How To “Search” A Patent: USPTO Is NOT PubMed

It is often easy to spot a physician’s patent.

The “References Cited” section has a long list of peer-reviewed papers from the indexed literature. While relevant to the clinical problem, these references are almost always less relevant to the physician’s patent than the patents in the “References Cited” section. This happens because medical papers use citations for different reasons than do patents.

When we write a paper for an indexed journal, we tend to include a lot of references for a lot of reasons. We want to impress the readers and reviewers that we know the literature and it adds bulk to the paper. More legitimate reasons include supporting the Background section (do we really need to see 4 or 5 references supporting the claim that a myocardial infarction or a malignant tumor is a bad thing) and the Discussion section (you see, my results are not really different from the other papers — so they must be reasonable and printable — yet they are different enough so that this paper should not be rejected as being repetitive).

The point of citations in a patent is to list all of the references that would be material in considering whether to grant the patent. So, their purpose is almost the opposite of the references in a medical paper. The citations in a patent are referred to as “prior art” and consist of patents that are close enough to your idea that they should be considered in deciding whether to allow or reject your patent. Another difference is that both you and the patent examiner get to add citations. (The ones actually considered by the examiner have an asterisk* by them.) These citations come out of the examiner “searching” the claims in a patent application. Patent examiners usually do not go beyond searching the patent database. However, anything published anywhere can be fair game if it has a publication date before the date of your invention. (Remember, there are two sets of rules for whether something is “prior art” – the US which currently has a grace period of 1-year before the filing date of your application, and the rest of the world which generally allows anything published or publicly available before the filing date of your invention to be considered as prior art.)  If they do search beyond existing patents, they tend to go to IEEE Explore which is a very large database of technical articles.

When you think that you have a patentable invention the first thing to do is verify — with a simple Google web search — that there aren’t any existing devices “practicing the invention”.  If this initial search doesn’t turn up anyone currently practicing your idea for an invention, it is time to “search” the patentability of your idea for an invention. Go to USPTO.gov or Google Patents  and put in some key words and search on your idea. (Advantage of the Google site is that they make it easy to print out the full patent but thay appear to be missing many patents. For this reason, I never go to Google Patents first.) So, far this is just like going to pubmed.gov to search for literature references. Don’t forget to also search on published patent applications. (Inventors have the option to have their applications published 18 months after filing; this is required if any international filing is done.) A big challenge is ensuring that you have used the correct key words. Warning: the key words for a patent search may or may not be the same as those used in a Medline search. And, the Patent Office search function does not include synonyms such as provided by Medline with their fabulous MESH feature. They also have the ability to search by classifications (“art area” in patent parlance).  While this option can be helpful if you know how to use their classification system, I find this confusing.

A good trick is to do “forward” and “backward” trolling in the USPTO database until you have fleshed out the prior art. After your first key-word search you will get a list of patents.  Click on those and find the patents that are closest to your idea. Note if there are any unexpected key words, as well as alternative synonyms. In the first web-page of each patent that you found there is a section called “Referenced By” which gives newer patents that cited this one. Following that trail forward will often find recent patents that you might have missed with a simple key-word search. Study the patents you found with your forward trolling. Each of them will have a listing of older patents that were cited by the one you are looking at. You can then troll backwards to read these older relevant patents. If you do this a few times you will typically find the patents most closely related to your idea.

Don’t be tempted to ignore a patent that is close to your idea. Not only is it your legal responsibility to list the close prior art, it is a very good idea to be proactive. The patent examiner knows how to search patents better than you do and is not going to miss these.  Citing the most relevant art can give you an edge with an examiner as they can see that you have done your homework. As a bonus, you now get a chance to argue in your application why your idea is better and how the other prior art really missed the key point of your idea for an invention. Any company considering licensing your patent will certainly do a careful search long before they write you a check.

There is another — more psychological — incentive for you to carefully search and list the closest art. Some patent examiners tend to ignore the citations that you submit (possibly reasoning that your list is self-serving) and prefer to rely only on the prior art that they find (the Patent Office version of NIH – “not invented here”). Thus, some patent lawyers suggest (only half-joking) that often the best way to get a threatening reference ignored is to list it.           

If you ever get into litigation over your patent, you will quickly find that a comprehensive and brutally-honest prior art list is extremely helpful. The parties on the other side always try to “invalidate” your patent by convincing the judge or jury that someone else beat you to the invention. They will pull out the closest patent that they can find and wave that in court. However, if you have already listed that patent it is harder to overcome the presumption that the patent examiner considered it and decided that your idea was different enough to warrant issuing you a new patent.

Comments

[Arlen D Meyers, MD,MBA]

A few other tactics;
1. Hire someone to do this for you. There is an entire food chain of patent experts short of patent lawyers who do patent research, prior art searches, patent mappting, etc
2. Check the databases of Federal Agencies , like NIH, NASA and NSF, for previously awarded SBIR/STTR grants. You'll be surprised how many people are already funded to get your idea to market
3. Work closely with your university technology transfer office. Once you file an invention disclosure, the patent search will be next step.