In this article, we go into further detail about where we think the Hearing Officer – and indeed the Patent Office generally – has been led astray, as well as why we think the Celgene claims are not patentable.
It seems that the Patent Office has now built up such a body of its own decisions on the patentability of ‘business methods’ (a term which now appears to apply to a range of computer-implemented inventions which do not necessarily have to relate to financial data processing) that it is in clear danger of losing sight of the basic principles underlying the ‘manner of manufacture’ test.
As we shall explain, the Celgene claims can be shown to be unpatentable based upon some very simple and long-standing principles. However, by arriving at a conclusion via a circuitous route involving previous Patent Office decisions, the Hearing Officer produces a line of reasoning that just does not add up. If we remove the specifics relating to the information content in Celgene’s system, leaving only a generic statement relating to structure and function, then paragraph  of the recent Celgene decision says something absurd:
The system in claim 1 can be a computer system. While a computer system is a tangible article, in the present case that computer is characterised solely by the processing that it carries out. As a consequence, the essence of the claimed system is the processing of information about X, and the generation of a Y. The purpose of the processing is Z. While this is clearly a sensible thing to do, the way in which it is achieved by the system does not lie in the useful arts, or the field of economic endeavour.
The logical conclusion of this reasoning is that no computer-implemented invention would be patentable. Not only is this not the current state of the law in Australia, it is not even consistent with IP Australia’s current public position on ‘software patents’.
PATENTABILITY AND SOFTWAREWhen Deputy Commissioner Phil Spann got up to speak in Sydney, at the recent IP Forum event on software patents (which can now be viewed on YouTube), he said a few things which are pertinent to the question of patent-eligibility and software:
- ‘It is not a patent examiner’s job to decide whether a patent for a particular product is economically justifiable, or socially justifiable.’
- ‘Software per se is not patentable. It is what the software does, what its effect is in the machine, that is patentable.’
- The Patent Office ‘routinely grants’ patents for software inventions, on the strength of the Federal Court decisions in IBM v Commissioner of Patents (a method and apparatus for producing curves on a computer graphics display) and CCOM v Jiejing (apparatus for assembling text in Chinese characters).
Secondly, since ‘what the software does’ is the thing which is patentable, the fact that a computer-implemented invention may be characterised solely by the processing that it carries out cannot disqualify it from patentability. Both IBM and CCOM involved of inventions in which the input and output means were entirely conventional, and which were totally characterised by the processing carried out by a general-purpose computer under instruction from a suitable software program.
Thus if the ‘essence’ of an invention is that it involves ‘the processing of information about X and the generation of a Y for purpose Z’, the mere selection of X, Y and Z cannot be the determining factor of patentability. It cannot be the case that if X is ‘points’, Y is ‘curve’ and Z is ‘more efficiently generating the curve on a display’ (as in IBM) then the claim passes the ‘manner of manufacture’ test, but if X is ‘registration of persons and the risk to a patient’, Y is ‘prescription approval code’ and Z is ‘minimisation of the risk of harm to patients’ (as in Celgene) then it does not.
Nor can it be said – with a straight face – that the practice of medicine, and the safe dispensing of medications, does not lie in ‘the useful arts, or the field of economic endeavour.’
What matters is not the identity of X, Y and Z, but the nature of the ‘processing’ and ‘generation’ aspects. In the words used by the Full Federal Court in Grant v Commissioner of Patents  FCAFC 120, ‘a physical effect in the sense of a concrete effect or phenomenon or manifestation or transformation is required’ (at ). Again, this is not about X and Y (which are, inevitably, just going to be some form of data records in any practical application), but about the aspects of the invention that can bring about change or creation.
CELGENE’S CLAIMCelgene’s claim 1, to which the Hearing Officer’s statement above purportedly refers, is in the following terms:
A system for dispensing a drug to a patient in need of the drug, so as to minimise or prevent the occurrence of an adverse side effect known or suspected of being caused by the drug, the system comprising:This is, quite simply, not a claim directed to ‘the processing of information about X and the generation of a Y’, however you try to define X and Y.
- means for receiving, from a remote pharmacy via a client application, an approval request for a prescription for the drug presented by the patient including:
- a unique patient identification number corresponding to the patient, the patient having been prescribed the drug by a prescriber;
- pharmacy registration data from the remote pharmacy; and
- data representing factors which influence risk of occurrence of a known or suspected adverse side effect if the patient receives the drug;
- a computer readable storage medium storing thereon a database including:
- patient registration data, said patient registration data including the unique patient identification number assigned to the patient by a risk management system;
- prescriber registration data;
- pharmacy registration data; and
- a prescription approval code for association with the prescription to be dispensed upon verifying from the database that the:
- patient, prescriber and remote pharmacy are registered in the medium; and
- risk of occurrence of a known or suspected adverse side effect in the patient is acceptable; and
- means for communicating the prescription approval code associated with the prescription to the client application of the remote pharmacy.
WHY THE CELGENE CLAIM IS NOT PATENT-ELIGIBLEThe above claim is not for a manner of manufacture for the very simple reason that it does not define any meaningful function, processing or generation whatsoever. There is an input device (‘means for receiving’) which is capable of accepting input in the form of patient and risk information from a pharmacy. There is a storage medium (e.g. a hard disk) containing a database which stores collections of information, all of which appears to be either predefined, or merely the same information received via the input device. Finally, there is an output device (‘means for communicating’) which is capable of sending information, in the form of a ‘code’, back to the pharmacy.
The input device, storage medium and output device are a mere – and wholly conventional – collocation of parts doing precisely the job they do in every single computing device ever made. In themselves, they provide nothing functional which could contribute in any way to patentability. The information received, stored, and sent, as it is defined in the claim, has absolutely nothing special about it, and there is no new working relationship created between the parts within the terms of the claim.
As Section 184.108.40.206 of the Australian Patent Office Manual of Practice and Procedure clearly and correctly states: ‘Any presentation of information characterised solely by the content of the information has traditionally been not patentable.’ Equally clear and correct is the statement in Section 220.127.116.11.1 of the Manual that an unpatentable collocation is ‘a number of known integers, process steps, or the association of materials or substances in such a way that no working interrelationship or potential working interrelationship exists between its various constituent parts. That is, each part performs its normal function, and is not functionally dependent on any other part.’
The only remotely ‘functional’ element of the claim is the fact that the ‘prescription approval code’ is ‘for association’ with the prescription ‘upon verifying from the database that’ the parties are all registered in the database, and the risk level is acceptable.
In fact, we would argue that this is not actually at all functional as claimed. The claim is written using passive voice, and specifies only a use to which the available information may be put. No actual processing is defined. Even read as defining ‘potential processing’, the most that could be involved is a mere comparison of information which is already available in the appropriate form for making the comparison.
The claim contains no language that could be said to result in anything being processed, transformed or created. If the claimed invention processed information, and re-presented it in a different and useful form, then it might be patentable. In this case, however, there is nothing brought into being by the operation of the invention that did not exist beforehand in exactly the same form.
The Hearing Officer is wrong to say that the claim is directed to a computer which is ‘characterised solely by the processing that it carries out’. The claim is actually directed to a collocation of computer peripherals characterised solely by the information content that they receive, store and transmit. And for this simple, and very traditional reason, the claim does not define a manner of manufacture.
It is equally incorrect to say that ‘the essence of the claimed system is the processing of information about registration of persons and the risk to a patient, and the generation of a prescription approval code’. Where in the claim does it say that a prescription approval code is ‘generated’? What ‘processing’ is performed? The registrations are either present or they are not, and the risk information is already part of the received input data.
Worst of all, the statement that ‘minimisation of the risk of harm to patients’ is ‘a sensible thing to do’ but ‘the way in which it is achieved by the system does not lie in the useful arts, or the field of economic endeavour’ is not only inaccurate, but dangerously so. Whatever deficiencies the claimed ‘invention’ may have, there is absolutely no question that its intended application lies in the ‘useful arts’. This particular terminology comes from the High Court’s decision in the NRDC case (National Research Development Corporation v Commissioner of Patents  HCA 67, at ):
The point is that a process, to fall within the limits of patentability which the context of the Statute of Monopolies has supplied, must be one that offers some advantage which is material, in the sense that the process belongs to a useful art as distinct from a fine art … that its value to the country is in the field of economic endeavour.
The problem in this case, as we have explained, is not that the alleged invention is not directed to a ‘useful art’. The problem is more fundamental: the claims do not recite any patentable functionality (i.e. the ‘process’ required by the above passage from NRDC) which could belong to any sort of ‘art’ at all!
Unfortunately, however, seemingly ‘clear’ explanations, such as the Hearing Officer’s reasons for rejecting the claims in this case, have a bad habit of coming back to bite later applicants.
CONCLUSIONThe Australian Patent Office appears to have become so fixated on its own string of recent decisions, rejecting claims for not meeting the ‘manner of manufacture’ requirement, that it has completely lost sight of the basic principles involved.
In this case, the Hearing Officer was presented with claims that can be shown to be unpatentable based upon simple principles – the ‘collocation’ principle, and the ‘mere presentation of information’ (or ‘printed matter’) principle – which predate even the High Court’s NRDC decision. Instead, in his eagerness to rely on his own earlier decision in Celgene Corporation  APO 12, as well as the Patent Office decisions in Invention Pathways Pty Ltd  APO 10 and Jumbo Interactive Ltd & New South Wales Lotteries Corp v Elot, Inc.  APO 82, the Hearing Officer produced a decision which may be correct in its outcome, but where the reasons bear little relationship to what is actually claimed.
Indeed, as our analysis above demonstrates, the approach in this case has led to a conclusion with regard to claim 1 which not only misrepresents the claim, by suggesting that it actually involves some ‘processing’ which ‘generates’ something, but in doing so is demonstrably at odds with the principles developed by the courts, and with IP Australia’s own publicly-stated position on software patents.
It is time to go back to basics:
- properly construe the language in the claims;
- determine what is actually being claimed as the invention; and
- apply the principles which have been developed by the courts for the application of section 6 of the Statute of Monopolies.
We know that there are IP Australia staff reading this blog. Last month 574 distinct visits to the site have come from within Australian federal government agency networks. We beg you: please put a stop to this madness. It has gone far enough!
What we need is coherent and consistent decision-making which will allow users of the patent system to make meaningful decisions about investment in, and protection of, innovative technologies.