Design Strategy
Design patents have long been the quiet, dependable part of a medical device portfolio. They issue quickly, they cost a fraction of a utility filing, and until recently they were remarkably difficult to invalidate. The last of those three is no longer true.
In May 2024 the Federal Circuit sat en banc — its first en banc patent decision in more than five years — and threw out the test that had governed design patent obviousness for four decades. Most device companies have not adjusted. The filings in your portfolio were drafted and prosecuted under a standard that no longer exists.
Under the Rosen-Durling test, a challenger had to clear two gates before an obviousness argument could even begin. First, they needed a primary reference that was basically the same as the claimed design. Second, any secondary reference had to be so related to that primary reference that features in one would suggest applying them to the other.
That first gate did most of the work. If no single piece of prior art looked substantially like the patented design, the analysis stopped there. In practice it meant that a device design with any meaningful visual distinctiveness was close to unassailable, and everyone drafting and enforcing design patents priced them accordingly.
In LKQ Corp. v. GM Global Technology Operations LLC, 102 F.4th 1280 (Fed. Cir. 2024), the court held the Rosen-Durling framework improperly rigid and inconsistent with Supreme Court obviousness precedent. In its place it applied the same Graham v. John Deere framework that has governed utility patents since 1966: the scope and content of the prior art, the differences between the prior art and the claimed design, the level of ordinary skill in the art, and objective indicia of nonobviousness.
A primary reference is still required, and it must still be visually close to the claimed design. What is gone is the requirement that it be basically the same. The court also confirmed that the prior art must be analogous — in the same field of endeavor as the article of manufacture, or otherwise reasonably pertinent — and that the differences are assessed through the eyes of an ordinary designer in the field of the article, not an ordinary observer.
The USPTO moved immediately. A memorandum to the examining corps issued the following day, effective at once, instructing examiners to apply the new framework.
Two features of device design make the change bite more than it does elsewhere.
The first is that device form is heavily constrained. Handpiece geometry follows the grip and the approach angle. Housing volume follows the board and the battery. Ports, seals, and surfaces follow sterilization and cleaning requirements. Console layouts follow the workflow and the standards that govern alarms and indicators. Those constraints make designs across a category converge, and a converged field is a field where more prior art is now close enough to serve as a starting point.
The second is that the ordinary designer standard has real content here. An ordinary designer of surgical instruments knows the ergonomic and regulatory constraints, and a challenger can now argue that what looks like a creative choice was the predictable response to a known constraint. Under the old test that argument rarely got a hearing, because the challenger could not get past the primary reference gate.
The categories most exposed are the ones where visual differentiation is genuinely narrow: instrument handles and housings, trocars and cannulas, cart and console enclosures, disposable sets and trays, and the graphical user interfaces that increasingly ship as part of an AI-enabled device.
Treat design patents as a portfolio layer rather than a standalone fence. A design patent covering the housing of a device whose novelty lies in its mechanism or its signal processing was always a supplement. It is a thinner one now, and a portfolio built on design filings alone is more fragile than its issue rate suggests.
File more, and file narrower and broader. The claim in a design patent is the drawing, so a family that claims the whole article, subassemblies, and specific ornamental features separately gives more places to stand when one of them meets a close reference. Broken-line disclaimer practice is the mechanism, and it should be a deliberate decision at filing rather than an afterthought.
Build the record while the design work is happening. The Graham framework makes objective indicia relevant in a way the old test did not reach, and those are hard to reconstruct later. Document the alternatives considered and rejected, and why. If a design choice was driven by something other than the obvious constraint, that reasoning is evidence.
Be realistic about commercial success. It is the strongest of the objective indicia and the hardest to establish for a medical device, because devices are bought on clinical performance, reimbursement, and contracting relationships. Success attributable to a design has to be shown, not asserted.
Reassess what is already granted. A design patent you are relying on to block a competitor is a different asset than it was in early 2024. Where one is doing real commercial work, it is worth knowing now what a challenger would find, rather than during an assertion.
The court declined to spell out how every part of the utility framework maps onto designs, and left it to later cases. What counts as analogous art for an ornamental design is the largest open question, and it matters most in fields like medical devices where an argument can be made that a designer would look to adjacent product categories. Whether indicia such as long-felt need and the failure of others have any sensible application to a design is also unresolved.
That uncertainty is itself a reason to act rather than wait. Portfolios are being built now against a standard whose edges will be drawn by the cases that get litigated first, and the filings that survive will be the ones that gave themselves more than one position to defend.