistent submissions get faster approvals. The regulatory framework varies; the preparation discipline required does not.

Official resources for sponsors

Use these sources directly. Don’t rely on summaries.

• RDC 837/2023: Available on AnvisaLegis (legislacao.anvisa.gov.br). Search “RDC 837” to access the full text.
• Guia de Investigação Clínica de Dispositivos Médicos (Guia 30): Published on gov.br/anvisa. Search “Guia 30 dispositivos médicos” or navigate to the clinical research section under “Produtos para a Saúde.”
• Perguntas e Respostas – Pesquisa Clínica com Dispositivos Médicos: ANVISA’s official Q&A page, updated to align with RDC 837/2023. Available at gov.br/anvisa under the clinical research / medical devices section.
• Documentos de Instrução for DICD (subject 80104): Accessible through ANVISA’s electronic petitioning system (peticionamento.anvisa.gov.br). Search under “Produtos para a saúde” for the DICD anuência petition type.
• CEP/CONEP system: Plataforma Brasil (plataformabrasil.saude.gov.br) is the national system for submitting research protocols to CEP and CONEP. Registration and submission guidance is available at conep.saude.gov.br.
• Resolução CNS nº 466/2012: Full text available at conselho.saude.gov.br. This is the foundational Brazilian ethics regulation for research with human subjects.
• Panama MINSA research regulation: gov.pa/minsa under “Regulación de Investigación para la Salud.” RESEGIS platform for protocol registration is linked from MINSA’s research section.

If you’re preparing a first-in-human or early feasibility study for Brazil or another Latin American jurisdiction and need a submission-ready dossier structure that’s already mapped to both local regulatory requirements and U.S. FDA planning, contact bioaccess® to discuss how to sequence ANVISA authorization, CEP/CONEP ethics review, and import logistics for your specific device program.