4 min read

The sentence above is a quote from a journalist, Sharon Begley, in an article interviewing the astronomer, Carl Sagan, and is often misattributed to him.  How does it apply to patent searching?  You’re the one with an idea for the incredible something, and you will want it known, but as I often see, just because that something isn’t known to you, doesn’t mean it isn’t known.  A brainwave today just may have been someone else’s one hundred years ago too.

I’ve recently been telling you how to conduct aspects of your search, but I’ve never mentioned where you should go to conduct it, or why.

If you want to know if your idea is new, you’re going to conduct what is known as a novelty or patentability search.

You’re in Australia, you’ve heard of IP Australia, and that leads you to AusPat, or you’re in New Zealand and you find your way to IPONZ, and so on around the world.  Is a search there enough?  Not really.  Novelty is global these days, and goes back further in time than either of those two, or any, databases cover.

While there are a few commercial patent databases available, you’re going to have search one (or more) of the freely available databases that contain a collection of patent applications from around the world.  There are a couple of good collections by intellectual property organisations.  These are WIPO’s PatentScope and the European Patent Office’s Espacenet.  They don’t cover every country but they do have country coverage pages, so if you are particularly interested in a country not covered, you may have to go directly to that country’s national patent database.

A few other freely available collections not run by intellectual property organisations include Google Patents and The Lens.  Try them all out as the interfaces are different, and you will find one is more comfortable and intuitive to use.

A second type of search that might commonly be undertaken at a layperson’s level, although far more risky if you get it wrong, is a freedom to operate search, sometimes known as an infringement or clearance search.  This time you’re interested in knowing if you will be infringing someone else’s rights by importing, manufacturing or selling a product.

In this case you generally only need to consider the country or countries you will be doing any of those things in (bearing in mind that imporrting and exporting patented goods and processes is something of a legal minefield), so if that’s Australia you only need to search in AusPat.  The other aspect of a freedom to operate search is that you only need to search applications or patents with a live status, or in AusPat speak, an ‘active’ status.  You may also need to consider any application that recently became inactive as they could be restored to an active status.  If you’re searching in a database that doesn’t allow status searching then a useful substitute is restricting the search to filing dates less than 20 years old (or 25 if you’re searching for a pharmaceutical).  You’ll be searching through some inactive statuses but you will capture all active applications.

You can still use the collection databases above but you’ll also have to add a country restriction to your search.

If you want to find the location of your national database, here is a good place to start.

A third useful type of search is a state of the art search, or landscape search.  This is where you want to know what’s out there in a particular field.  As you’re more interested in what’s new, think of it as a date restricted novelty search, so instead of searching back as far as you can, possibly beyond the first instance of the field, you should restrict the filing date to a recent period such as the last ten or five years, depending on how fast the technology is moving.  The most useful databases in this instance are the collection databases.

There is one last type of search you may have heard of, which is an invalidity or validity search.  The two names appear contradictory but it just depends on which side of the pitch you’re on, whether you’re attacking or defending.  It’s unlikely a layperson would ever have to conduct a search such as this, but think of it as a novelty search on steroids.  You not only look at patent literature but also scour non patent literature as well, digging and digging until you’ve exhausted your options or your budget.  A good novelty search takes me somewhere around five hours to complete, but an invalidity search will take many more than that.

So, go somewhere and look for that incredible something, and if you can’t find it, it’s already yours.

6 min read

You’re probably already lacing up your running shoes at the very thought of the mention of the International Patent Classification (IPC), but please stick around; I promise this will be almost painless.

If you’re like me, your introduction to the IPC was a seemingly random string of letters and numbers rolling off someone’s tongue.  It made no sense at the time, and yet, despite the years, when one might think they have it down pat, there are still times when it’s a head scratcher.

I’m going to try to help you, firstly, find the right patent classification for your idea, and secondly, help you use that information in a meaningful way, but before all that, a few letters and numbers.

A typical IPC classification looks like A61K31/192.  This one happens to relate to pharmaceutical compositions containing ibuprofen, but let’s break it down into its constituent parts.

An IPC classification is composed of five parts: section; class; subclass; group; and subgroup, each building on what has come before.  In the example above the section is A; the class is A61; the subclass is A61K; the group is A61K31, and the subgroup is A61K31/192.

Sometimes you’ll see IPC classifications with a space between the subclass and the group, or with some extra zeros in the group numbers.  They all mean the same thing.

There are eight sections, labelled A to H, and these cover broad technology areas as follows:

A             Human necessities

B             Performing operations; Transporting

C             Chemistry; Metallurgy

D             Textiles; Paper

E              Fixed constructions

F              Mechanical engineering; Lighting; Heating; Weapons; Blasting

G             Physics

H             Electricity

These sections get subdivided and further subdivided until you end up with these figures for the most recent iteration of the IPC published in January 2020.

Eight sections – 131 classes – 646 subclasses – 7518 groups – 68030 subgroups.

I know those are daunting numbers if you want to find the right classification but the IPC is a hierarchical system that is laid out in a very orderly fashion.

Here’s the IPC layout for A61K31/192.


There are a series of dots known as 1-dot subgroups, 2-dot subgroups and so on. All 2-dot subgroups are subsets of 1-dot subgroups, and 3-dot subgroups are subsets of 2-dot subgroups.  What we are aiming for is the last possible subgroup into which we could put our idea, or in this case, ibuprofen.

We’re looking for pharmaceutical compositions containing an organic compound which brings us to the group A61K31.  Ibuprofen is an acid, so A61K31/185 (1-dot).  It’s also a carboxylic acid, so A61K31/19 (2-dot).  It’s not an acyclic carboxylic acid so we drop down to the next 3-dot, and it is a carboxylic acid having aromatic groups, so A61K31/192.  It doesn’t have two carboxyl groups nor does it have an amino group, so we stop at A61K31/192.

Still here?

I can’t imagine I’ve made your views on the IPC more positive so far, but this is where I make it easier.

The bottom line is:  Forget about all those dots (if you want).

The easiest way to find the right classification for your idea is to use the knowledge of someone who knows the IPC inside out.  Conducting a narrow keyword search for your idea should bring up earlier patent applications in the same technology area.  All of these earlier applications will have an IPC classification on their front page, and they are there because a patent examiner or searcher has determined they are relevant to each of those inventions.  The subgroups (e.g., A61K31/192) might be different but you will find that they all have the same group (e.g., A61K31), and it’s the group you’re after.

Now you can look up that IPC group online and see if you can further classify your idea.  Then, in AusPat for example, you can conduct a search for an IPC classification (or IPC Mark as it’s described there) by looking for A61K31/* as in our example.  The /* after the group is necessary for the search to capture all of A61K31 in this database.

Another method of finding the right IPC classification is to use a tool such as the IPC Search tool.  For this I’ll go back to the battery powered surfboard idea from my previous article on quick patent searches.

A search in the IPC Search tool for “surfboard” brings up two classes which are for two different subgroups in the group B63B32 (water sports boards).

Unfortunately for my example’s purposes, WIPO has just created a new class for surfboards or water sports boards in general.  In the 2019 edition surfboards were classified within a class for vessels adapted for special purposes, so until the relevant patent applications are reclassified we also have to search the subgroup B63B35/79.

Doing the same thing for “battery” gives 75 or so results, which should be expected for something that occurs in multiple applications including vehicles and circuits, and also has a military definition.  We need to check each classification until we find the one most appropriate for our needs.  In this case it is H01M.  We are at the subclass level now as the different groups cover various types of battery.

In the previous article we focussed solely on keywords to reduce the number of patent applications to consider in a search for a battery powered surfboard.  The keyword search gave us about 300 hits in the search results.

It would be possible to combine the classifications we have found in a search that looks for patent applications in both the surfboard and battery classes.  It’s not something I do a lot of, and is mainly for trying to pick out a combination of features that has been described in an unusual way or misclassified within the class of interest, such as describing a surfboard as a water sports board.

If we do that search we get about 30 hits in the search results.  It’s not too many, and useful to consider them even if they would also appear in the next couple of searches as well.

The following searches are essentially backups to the earlier keyword search, where we will swap one of the keywords (or set of keywords) for the equivalent IPC classification.  I previously identified the following variations on describing a surfboard:  a surfing board, a watercraft board, a water sports board, a surfing device, and a surf board.  There could be others, but chances are most of them have been classified in B63B32 or B63B35/79.

The search then is, (IPC = B63B32 OR B63B35/79) AND (keyword = battery OR batteries OR …).  You would also need to find a set of keywords that describe a battery.  Doing this search gives about 350 hits, which is consistent with the keyword only search.

Searching the other way round, i.e., (IPC = H01M) AND (keyword = surfboard OR surf board OR …) gives about 50 hits.

Putting all these four searches together gives about 400 hits because there is a lot of overlap between them, but you are getting more than a keyword search alone, so the IPC can be invaluable even at a very high level to provide additional patent applications you would otherwise have been unaware of.

So, take a deep breath and dive into the IPC as far down as you dare.  Even just having a paddle can make your searching much more effective.

Authored by Frazer McLennan and Charles Tansey, PhD

6 min read

I often hear that a patent search won’t take long, or it will be a small search, or get asked to do a quick search, and in some cases that’s probably right, but in other cases, by which I mean most, things are not always as small as they appear.  I’m going to risk expulsion from some secret society by pulling back the curtain and revealing some the black magic that occurs when someone asks for a patent search.

Why do people think a search will be small or not take long?

Perhaps they only know of three or four other companies that do what they do, or that the idea is a relatively simple concept, or that you only need to look for the doohickey part of the whole thingamajig, or that whenever they’ve done a search they only get a few results to look at, or maybe they just don’t understand the scale of what constitutes patent literature.

Two estimates, made in the last three or four years, of the total number of patent applications filed put the figure at about 90 million, and between 80 and 100 million.  If you want something a little more concrete, WIPO says that there are just over 56 million patent applications up to the end of 2018, which is close, but they’ve only been keeping count since 1985, so it’s definitely an underestimate given there are a couple of hundred years of applications they haven’t counted.  This number is set to rise, and rise exponentially.  To single out the highest filing country of recent times, China became the first country to file over one million patent applications in a calendar year, which happened in 2015, and just three years later they have increased that by half again.

The exact number, or even a credible estimate, doesn’t matter.  It’s a big number, and there’s no way any one of us can look through them all.  The thing is though, we have to, and when it comes to your idea, and I mean Your. Actual. Personal. Idea., at the point you say “Can you do a quick search on…”, every one of those 100 or so million patent applications is relevant, and they remain relevant until I can find a reason to exclude them.  What’s left after I find those reasons are your search results.  That’s going to be a relatively tiny number, and that’s what makes a patent search look small or quick.

So how can you exclude millions of potential search results?

The fastest way to get a more manageable set of results is to apply a patent classification code.  A patent classification code is assigned to a patent application when it is filed, and then further refined upon examination of that application.  You’re relying on people with experience in a subject matter field to say that this patent application belongs in this patent class, so when you apply that patent class in your search, your reason for excluding some of those 100 million patent applications is that other people­—experts—say they aren’t relevant.

If we take one patent classification code relating to pharmaceutical compositions containing a chemical compound as the active ingredient, and apply it at its broadest level, A61K31, to that 100 million, we already bring ourselves down to fewer than 3 million patent applications.  We’ve already excluded 97% by broadly applying a patent classification.

Patent classification systems are something for another day (but if you’re keen here’s the IPC and the CPC) so I won’t go too deep, but if you’re looking for something common such as esomeprazole (Nexium), the relevant classification is A61K31/4439, which brings us down to fewer than 70 thousand patent applications, so you have excluded over 99.9% of all patent applications.

Keywords can also be useful tools for excluding patent applications you don’t need to consider.  Some aren’t so useful, such as the brand name for midazolam, which is Versed.  Versed is also a word used in patent applications, commonly found at the end of a specification in a sentence that includes the words “… those versed in the art will appreciate …” that indicates subject matter experts will understand that some minor differences in the invention are obvious, but that doesn’t assist anyone in a search for midazolam by increasing the number of search results by tens of thousands.

Even if you have a keyword that will be useful and restrict your results to a manageable level, you still have to be aware of variations on that keyword.  The English language is very colourful and allows for a wide range of potential descriptions for a single object.

Consider a surfboard.  It’s a well known object and you would be hard pressed to find someone who described it otherwise when asked.  I’ve had a quick look and I can find reference to a surfing board, a watercraft board, a water sports board, a surfing device, and of course, a surf board (with the space between the words).  In terms of how much that restricts your search, ‘surfboard’ provides 15 thousand results while ‘surf board’ provides 80 thousand.

What’s even more useful than a single keyword is a combination of two or more.  We can take an object (the surfboard) and combine it with an essential feature (and apologies to anyone whose hopes and dreams I’m just about to destroy, a battery).  Combining these two sets of keywords so that we are locating them essentially within the same paragraph in a patent specification, i.e. (surfboard OR surf board) NEAR (battery) we can whittle down the 100 million patent applications to just 300 or so, which is eminently searchable.

The last method of reducing the number of search results you have to consider relates to the type of search you are undertaking.  It’s really a novelty search or a validity search that has to consider all 100 million patent applications as relevant at the start of a search.  An infringement search or freedom to operate search has features that automatically reduce the overall numbers to consider, and that’s before you start to think about patent classifications or keywords.  These searches are restricted to a single country, so with the pharmaceutical class A61K31 above, we have already reduced that number by 75%.  These searches are also restricted by filing date to patent applications that have a live status, which in most cases is within the last 20 years, or the last 25 years for pharmaceuticals.  To continue our pharmaceutical example we have removed a further 50%, and that’s before we even consider what the actual status is.  If you were to look through Australian patent applications in the broad pharmaceutical class A61K31 that have a live status, you would be looking at just over 20 thousand records, which is a lot less than the 3 million we found above.

With most of these examples the numbers of search results are still enormous and I’m not saying you would search them but it shows you how the application of a few parameters, alone or in combination, can substantially reduce them to a manageable level, and make that ‘quick’ search, well, quick.

Authored by Frazer McLennan and Charles Tansey, PhD

First of all, what is a patent family?  Like a family made up of people, it’s a collection of things that are related in some way, where the things just happen to be patent documents.  Although all methods of constructing a patent family are artificial, some are more artificial than others, which, I guess, implies that some are almost natural.

There are really two main types of patent family:  simple and extended.  These two are the most logical, but there are a few that aren’t so logical, and really just exist so as to give a convenient name to a particular collection of patent documents.

The table below can be found in quite a few publications explaining how patent families are constructed.  I’ll add a bit of colour to it as we go to help explain the difference between the various types.

Patent families table 1


In a simple patent family all of the patent documents have exactly the same priority document (or documents), which means they all include basically the same inventive subject matter.

Essentially we’re looking at a nuclear family.  Mum and/or Dad are the priority documents, and their offspring are the patent documents.

Family 1 is a single parent family with one kid, while Family 2 is a two parent family with two kids, and so on.  Things are kept, well, simple and easy to understand.

Patent families table 2

While each patent document belongs to a single family, this table clearly shows that a priority document can belong to more than one family, so where I say above that a patent family covers an inventive subject matter, it’s not necessarily mutually exclusive with the subject matter of a second patent family.  For example, Family 4 has the elements of Priority P3, which must also form part of Family 3’s subject matter, but Family 3 also contains the elements of Priority P2, which is sufficient to differentiate it from Family 4 and to call it a separate patent family.

So, why don’t we define a patent family by a single priority document instead…


A complex patent family is generally defined as one where all patent documents are linked by at least one priority document.

The definition says ‘at least one’ but the table below doesn’t provide an example.  In the table above, Family 2 could also be considered to be a complex patent family as it has a pair of priority documents that are both shared by Patent documents 2 and 3.  It doesn’t include Patent documents 1 and 4 as they don’t share both priority documents.

Otherwise, this is a people family where the parents have separated.  You have Family 1 with Mum with three kids, and Family 2 with Dad, also with three kids, but where two of the kids are in common.  Depending on which family you are looking at, sometimes the kids in common are staying with Mum, and sometimes with Dad, and Dad’s just hanging out for that one weekend in a blue moon when he has no kids at home and he can go and play a round of golf.

Patent families table 3

I prefer the simple patent family approach to this one as, in complex families, a single patent document can belong to more than one family.  The simple family allows for the concept of patent equivalents.  As the priority document and subject matter are the same, you can say that Patent document 2 is an equivalent of Patent document 3.  You can look at the claims of one and be certain that they will be substantially the same as the other.  It’s handy when one of the patent documents is not in English as you can read an equivalent that is.  You can’t say the same about Patent documents 1 and 2 in complex Family 1.

So, what’s the solution…


In an extended patent family all patent documents have at least one priority document in common with at least one other family member.  An extended patent family is the marriage between a simple patent family and a complex patent family.

This can lead, in some instances, to extraordinarily large patent families where the family members are in a similar technical area, but with potentially a larger diversity because two family members with different priorities may cover different inventions.

Now you have Mum, your stepdad, Dad, Dad’s new girlfriend, half brothers and sisters, cousins, second cousins once removed, aunts, and mad Uncle Bob, all of whom may live in different cities or countries, perhaps never having had met except for that one time when your grandmother died, and you wouldn’t know them if you passed them in the street.

Patent families table 4

An extended patent family does provide the whole picture by putting everything you need to know in one place.  From there you can tease out a number of simple patent families (or complex ones if that’s your thing), or report the family as a whole.  It really depends on the question you are looking to answer.


This is perhaps the most artificial of all the patent families.  It draws together patent documents that are not related by a common priority, but is determined intellectually, i.e. by sitting down and using your necktop computer.  Patent documents in technical families will sometimes have priority documents in common, but others will have same inventor(s), and the same or similar title, subject matter and drawings, and so should be included for a sense of completeness.

Patent documents can become separated from what would have been their simple patent family as a result of missing a convention filing or divisional applications not being connected properly with their parent application.

This has to be Kevin from Home Alone.  Mom and Dad (Inventors 1 & 2) have applied for “Christmas Vacation” in the United States but have forgotten one of their patent documents when it came to flying out to the “Paris Convention”.

Patent families table 5


A national patent family consists of any patent applications from a single country having at least one priority in common, so for example, in addition to the original application, patents of addition, divisional children, and continuations.


A domestic patent family relates to subsequent publications of same application, i.e. with same number but with different kind codes. For example the published application (A1) and the granted patent (B1).

Where can you find patent family data?

In an automated world, patent family information has to be constructed from priority data, which is why the technical family above has to be constructed manually.  It also goes some way to explaining why some members can get separated from their patent family as priority data isn’t particularly standardised.  Even if you apply a reasonable standard, a missing or extra zero somewhere can throw the system out, but that’s up to people like me to notice and correct.

The best free aggregating sources are the European Patent Office’s INPADOC database available via Espacenet, and to a large degree, WIPO’s PatentScope.  Google Patents should also provide a reasonable patent family.

There are other free patent information databases, like Depatis and the USPTO, among many others, that you can use to construct your own patent family if you are so inclined, but that level of enthusiasm is best reserved for looking for any outliers in countries not usually covered by the aggregating sources.

The other sources are commercial patent databases such as PatBase or Derwent Innovation, among a few others, but of course you will need to pay for the privilege.  Commercial patent database providers define patent families for their own purposes, whether it’s to differentiate themselves from their competitors, or it’s just easier to do from a technical viewpoint, but they are always mixtures of the types described above, and sometimes you have the option of choosing your family, unlike the real world.

Authored by Frazer McLennan and Charles Tansey, PhD

It’s time to talk of mice and McCaw. While that Richie won’t be playing in 2019 edition of the Rugby World Cup soon to be held in Japan, it is time to wonder which ball the new Richie (Mo’unga) will be kicking off with in the final on November 2.

This will be the seventh rugby world cup for which Gilbert have supplied the official match ball, and they continue to make changes to improve handling and aerodynamics. Gilbert have been producing footballs since 1823, and it is said that the ball picked up by William Webb Ellis all those years ago was a Gilbert.

While Adidas have sought to improve their world cup soccer balls by reducing the number of seams, getting down now to a six-piece ball, rugby balls have been settled in their four-piece configuration for many decades. It’s even defined as such within the Laws of Rugby.

So, where then have the changes been made?

After problems with the aerodynamics of the ball used in the 2003 tournament, the shape of the ‘pimples’ on the surface of the ball were changed from round to star shaped to improve handling and kickability for the 2007 tournament. Remarkably, Gilbert (and their new owners, Grays International) don’t seek protection for very much of their innovation, but in this case they did, which was published as WO2006/061608.

The next development was for the 2011 tournament where a new valve was produced, which distributed weight along the seams of the ball to improve rotational stability and accuracy.

Both these features were retained for the 2015 world cup ball, and most of the improvement appears to have been in the polymers and laminations of the synthetic cover material.

Moving forward to the latest ball to be used in 2019, Gilbert have made further improvements to the pimples, making them ‘dual height’, and altering their distribution across the surface. This needed to be done without increasing the actual height of the pimples as it would have affected the aerodynamics, but as there doesn’t appear to be any patent protection and hence publication of the details, all I can say is they claim to have produced a ball with the largest total surface area of any international match ball.

These changes seem to be getting more and more infinitesimal and highly technical, which someone has suggested is creating a rugby ball that is close to perfection, so it will be interesting what Gilbert come up with as we approach 2023.

Authored by Frazer McLennan and Charles Tansey, PhD

IP Australia have just released the new Australian Design Search into the wild, replacing the (now) old Australian Designs Data Searching system, or ADDS.

It’s a vast improvement over the old system, although if you’re familiar with IP Australia’s efforts over the last few years you’ll see a lot in common with the Australian Trade Mark Search.  None of the glitchy bits or unsearchable fields of the old system have been carried over, which will be a relief for those of you who search designs regularly.

The link above will take you to the Quick Search.  It has a limited range of search options, most of which are non-technical; a name or a number.  You can also search images from here but I’ll go into that later.  If you’re not sure what to do, there is a range of colourful help tabs relating to each of the options below the searching field, otherwise try the Help link at the bottom of the search page.

Here’s the Quick Search page.

quick search page

The Advanced Search is exactly what it says.  You can now search multiple search options including images, numbers, keywords, statuses, dates and classification codes.  If you enter data into more than one field it treats the search as if you had an AND between the two fields.

Here’s the Advanced Search page.  You can open this page using the link towards the top right of the page.

design numbers

Many of the fields are similar to the old ADDS search, but are easier to use.  I’ll focus here on some that are different.

The major changes in what you can now search are the addition of status searching, and being able to search the text of the Statement of Newness and Distinctiveness, or the SOND.

Status searching is important to have when you want to limit your search to relevant designs.  I’m simplifying things, but for instance a freedom to operate search is generally limited to active statuses, i.e., pending or registered designs, or you may be interested in whether a registered design is under examination so that you can anticipate it being certified.  The point is, you now have options that you never had before.

Being able to search the SOND is also important.  I’m not saying they are always brilliant works of literature as they often just refer you to the images, but occasionally they contain relevant keywords not found in the product name, allowing you to locate that hidden gem that has been inaccessible in past searches.

Another major change is being able to look up classification codes, and enter them directly into the search page.

The symbol to the right of the classification codes text field, as shown below, opens up the Australian Designs Classification Codes popup.

classification codes

If you’re not familiar with the designs classification, you can search for a class in the text field at the top of the popup.  It looks for words appearing in the classification descriptions so be aware that you may need to try a range of keywords to find what you want.

design classification codes

Otherwise you can click on the text to open up the next level of classification, and in some cases a third level, and select the classes you want by clicking on the plus symbol.  You can select multiple classes as well.

The last new search option is image searching.  It’s an interesting exercise to search for an image, and I’m sure further refinement will happen, but for now you should treat an image search as supplementary to a search of classes or keywords.  It’s simple enough to conduct an image search.  Clicking on the Select Images link on the right hand side of the page allows you to choose an image file from your computer.

select images

Once that’s open you can crop it to any part of the image you wish to search for, before conducting the search.  I think best results are obtained from an image that is a drawing on a white background, but try it for yourself and give feedback to IP Australia about how it might be improved.

Once you’ve conducted your search there are a number of ways you can view the results.  The default grid view is customisable to show between two and five images across the screen.  Mousing over each design provides basic details, and clicking through reveals more detail including the entire gallery of representations for that design.  You can also choose a list view with one line per design.

From your search results, or from My List once you’ve selected a few designs that are relevant, you can export your results to Excel, or produce a pdf search report that outlines the searches you conducted, the relevant designs you selected, and the details of those selected designs.  The Excel spreadsheet is good if you like a lot of detail, including a thumbnail image, and want to customise a report, and the pdf provides a nice summary of your efforts for distribution purposes to management or investors for example.

IP Australia have done a nice job on the new Australian Design Search, and this has been done by consulting widely with users of ADDS and other interested parties to get the best possible outcome.  I say give it a go and conduct some design searching today.  You never know what you might find.

Authored by Frazer McLennan and Charles Tansey, PhD

On May 1, Japan entered a new era, known as Reiwa, following the abdication of Emperor Akihito, and, consequently, ending the Heisei era.  You may be asking, “What does this have to do with Japanese patent numbers?”, and these days the answer is “not much”, but that hasn’t always been the case.

Prior to 2000, the Japanese patent numbering system used the Emperor year (the number of years since the current Emperor ascended the Chrysanthemum Throne) as part of its numbering system to indicate the year in which various events occurred.  The main events that affect those of us searching the Japanese patent database are the publication of patent or utility applications, and publication of the examined patent or utility applications.

The numbering format for these four publications is Yyy-nnnnnn, where Y refers to the Emperor era, yy refers to the Emperor year, and nnnnnn is a six-digit number.

Emperor eras relevant to us are Showa which ranges from 1926 to 1989, and Heisei which ranges from 1989 to 2019, although it is the period from 1989 to 2000 that is most of interest here.

Now things start getting a little more complicated, but stick with me.  The numbering of years within each era restarts with each era, so 1926 is year one of the Showa era, or S01, and so on, up to 1989, which is S64.  1989 is also the first year of the Heisei era, so it’s also known as H01.  2000 is therefore H12, and H30 has just concluded.

One more thing, the six-digit number, nnnnnn, restarts each year at 000001, and the number for the unexamined application is different to the number for the equivalent examined application.  What this means is, for a random number such as H01-066151, this could be the publication number for the unexamined or the examined application.

That’s generally fine within the Japanese system as it only takes one or two searches to find what you’re looking for, but here in the West we removed the Emperor era symbol, and we also took to removing, and I’m not sure it was consistent, some of the leading zeros within both the Emperor year, and the six-digit number.  So the number above became JP 1066151.  Is that H01-066151 or H10-066151? Is it examined or unexamined? Is it a clamp for a sofa or an antidepressant or even a piece of radio telephone equipment?  Thankfully the European Patent Office realised there was a level of confusion, and now insist on the Emperor era as part of the number, although the leading zeros thing is still an issue.  Many other sources of Japanese patent information haven’t made this leap.

After 2000, the numbering system moved to yyyy-nnnnnn, where yyyy is the Western year, and the application keeps this number until grant, regardless of where it is in the application process.  It’s a much more straightforward system.

I also haven’t mentioned anything about granted patent numbers.  They are numbered consecutively, regardless of a change in year or Emperor, although there have been a couple of jumps forward to a round number when the law changes.  The last was to JP 2 500 000 in 1996 for patents.

What you need then, in order to locate older Japanese publications, is a sense of what the Emperor era is, a kind code such as A or U for an unexamined application or B or Y for an examined application, and an applicant name or an idea of the subject matter to help pinpoint the right one.

If not, there’s always this for afterwards.

Authored by Frazer McLennan and Charles Tansey, PhD

A new database has been launched with the principal purpose of providing easily accessible and understandable information about the patent status of a specific medicine in a particular country, by identifying key patents related to the medicine worldwide.

This database is meant to have all this information in a single place, and to complement other patent databases including commercial products, national patent offices and WIPO’s PatentScope.

It is called Pat-INFORMED, or more fully, the Patent Information Initiative for Medicines.

Establishing the patent status of medicines can be a difficult task.  The Orange Book has been providing the relevant information for many years but only contains US patent numbers. Pat‑INFORMED aims to provide similar information on a worldwide basis.

It currently contains information on 160 products in six therapeutic areas plus additional products on the World Health Organization Essential Medicines List. There are over 14,000 patents from over 600 patent families at this time.

The six therapeutic areas are HIV/ AIDS, cardiovascular disease, diabetes, hepatitis C, oncology and respiratory conditions.

Pat-INFORMED is a collaboration between WIPOIFPMA and 20 pharmaceutical companies, namely Abbvie, Astellas, Bristol-Myers Squibb, Daiichi-Sankyo, Eisai, Gilead Sciences, GSK, Ipsen, Johnson & Johnson, Leo Pharma, Lilly, Merck, MSD, Novartis, Novo Nordisk, Pfizer, Roche, Shionogi, Takeda and UCB.

This is undoubtedly a valuable addition to the Orange Book but as a patent information professional with many years of searching pharmaceuticals notched on my belt, it is clear that Pat-INFORMED is not without its shortcomings.

It is limited to granted patents. Pending applications are not listed, for the reason that pending applications can be refused, amended or abandoned, where they may no longer be applicable to a pharmaceutical product. To be fair the Orange Book is the same. These databases are to start you off in the right direction, but patents are not listed if they are solely directed to metabolites, intermediates, packaging or methods of manufacture. The problem is there are many more patent applications, and granted patents for that matter, that will be relevant to the drug you are interested in, and these are not discoverable here.

Some shortcomings are listed in Pat-INFORMED’s terms and conditions of use.

Any participant who lists patents or provides information or other services does so on a purely voluntary basis. This can be contrasted with the Orange Book, where providing patent information is mandatory. Pat-INFORMED is already limited to just twenty participants and their patents. Many drug names do not appear. Hopefully this picks up and we see more pharmaceutical companies adding information.

The next issue is twofold. The terms state that “Once fully populated, information in the database will be updated at least annually”. Firstly, it points to the fact that the database is essentially incomplete, or a work in progress. There’s no indication as to when it will be complete. Secondly, it won’t be updated very often, which can lead to out of date information being relied upon. Commercial patent database providers and many national patent offices update their information daily, or at the very worst, weekly.

There are some challenges to searching for pharmaceuticals, and this database touches on one of those in particular. Pat-INFORMED relies on the International Nonproprietary Name (INN) or ‘generic’ name of the drug as the basis of the search. This is deliberate. It is relying upon participating pharmaceutical companies to provide patent information that might otherwise be undiscoverable by someone searching for an INN in PatentScope, for example.

Generic or brand names are not likely to appear in patent specifications early in the lifecycle of a particular drug, typically because the INN is approved some years after the first, or ‘molecule’ patent.

The language of the INNs and medical sector is different to the language of patents, which describe drugs in different ways or with different names.

It is therefore difficult to identify relevant patent applications filed before the INN is approved through a search using the common, generic name.

A variety of sources of information is required to bridge the information gap between the molecule patent of a pharmaceutical and the recommendation of its INN.

Some of the pieces of information are non patent, such as the chemical name, the manufacturer’s code name and the drug type.

For example, the drug Brexpiprazole, approved in the United States in 2015, was originally filed as a patent application in 2006 and obtained its recommended INN in 2012,        and is known by its chemical name (7-{4-[4-(1-benzothiophen-4-yl)piperazin-1-yl]butoxy}quinolin-2(1H)-one), its manufacturer’s code (OPC-34712) and the drug type (antipsychotic/antischizophrenic).

Two of these pieces of information should be enough to identify the molecule patent as US7888362 as its chemical name appears in claim 7 and the drug type appears in the preamble. The third piece of information has been used in eleven patent families to identify the drug as being of interest, with some prior to the INN being proposed.

Other sources of information are relevant to a patent search. These include the IPC (International Patent Classification), or more recently, the CPC (Cooperative Patent Classification), and citation searching.

The patent classes that a drug such as Brexpiprazole is placed into are very straightforward. The compound class (C07D409) focuses on the benzothiophene ring and what is attached to it; the active ingredient class (A61K31) considers the piperazine ring as the primary ring; and the indications class (A61P25) relates to disorders of the nervous system and specifically mentions antipsychotics and schizophrenia. It is unlikely that a patent application relating to Brexpiprazole will not be placed in the correct patent class.

There are also other patent classes that may be of interest that relate to the pharmaceutical form or the inactive ingredients making up the form.

There are a number of searching techniques available that use these patent classes either alone, together or in combination with other pieces of information such as keywords and relevant parties (especially the manufacturer) that enable a searcher to locate relevant patent families filed between the filing date of the molecule patent and a time when the INN is in common use.

It is also worth remembering just because there is an INN, and it is in common use, that patent applicants will use it. There is no requirement to do so. For instance, use of the generic name ‘Brexpiprazole’ peaks in 2016, some four years after the INN was approved. Some advantage can be sought by not using the INN and instead referring to the chemical name, the manufacturer’s code or sometimes “the compound of Formula I” in order to disguise the application to avoid early discovery of a competitor’s intentions.

A further method of discovering patent applications filed before the INN is approved is citation searching. A patent application filed today may cite some related, earlier filed patents in the preamble, so while those earlier patents may not mention the INN, the later patents make the link.

Here’s a line from the preamble of a patent application filed in 2016, clearly referencing an application that could not have mentioned the INN.

WO 2006/112464 A1 discloses brexpiprazole and its use for the treatment of schizophrenia…

Citations can not only be found in the preamble but also on the front page and in the search report.

Pat-INFORMED has the potential to reduce the amount of searching required when establishing the patent position around specific medicines. However, it should be regarded as a first step for the global health community and pharmaceutical procurement agencies, rather than a complete source of information, particularly given the consequences of missing a critical pharmaceutical patent.

Authored by Frazer McLennan and Charles Tansey, PhD

Calculating patent expiries for most countries is relatively straightforward. Some countries even calculate it for you and display it in a field in their online databases. Why then, is it so difficult to accurately calculate expiry dates for US patents, and be confident in your answer?

The commonly held view of how to calculate US patent expiry dates is as follows:

  1. For applications filed prior to, and still in force or pending on, 8 June 1995, the expiry date is the later of 17 years from the issue date, and 20 years from the filing date.
    • For example, US 5219846 has a filing date of 20 December 1991 and an issue date of 15 June 1993. Adding 20 years to the filing date gives 20 December 2011, and adding 17 years to the issue date gives 15 June 2010. The expiry date is the later date, 20 December 2011.
  2. For applications filed on or after 8 June 1995, the expiry date is 20 years from the date of the earliest related application, or where there are no earlier applications, the filing date.
    • For example, US 6472448 has a filing date of 27 March 2001. There are no earlier related applications, so adding 20 years to the filing date gives 27 March 2021 as the expiry date.
    • A further example, US 7105503 has a filing date of 15 February 2002, but in this case there are earlier related applications. These are a continuation-in-part, filed on 22 December 2000, of US 6436915, which is a continuation-in-part, filed on 7 April 2000, of US 6214613. The earliest related application has a filing date of 7 April 2000, so adding 20 years gives 7 April 2020 as the expiry date, and not 15 February 2022.

With simple examples such as these you can be confident, but as the number of continuations or continuations-in-part or divisionals grows it becomes more difficult, and situations where the continuing applications span more than 20 years are particularly troublesome.

Precise details for claiming the benefit of an earlier filed application can be read in the full article Considerations, concerns and recommendations in calculating patent expiry dates for United States patents, but in short, if an applicant wishes that a pending application has the benefit of the filing date of an earlier filed application, they must make reference to the earlier application. The reference must identify all of the prior applications and indicate the relationship between each application in order to establish copendency throughout the entire chain of prior applications, and usually forms the first paragraph of the specification.

In most cases the earlier filed application upon which the benefit claim is made, is also the earliest filed application, and calculation of an expiry date is straightforward as outlined in (2) above, but sometimes it appears that the later application is still “live” even though the expiry date calculation indicates that it should have expired.

For example, take US7477268. This application contains a paragraph that references five earlier filed applications.

The first filed application has a filing date of 20 December 1991. As US7477268 references all earlier filed applications it would be reasonable to say that it is claiming benefit from the first filed application. This would give it an effective filing date of 20 December 1991 and as such an expiry date of 20 December 2011.

Everything is clear up to this point, except when you view the relevant record in the USPTO register, PAIR, you find that the applicant paid the 4th year maintenance fee on 13 June 2012, which is about 6 months after the “expiry” date. It would appear that the applicant, although having referenced all earlier filed applications in the specification, is not claiming benefit from the first filed application, and there is no indication as to which earlier filed application they are claiming benefit from.

We are often told not to rely entirely upon patent term expiry calculators or other indications as to a patent’s status or expiry date, but instead to seek legal advice or confirmation from the relevant national register as this will be the most accurate and up to date information. However in the case of the USPTO, unless there is a clear indication that a patent has expired or been abandoned, we have to assume that a patent is still current. There is no calculation of the anticipated expiry date, and no clear way of calculating it ourselves.

I understand that many factors come into play here. There are some factors that the USPTO has definitive answers for; factors such as the filing date of the application, patent term adjustments, patent term extensions, terminal disclaimers and statutory disclaimers. Then there appears to be a factor for which the USPTO has no answer; and that is benefit claims. If a clear indication could be made as to the benefit claim, then the rest is easy, and an expiry date calculated. Once we can confidently predict an expiry date then viewing the maintenance fee records in PAIR will make it clear if a patent is still current.

This is a shortened version of a full article which can be accessed at Considerations, concerns and recommendations in calculating patent expiry dates for United States patents.

Authored by Frazer McLennan and Charles Tansey, PhD