Authors of research papers should use and cite online patent databases more frequently, according to a recent Correspondence (Nature 461, 340; 2009), but from a US perspective, this is unsound advice.

An employee publishing a patent citation may be exposing his or her employer to liability for triple damages, and many firms ask their technologists to remain ignorant of the patent literature. US patent law awards triple damages for the period in which an infringer wilfully violates intellectual property. Ignorance is a valid defence against such a claim, so, before contemplating legal action, owners of intellectual property will often alert potential infringers, preemptively starting a wilful-violation clock. Widespread patent citation could even lead corporate lawyers to advise corporate scientists to avoid even reading or citing papers that cite patents.

Although patents represent rigorously reviewed novel work, they cannot be compared to peer-reviewed academic articles. The patent defines the invention (in its claims) and argues for its priority and novelty, minimizing the relevance of prior publications, whether legal or academic, along the way.

An examiner issues a patent only after rigorous review, but his or her report focuses on the patent's claims. Therefore the scientific validity of the patent's arguments and supporting data are not necessarily central to this review.

If the inventors (or others) later recognize serious flaws in the patent's data or scientific reasoning, they are under no obligation to retract or correct the patent. An invention is still valid and enforceable, and remains in public databases, even when the inventor got the science wrong.