Is it that important to do a prior art search?

[Arbu]

It seems I could spend a lot of time doing a prior art search on the various databases. It doesn't seem that there's much obligation on me to do this - when I draft my patent application I understand that I'm only expected to cite prior art that I'm aware of and that is particularly relevant. I'm a micro entity so the filing fees aren't going to be that much - less than I would value my time for doing a prior art search. So can I just skip doing a prior art search and take the view that if my application gets rejected because of prior art, well, too bad?

Also I'm finding that when I look at a patent on Google patents I can't find it on the USPTO patFT database. For instance I'm looking at patent no. US6041312A and Google patents finds this really easily. But if I enter this as a search time in the patFT database I can't find the patent at all, whether I search under Patent No. or All Fields and whether or not I leave the "US" prefix out. Why is this?

[Robert K S]

So can I just skip doing a prior art search and take the view that if my application gets rejected because of prior art, well, too bad?

To quote this FTC report, "[t]he applicant must reveal material prior art that already is known but has no obligation to conduct a search that would bring additional prior art to his or her attention."  See MPEP 2001.06.  You seem to understand the duty of candor and the risks of not conducting a search, so I'll say no more.

Don't know the answer to your specific question about the search databases but in my experience all the search databases suck.  No exceptions.  The free ones suck.  The high-priced ones suck.  Google has two of them(!) and they both suck for different reasons, but usually I can get by using them in combination.  Fortunately, there are enough options that if you don't find what you're looking for using one option, another option may prove useful.  You may also want to check to see if a library near you is a Patent & Trademark Resource Center (PTRC) which may be able to provide you with even more search options.

[Arbu]

Thanks, that's helpful.

[inventurous]

A counterpoint to your instinct regarding obligations and Robert's excellent answer is that, if you're not somewhat intimate with the field, you may later wish you had some awareness of what was out there when you were drafting your application.  When you're dealing with an Office Action two years from now it might be a lot more difficult to deal with references that you could have handily worked around at time of drafting.

[Arbu]

I've found a patent which says it will do what my invention does without giving any explanation as to how. To my knowledge the owner of that patent hasn't achieved what he said he would do. Perhaps I should mention that as evidence of the utility of my invention.

[still_learnin]

I've found a patent which says it will do what my invention does without giving any explanation as to how. To my knowledge the owner of that patent hasn't achieved what he said he would do. Perhaps I should mention that as evidence of the utility of my invention.

First, "utility" is a term of art in patent law. From Wikipedia:

In United States patent law, utility is a patentability requirement.[1] As provided by 35 U.S.C. § 101, an invention is "useful" if it provides some identifiable benefit and is capable of use.[2] The majority of inventions are usually not challenged as lacking utility,[3] but the doctrine prevents the patenting of fantastic or hypothetical devices such as perpetual motion machines.[4]

So it's very rare that the Examiner needs to be persuaded of utility, and I don't know of any patent attorneys who write anything in the spec about utility.

I've found a patent which says it will do what my invention does without giving any explanation as to how. To my knowledge the owner of that patent hasn't achieved what he said he would do. Perhaps I should mention that as evidence of the utility of my invention.

Perhaps by "utility" you really mean "I have actually figured out how to accomplish the desired result while the prior art clearly has not."

It's true that the specification of a patent application must explain how to make/use the claimed invention. This is called the enablement requirement (Section 112 1st paragraph). Your spec must also demonstrate that you had "possession" of the invention. That's the written description requirement, also in Section 112 1st paragraph. Written Description is somewhat related to, though not the same as, enablement.

Claiming a result without explaining how that result is achieved is likely to lead to Enablement and/or Written Description rejections. For example, a claim to a Star Trek transporter would trigger such rejections -- and likely a rarely seen Utility rejection as well.

None of the above requirements/rejections have anything to do with differentiating your invention from the prior art. Differentiating your invention from the prior art is, however, how you avoid/get over prior art rejections (102 and 103).

Sounds like the Examiner might use the patent you found as prior art against you. You avoid this by claiming something that is a) not found in the prior art and b) not obvious in light of the prior art.

If you actually receive a prior art rejection from the Examiner, you can then explain how the invention you *claim* is different than what is *described* in the prior art. Or you can provide evidence and/or technical reasoning that explains why/how the prior art patent does not/cannot do what you claim. (This is called arguing that the reference is not enabled.) Merely asserting that the prior art patent won't achieve the result is generally not enough. And you should be aware that this argument is not usually persuasive.

Some applicants file a specification which has a section explaining how the claimed invention is different than the prior art. I'll let someone else here explain why that's generally a bad idea.

[bartmans]

I've found a patent which says it will do what my invention does without giving any explanation as to how. To my knowledge the owner of that patent hasn't achieved what he said he would do. Perhaps I should mention that as evidence of the utility of my invention.

It may well be the case that you do not have to explain how the invention works. The only requirement that relates to the 'how'-question is the requirement of enableemnt: a person of skill in the art should be able to reproduce the invention (or to do whatever is claimed).

That these are two different things may be illustrated with an example in the life sciences. You may claim a therapeutic composition which is able to solve headache (let's assume a tablet with aspirin in it). Then - as an applicant - you only have to show that it works, not how it works: the reasons why the aspirin would take away the headache do not need to be given (if they would be known altogether).
In this sense, the how can be hidden in a black box: you put something in a black box and a result comes from the black box. If this is novel and unobvious, it may be patented without knowing what happens in the black box.

In such a case the 'discovery' how the invention works will even no longer deemed to be patentable.

[PaulWeathers]

Hi

So I may file the following application:

Method of treating a patient with headache.

Headache is a very unpleasant condition and a treatment for this condition is requested by patients with headache.

I have surprisingly found out that administering an effective amount of aspirin to a patient suffering from headache is successful in treating the patient suffering from headache.

The amount necessary for being effective can be determined by gavage of increasing amounts of aspirin and evaluating the severity of headache that remained after gavage. It is preferable to wait some time after gavage of the aspirin, preferably from 1 minute up to 60 minutes, before the remaining severity is evaluated.

I claim

1.  A method of treating a patient with headache, the method comprising administering an effective amount of aspirin to a patient suffering from headache.

So as long as the examiner does not find prior art that takes away novelty or makes the skilled person assume that administration of aspirin bears a reasonable chance to treat headache, I will get my notice of allowance?

Do I need proof?
What happens if someone (examiner, 3rd person...) declares "I have taken loads of aspirin and it did not treat my headache at all"?

Thx
Paul

[dbmax]

For instance I'm looking at patent no. US6041312A and Google patents finds this really easily. But if I enter this as a search time in the patFT database I can't find the patent at all.

Leave out the "A" as well.

When searching PatFT for patent no 6041312 I find this:

http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO2&Sect2=HITOFF&p=1&u=%2Fnetahtml%2FPTO%2Fsearch-bool.html&r=1&f=G&l=50&co1=AND&d=PTXT&s1=6041312.PN.&OS=PN/6041312&RS=PN/6041312

[bluerogue]

1.  A method of treating a patient with headache, the method comprising administering an effective amount of aspirin to a patient suffering from headache. [/i]

As medical processes are not my area, I could be completely wrong, but at a glance, I'd say there might be some 101 issues as a person could conceivably be performing the method and you're fairly likely to run into some Mayo/Alice.  There might also be some questions as to undue experimentation as the effective amount may vary from person to person and an amount sufficient to kill a person would also effectively end the headache.  Even assuming no PA issues, I would not be surprised if there were other rejections based on that claim.

[PaulWeathers]

Thanks for the comments.

I almost literally (except the disease and the medication) from the USPTO's "Subject Matter Eligibility Examples : Life Sciences" found here:

https://www.uspto.gov/sites/default/files/documents/ieg-may-2016-ex.pdf

"7. A method of treating a patient with julitis, the method comprising administering an effective amount of anti-TNF antibodies to a patient suffering from julitis." (Claim 7 on page 11 and analysis beginning at the end of page 15)

There might also be some questions as to undue experimentation as the effective amount may vary from person to person

That's why I included the section that explains how to determine the effective amount. As administering some powder and asking the patient how he feels should be quite common for the skilled person I doubt that following the guidance in the specification would count as undue experimentation.

I think it would actually be a nice experiment to file a PA with the exact content of my example, just substituting something for the aspirin that is not knowingly correlated with treating headache. Uranium hexafluoride?

[bartmans]

You do not have an obligation to show how it works, but you do have an obligation to show that it works. How scientifically sound such evidence needs to be is debatable.
This can be illustrated by patents for food compositions with an improved mouthfeel. The improvement can only be 'demonstrated' by results from a taste panel in which people taste the different compositions and score for the properties thereof, such as mouthfeel, color, firmness, etc. How scientifically sound is such a method? Not to speak of disease ratings by questionnaires, which are typically very common in psychiatrics. Can they 'prove' that patients become less anxious or less depressed?

[PaulWeathers]

Is this a "yes"?

[bartmans]

Sorry for the delay in responding, but I have been traveling.

It is indeed a yes, and you will find numerous patents that indeed do not (or at least not fully) explain the working mechanism of the drug.
Often an attempt is provided to hypothesize what the working mechanism is, but this is more to be able to argue that also variants of the drugs (that e.g. would have the same active group or that bind to the same receptor) would also be able to work (without even showing that any or all of these variants work).

Also very often, especially with new drugs, the applications are filed at a stage where there is hardly any testing involved and only the claimd activity of the drug is derived from in vitro or at best animal tests. In those case you need to show that the the therapeutic activity that is claimed is plausible from the data that is provided. The term 'plausible' is a term from the European patent field; I assume that in the USA they rather use the word 'credible'.