FDA pathways for founder planning

Use this page as a planning mirror

Even if Canada is the first market, U.S. pathway language helps teams pressure-test intended use, risk, predicates, novelty, and evidence burden. The goal is not to self-diagnose a pathway from a webpage. The goal is to ask better questions before product claims, architecture, and fundraising timelines become too fixed.

What this page helps you decide

This page helps Canadian founders understand U.S. pathway language well enough to ask better planning questions. FDA terms such as 510(k), De Novo, PMA, predicate, Q-Sub, and SaMD often appear in investor, advisor, and market-entry conversations.

Use it as a planning mirror, not as a substitute for regulatory counsel. The practical value is clearer claims, better evidence assumptions, and fewer false timelines.

Pathways at a high level

This page focuses on U.S. pathway language because many teams compare North American go-to-market options. If you are building primarily for Canada, pair this page with Health Canada and ecosystem strategy and Health Canada’s guidance on software as a medical device (SaMD) to keep planning aligned. The U.S. and Canada use different filing structures, but the underlying questions—intended use, risk, and evidence—overlap.

510(k) in plain language

A 510(k) submission argues that your device is at least as safe and effective as a predicate that is already on the U.S. market for a comparable intended use. The heart of the work is the comparison: intended use, technological characteristics, and performance data where differences matter. If your product is “software in a box” or a connected system, the agency still evaluates the whole function patients rely on, including software lifecycle and update practices when applicable. A weak or shifting predicate story is a common source of late surprises, so test predicate choices early with advisors who can challenge your reasoning.

De Novo and PMA, briefly

De Novo exists for devices without a true predicate that still sit in lower or moderate risk bands after classification. Expect to describe novelty clearly, to justify the risk class, and to help FDA establish or refine how the device type is regulated. PMA paths apply when the law and regulation require premarket approval—often higher-risk or life-sustaining contexts—and when the body of needed clinical and quality evidence is commensurately large. For founders, the practical signal is: if experienced regulatory counsel whispers “PMA,” your fundraising, team, and traceability to clinical evidence all need to scale together.

Software as a medical device (SaMD)

Many products in this track are, in whole or in part, software as a medical device—functions that meet the definition of a device on their own, without being part of a specific physical instrument. The FDA publishes dedicated thinking on device software functions and artificial intelligence in SaMD. The parallel Canadian framing appears in the Health Canada SaMD guidance linked from this curriculum’s Health Canada page. You still need a single coherent story: what the software does for patients, how it is validated, and how changes are controlled after release.

Q-Sub: when a conversation helps

The FDA’s Q-Sub (Pre-Submission) program is a formal way to ask questions and get feedback on plans before you file a marketing submission. It is not free consulting and not a guarantee, but for novel software interfaces, ML claims, or borderline device versus wellness positioning, a well-prepared Q-Sub can reduce interpretive risk. Your regulatory advisor will time it against prototype maturity and the specific questions you need answered in writing.

Questions that expose weak assumptions

• What exact clinical or workflow claim would appear in the first public-facing materials?
• Which existing products are similar, and where is your product meaningfully different?
• What evidence would convince a skeptical reviewer that the difference is safe and effective?
• Which future feature would change the risk story enough to require regulatory rethinking?

What to decide early

• Intended use and indications for use language, stable enough to test against.
• Risk class range and a documented list of what would move you up or down the range.
• Core evidence strategy, including verification, validation, and software or hardware design controls.
• Postmarket update policy for model, firmware, and cloud changes, aligned to U.S. FDA expectations for cybersecurity and change control.