« Compact Prosecution | Main | VoIP »

December 18, 2007

Citing Prior Art

In many instances, the proposed IDS rules will create a prior art search dilemma for patent prosecutors. The new rules require extensive analysis of cited art, reportage that characterizes the claims and thus invites prosecution estoppel, hence hampering enforcement potential, particularly applying the doctrine of equivalents. In a field crowded with art, to minimize cost while maximizing patentability under the new IDS rules, a crafty prior art strategy is imperative.

The proposed IDS rules require for an "explanation" of any prior art citation over 25 pages, or over 20 cited references:

An “explanation” of a document would include, for each reason for which the document is being submitted:
1. An identification of:
(i) The specific feature(s), showing(s), or teaching(s) that caused a document to be cited, and
(ii) a portion(s) of the document where the specific feature(s), showing(s), or teaching(s) may be found (e.g., by page and line number); and
2. A correlation of the specific feature(s), showing(s), or teaching(s) to corresponding specific claim language, or to a specific portion(s) of the supporting specification, if the document is cited for that purpose.

Correlating prior art to "specific claim language" is a recipe for declaring a claim construction. During enforcement, prosecution estoppel looms large as an outcome of meeting the "explanation" requirement.

Prof. Dennis Crouch prepared a historical chart that lists number of references.

The rules will almost certainly have a disparate impact on various art units within the PTO because some art units typically receive more cited references than others. Semiconductor patents, for instance, rarely have 20+ references, while most molecular biology patents cite 20+ references. There is also some correlation with the number of references cited and the expected examination delay within the PTO.

Crouch's analysis overlooks the "long disclosure" problem. Computer-related technologies are commonly long disclosures.

Every patent application filed comes with an oath compliant with 37 CFR 1.63, acknowledging "the duty to disclose to the Office all information known to the person to be material to patentability."

Thus, the new IDS rules impel an economical prosecutor not to search the prior art. While this has not been uncommon practice under the present disclosure regime, it has always been short sighted, and a disservice to an applicant. Post-KSR, not knowing the most relevant prior art is a formula for getting blind-sided, and accusation of malpractice.

The solution is to employ a patentability expert, an outside prior art searcher that can cull references to the necessary minimum of high relevance, eliminating the prospect of inequitable conduct; and, ideally, litigation savvy, able to craft "explanations" that give the most leeway for claim construction during enforcement by minimizing prosecution estoppel. What's more, an out-sourced search frees up prosecutor time to focus on the specification and claims.

Under the new IDS regime, the cost of filing a patent application, especially patentability analysis, is bound to rise. Bang for the buck thus becomes paramount. More than ever, the issues become quality and efficient allocation of resources. The fling-it-and-see-what-sticks prior art searcher isn't going to cut it anymore. You need Patent Hawk.

Posted by Patent Hawk at December 18, 2007 12:21 AM | Prosecution


I disagree with respect to the characterization. What will happen is that patents will be stronger. It will prove much more difficult for a litigator to overcome the presumption of validity. Let's say you have a claim with element A and you cite three references showing element A. You move forward with characterizing the element A of the references as not being your element but merely provided to allow the examiner an opportunity to verify applicant's argument. Then you make the assertion unilateral, it is believed unless the examiner expressly avers otherwise that the applicant's assertions are correct. There is caselaw that would support this finding.
What the invalidating party is then faced with is providing additional references to the jury to overcome th presumption. The difficulty here is that the invalidating party will be hardpressed to show that the element A in their art is different from the element A as characterized. In short a diligent patent attorney can shut down volumes of prior art for use in support of invalidity by carefully characterizing a few references provided.
The best part is that we all know the examiner's will not even look at the applicant's characterization. So yeah you'll get prosectuion estoppel in that litigators will be estopped from successfully using non-cited prior art against issued patens.

Posted by: me at December 18, 2007 9:48 AM

Thanks for that, the effects you described enhance the role of patent searcher and patent translator. Japanese patents are a virtual treasure trove of citable prior art but since 80% never get filed outside Japan and the for-free abstract in PAJ/JAPIO is very short and often very poor, prior art hunters would be well advised to pay a modest amount to somone capable of searching the full text in Japanese. Translation can run expensive but if you get someone (e.g. me) to translate andor summarize the bits of interest YOU can decide whats worth translating.

Posted by: Mike O'Keeffe at December 19, 2007 1:24 AM