A Process for Managing Prior Art References

The law requires an Applicant for a U.S. patent to disclose to the United States Patent and Trademark Office to closest known material prior art references. When a patent is litigated, one of the most common defenses is to argue that the Applicant failed to do that and so deceived the Patent Office. Such failure renders the patent unenforceable. Separate from all that, a patent is strongest when it faces and overcomes the closest prior art during prosecution before the Patent Office. All this is powerful incentive to properly disclose the closest known prior art.

But medium to large size companies may own dozens, or hundreds, or even thousands of patents. The prior art that was considered in each of these may also be highly relevant to the current patent applications of that company. It would be cost prohibitive to closely review many hundreds of previously cited prior art references for each subsequent patent application. So the question is, how does one determine the closest prior art from an every growing jumble of many earlier possibly relevant prior art references?

I am not aware of any easy answers. Some attorneys create and maintain ever growing spreadsheets of prior art. For each pending application they aspire to somehow review their spreadsheet to determine what they will cite to the Patent Office. For other attorneys, they basically cite many hundreds of references in each case without determining the specific relevance of any of them (a practice I personally abhor and condemn).

For my own largest client, I use a litigation document management tool—Summation—to cross-reference and track all their prior art. Initially setting this up was quite time consuming, for each case, every cited reference was scanned and entered into the Summation database. And each reference is cross-tagged with all the other cases it ahs been cited in. Each month, a paralegal reviews all the substantive activity in the client’s matters and updates the Summation database.

Then for each new case for that client, I review the subject matter of the claimed invention and determine a list of somewhat related past cases. Sometimes, it is helpful to consider what prior cases we have handled for each inventor. This produces a handmade set list of possibly related cases. I then search the database for every reference cited in this set list. The results screen allows me to view at least the first page of each reference and at least momentarily consider its potential relevance to the case at hand and select any thing potentially relevant for inclusion in an Information Disclosure Statement (IDS) to the Patent Office.

My IDSs produced by this process can still be quite large, but at least I considered each reference and had a genuine reason for including it.

Honestly, the whole process of managing and citing prior art references can become enormously burdensome. But it does have to be done. Sorry about the cost…..

Why Do Patent Claims Have Unnecessary Elements?

There are “a plurality” of possible reasons why a patent claim you are looking at might seem to have unneeded claim elements that complicate assertion of the claim in litigation.

First and worst is low attorney skill. Beginners especially, but some attorneys at all experience levels, simply lack the claim drafting gift. Of course, where the working attorney is an associate at a law firm, this problem is really lack of good partner supervision. Not a surprising problem where the client pays a bargain price, but it can happen at all price points (ouch!).

But even skilled patent attorneys like a little meat on the bones of a patent claim. Imagine an invention of a new widget having an inventive gold sparkle. You could try a claim that simple: a widget having a gold sparkle. But such a claim invites a very direct validity attack in litigation—if you can find any gold sparkle in the prior art, grab it and argue it would be obvious to add to a widget… very reasonable.

One way to protect against that is to recite more widget structure in the claim language, especially if all widgets have this stuff any way: “A widget having at least one signal input and one or more signal outputs, widget memory for storing signals and values, and a processor in communication with the signal input, the signal outputs and the widget memory and transforming the input signal to generate the output signals so as to include a gold sparkle.” Effectively the same thing as a widget with gold sparkle, but it forces a much more complicated and long winded validity argument. And the longer the validity argument, the less powerful it is.

So that is one good reason for the extra unnecessary claim elements—stronger validity. But there is an even better reason to include such things in the claims: you have to. The Patent Office insists.

The Manual of Patent Examining Procedure has hundreds of rules that the Patent Office follows in processing a patent application. It is complicated and difficult to work with—hence patent attorneys have to pass a hard qualification exam to even correspond with the Patent Office. But there are some unofficial rules too. Like, we hate being embarrassed, don’t even try to do that. When having an interview with an Examiner, it is not at all unusual to hear them say: “This claim is too broad. We don’t issue claims that broad.” The attorney hearing that can get all huffy and outraged—That’s not in the rules! That’s not fair! But the reality is that the Examiner has discretionary control of a patent application. He has unlimited power to keep assembling combinations of references in obviousness rejections for much longer than you have time or money.
If the Examiner doesn’t want to issue the patent, he won’t. If the Examiner doesn’t issue the patent, it is never going to be around to be asserted in litigation. If you want a patent, you better hear the Examiner when he says we don’t issue claims that broad, and make him happy that your claims are not too broad. And claims two or three lines long—a widget having a gold sparkle—are going to look too short and too broad to many patent examiners. So adding on the extra unneeded claim elements—and hey, all widgets have this stuff any way—makes the claim look narrower. Suddenly the claim is ten or more lines long and has three or four required elements. The Examiner’s potential embarrassment alarm will not go off, and you just may get a patent issued, albeit with some unnecessary claim elements that your poor litigators will have to deal with.

So a given patent claim may have unneeded claim elements for two reasons, either poor attorney skill or high attorney skill. How do you know which reason applies? I’m not sure you can tell just by looking at one claim.

Why So Many Obvious Dependent Claims?

Why do patents have so many dependent claims that add zero to a validity analysis? As an experienced patent prosecutor, I have an immediate answer: It is a complicated subject affected by a lot of things that are hard to explain. But now that you ask, let me meditate for a bit….

Oh!

A new answer: You are wrong. They add a lot to the validity analysis. Allow the professor to explain.

Claim 1: A widget having three or four claim elements described in a mix of functional and structural language.

Claim 2: A widget according to claim 1, made of plastic.

Claim 3: A widget according to claim 1, which is disposable.

The patent is litigated. Let me go out on a limb and assume that the main defenses are non-infringement and obviousness (the unfortunate defendant cannot find an anticipating reference). A pretty safe hypothetical right? It only covers 90+% of all cases.

Let’s briefly consider the plaintiff’s infringement case. Assuming we can show Claim 1 is infringed, is it going to be any harder to show infringement of Claims 2 and 3? No.

Now the validity attack.

To the jury: Claims 2 and 3 are clearly obvious, all widget in the marketplace are made of plastic and disposable.

Jury: Okay, that makes sense, Claims 2 and 3 are obvious.

To the jury: Reference 1 teaches a whole of stuff including some elements of Claim 1 and Reference 2 teaches a whole lot of other stuff including other elements of Claim 1, and a person with a PhD in the field and six years experience with the complicated requirements of making widgets would be likely to borrow just the right things from Reference 1 and Reference 2 to arrive at all the elements required by Claim 1.

Jury: Huh? You lost me. But clearly Claim 1 is not obvious in the same way that Claims 2 and 3 are obvious, is it? If obviousness is why Claims 2 and 3 are invalid, then Claim 1 is not obvious in the same way, and not invalid.