Almost every first-in-human device sponsor I work with asks some version of the same question once the protocol starts taking shape: do we need a data monitoring committee, a clinical events committee, a medical monitor, or all three? Founders hear the acronyms from investors, from their FDA consultant and from the ethics committee, and they assume each one is a box to tick. In a 10- to 20-patient early study, buying all three can cost more than it protects. Leaving out the wrong one can stall the study the first time something goes wrong.
I am Julio Martinez-Clark, CEO of bioaccess®. This page explains what each of the three safety-oversight roles actually does in a first-in-human (FIH) or early feasibility device study, what FDA and Brazil’s ANVISA say about them, and how I size them for a small study in Latin America. It goes with our guide to stopping rules and staggered enrollment in an FIH device study, the ISO 14155 inspectable file and our early feasibility studies page. This is operator guidance, not legal or regulatory advice; check the current text of each regulation before you finalize your plan.
One-sentence answer
A medical monitor is the sponsor’s physician who reviews safety events as they happen, a clinical events committee (CEC) independently classifies events and endpoints against definitions written in advance, and a data monitoring committee (DMC, also called a DSMB) independently reviews the accumulating data and recommends whether the study should continue; a small FIH device study almost always needs the first, usually benefits from a lean version of the second, and needs the third when the risk analysis, the population or the regulator calls for it, which in Brazil includes every pivotal study of a Class III or IV device.
What each role does
Medical monitor
The medical monitor is a physician working for, or contracted by, the sponsor. They review serious adverse events as they come in, answer investigators’ protocol and eligibility questions, judge whether an event is a potential unanticipated adverse device effect, and make sure the reporting clocks start. They are not independent of the sponsor, and nobody expects them to be. Their value is speed: in an FIH study the first serious event usually needs a medical decision the same day.
Clinical events committee
A CEC is a small panel of independent physicians, often two or three specialists in the therapeutic area, who adjudicate events against definitions fixed in the protocol or a CEC charter before the first patient is enrolled. They decide whether an event meets the definition of a serious adverse event or a primary endpoint event, and whether it is related to the device, the procedure, or neither. They do not decide whether the study continues. Their output is consistency: the same event at two sites gets the same classification, and an investigator who is enthusiastic about the device is not the only person deciding whether it caused a complication.
Data monitoring committee
A DMC is a group of independent experts, typically clinicians plus a statistician and sometimes an ethicist, that reviews accumulating safety (and sometimes effectiveness) data at planned points and recommends to the sponsor whether to continue, modify or stop the study. FDA’s February 2024 draft guidance, Use of Data Monitoring Committees in Clinical Trials, which when finalized replaces the March 2006 guidance on establishing and operating DMCs, describes exactly this function. The DMC recommends; the sponsor decides and owns the consequences.
What FDA says
Under FDA regulations, sponsors are not required to use a DMC except in the narrow emergency-research setting of 21 CFR 50.24(a)(7)(iv). Device sponsors are, however, required to monitor their investigations under 21 CFR Part 812 (812.40 and 812.46), and unanticipated adverse device effects must be evaluated and reported within 10 working days under 21 CFR 812.150.
The tone has shifted. The 2006 guidance said DMCs were not needed for most studies, particularly early ones. The 2024 draft drops that language and notes that DMCs are now used in early-phase trials in serious diseases or conditions and where causality of adverse events is hard to judge without seeing the full data. FDA’s early feasibility study guidance (section 7.4.2) names a data monitoring committee of clinicians, scientific experts and people with ethics expertise as one possible risk-mitigation element for an early device study. Neither document makes a DMC mandatory for an FIH device study. Both make it something your reviewer may ask about.
What Brazil adds
ANVISA’s RDC 837 of 13 December 2023, which governs device clinical investigations in Brazil, is more specific. Under Art. 57 the sponsor must consider a DMC before the study starts, based on its risk analysis of the device and of study participation. Pivotal studies of Class III and IV devices must have one, and its recommendations go to ANVISA as an amendment (aditamento) within 30 calendar days. The DMC’s functions belong in the clinical investigation plan, and if you decide not to set one up, the plan has to justify that decision. Art. 54 adds that the sponsor, or an independent DMC, evaluates adverse events in aggregate, and that evaluation feeds the annual reports.
So for an FIH study in Brazil, “we didn’t think about it” is not an available answer. You either have a DMC or a written justification for not having one.
ISO 14155:2020, the standard most Latin American ethics committees and regulators expect device sponsors to follow, treats a DMC as optional and expects the clinical investigation plan to say whether one is used and how it works. Ethics committees in the region read that section closely in first-in-human protocols.
How I size oversight for a small FIH study
- Always name a medical monitor with a phone number the sites can reach on the day of the procedure. For a study run in Latin America, that person needs to be reachable in the sites’ time zone and comfortable reading source documents in Spanish or Portuguese.
- Use a lean CEC when the endpoints involve judgment. If your primary safety endpoint is “major adverse events at 30 days” and relatedness to device versus procedure matters to your FDA story, two or three independent adjudicators with a short charter and pre-written definitions are usually worth it. If the endpoint is a simple, objective measurement, a CEC may add little.
- Decide on the DMC against the risk file, not habit. Implantables, life-sustaining devices, fragile populations, follow-up longer than a few months, or a Brazilian pivotal Class III/IV study point toward a DMC. A short, single-country study of a low-risk procedure with staged enrollment may be well served by an independent safety reviewer instead, with the reasoning written into the plan.
- Combine roles carefully in very small studies. In a 10-patient study, one small independent committee can sometimes adjudicate events and also make continuation recommendations, as long as its charter says so, its members have no financial stake in the device, and investigators do not vote on their own restart.
- Write the charter before first patient in. Membership, conflicts of interest, meeting triggers (for example, after each of the first few subjects and after any serious event), what data the committee sees, how recommendations reach the sponsor, and who tells the ethics committee and regulator.
Common mistakes I see
- A DMC charter that exists on paper but whose first meeting is scheduled after the last subject is treated.
- The sponsor’s chief medical officer sitting on the “independent” committee.
- Event definitions that differ between the protocol, the CEC charter and the case report forms, so the adjudicators are classifying against a moving target.
- No plan for who translates source documents for an adjudicator who does not read Spanish or Portuguese.
- Treating the DMC’s recommendation as the regulatory report. In Brazil the sponsor still has to file it with ANVISA; in a US-bound program the sponsor still owns 812.150 reporting.
How bioaccess® runs this
In the first-in-human and early feasibility studies we run in Latin America, we set up the medical monitoring, the adjudication workflow and, where needed, the DMC charter as part of study start-up, and we map each committee’s output to the reporting clocks of every regulator involved: ANVISA, INVIMA, COFEPRIS, Panama’s MINSA, and FDA when the data will support a US submission. The goal is an oversight structure sized to a 10- to 30-patient study that still reads as rigorous to an FDA reviewer looking back at it.
FAQ
Is a DMC the same as a DSMB?
Yes. FDA’s guidance treats data monitoring committee, data and safety monitoring board (DSMB), data and safety monitoring committee and independent data monitoring committee as names for the same function.
Does FDA require a DMC for an early feasibility device study?
No. FDA does not require one outside the emergency-research exception in 21 CFR 50.24. Its early feasibility guidance lists a DMC as one possible risk-mitigation element, and reviewers may ask how you will oversee safety if you do not have one.
Does Brazil require a DMC?
For pivotal studies of Class III and IV devices, yes (RDC 837/2023, Art. 57). For other studies the sponsor must consider one based on its risk analysis and justify the decision in the clinical investigation plan if it does not set one up.
Can the CEC and DMC be the same people?
In a very small study, sometimes, if the charter defines both functions and the members are independent of the sponsor and the investigators. In larger studies they are usually kept separate.
Who pays the committee members?
The sponsor, under written agreements that disclose fees and conflicts of interest. Payment does not compromise independence as long as it is fixed and not tied to the study’s outcome.
Sources: FDA, Use of Data Monitoring Committees in Clinical Trials, draft guidance (February 2024) and Federal Register notice 2024-02849; FDA, Establishment and Operation of Clinical Trial Data Monitoring Committees (March 2006); FDA, Investigational Device Exemptions (IDEs) for Early Feasibility Medical Device Clinical Studies, section 7.4.2; 21 CFR 50.24, 812.40, 812.46 and 812.150; ANVISA RDC 837/2023, Arts. 54 and 57; ISO 14155:2020.