BOARDS OF APPEAL
Information from the Enlarged Board of Appeal
Referral by the President of the EPO to the Enlarged Board of Appeal dated 29 December 2003 regarding divergent decisions of two boards of appeal
(Reference: T 385/86, OJ EPO 1988, 308 ff – BRUKER et al.
T 964/99, OJ EPO 2002, 4 ff – CYGNUS, INC et al.)
To ensure uniform application of the law, the following questions of fundamental importance which have arisen over the interpretation of the expression "diagnostic methods practised on the human or animal body" as used in Article 52(4) EPC are hereby referred to the Enlarged Board of Appeal pursuant to Article 112(1)(b) EPC:
1(a) Are "diagnostic methods practised on the human or animal body" within the meaning of Article 52(4) EPC (hereinafter: "diagnostic methods") only those methods containing all the procedural steps to be carried out when making a medical diagnosis, ie the examination phase involving the collection of relevant data, the comparison of the examination data thus obtained with the standard values, the finding of any significant deviation (a symptom) during that comparison and, finally, the attribution of the deviation to a particular clinical picture (the deductive medical decision phase), or
1(b) is a claimed method a "diagnostic method" even if it only contains one procedural step that can be used for diagnostic purposes or relates to the diagnosis?
2. If the answer to 1(b) is in the affirmative: Does the claimed method have to be usable exclusively for diagnostic purposes or relate exclusively to the diagnosis? According to which criteria is this to be assessed?
3(a) Is a claimed method a "diagnostic method" if
(i) it contains at least one procedural step considered as essential for a "diagnostic method" and requiring the presence of a physician (Alternative 1), or
(ii) it does not require the presence of a physician, but presupposes that a physician bears the responsibility (Alternative 2), or
(iii) all procedural steps can also or only be practised by medical or technical support staff, the patient himself or an automated system (Alternative 3)?
3(b) If the participation of a physician (by being present or by bearing the responsibility) is decisive, does the physician have to participate in the procedural step practised on the body, or does he only have to participate in any procedural step considered as essential for a diagnostic method?
4. Does the requirement "practised on the human or animal body" mean that the procedural steps take place in direct contact with the body and that only such steps practised directly on the body can provide a method with the character of a diagnostic method, or is it sufficient if at least one of the procedural steps is practised directly on the body?
Reasons
I. Introduction
The referral concerns the interpretation of the expression "diagnostic methods practised on the human or animal body". Under Article 52(4), first sentence, EPC such methods, along with "methods for treatment of the human or animal body by surgery or therapy", are not regarded as inventions susceptible of industrial application within the meaning of Articles 52(1) and 57 EPC, even if they intrinsically comply with the latter article. According to Article 54(2), second sentence, EPC, this legal fiction does not apply to products, in particular substances or compositions, for use in any of these methods.
The rules governing the patentability of inventions relating to medical methods differ, in some cases considerably, from one patent system to another. That is because the way in which the legislator regulates such issues depends on fundamental considerations influenced by a whole range of factors - legal, social, cultural and especially ethical.
The policy behind the legal fiction in the EPC is to ensure that those who carry out surgical, therapeutic or diagnostic methods as part of the medical treatment of humans or animals should not be inhibited by patents (T 116/85, OJ EPO 1989, 13, Reasons 3.7; T 385/86, OJ EPO 1988, 308, Reasons 3.2; T 24/91, OJ EPO 1995, 512, Reasons 2.4; T 655/92, OJ EPO 1998, 17, Reasons 5.3; T 329/94, OJ EPO 1998, 241, Reasons 3; T 35/99, OJ EPO 2000, 447, Reasons 6; T 964/99, OJ EPO 2002, 4, Reasons 3.1; T 807/98, unpublished, Reasons 2.1; specifically with regard to therapeutic treatment: T 82/93, OJ EPO 1996, 274, Reasons 1.2). However, the intention is only to free non-industrial medical and veterinary activities from restraint by patent rights (G 5/83, OJ EPO 1985, 64, Reasons 22). Thus the exclusion under Article 52(4) EPC is based on socio-ethical policies and considerations in connection with public health (see also T 24/91, Reasons 2.4). Human illness should not be commercialised, so that doctors are free at all times to take the actions they consider suited to cure illnesses or to diagnose them by means of investigative methods (see Schulte, Patentgesetz mit EPÜ, 6th edition 2001, Section 5, point 16).
II. The divergent decisions
1. T 385/86, OJ EPO 1988, 308 ff – Non-invasive measurement/ BRUKER
In T 385/86, Technical Board of Appeal 3.4.1 examined the travaux préparatoires for Article 52(4) EPC to establish its meaning and purpose and concluded that it was intended only to exclude methods of therapeutic treatment from patent protection, so that no one could be hampered by patent legislation in the practice of medicine. Like any exclusion clause, Article 52(4), first sentence, EPC had to be construed narrowly, and this was underscored by the second sentence. In its decision the board was convinced that the only diagnostic methods to be excluded from patent protection were those whose results immediately made it possible to decide on a particular course of medical treatment. So determining whether a method was a diagnostic method for the purposes of Article 52(4) EPC entailed ascertaining whether the claimed method contained all the steps involved in reaching a medical diagnosis. Hence methods providing only interim results were not diagnostic methods within the meaning of Article 52(4) EPC, even if they could be utilised in making a diagnosis (Reasons 3.2; emphasis added).
The board continued: "The systematic list of the steps leading to a diagnosis contained in the relevant literature includes recording the case history, observing, palpating and auscultating various parts of the body and carrying out numerous medical and technical examinations and tests - the examination and data gathering phases - and comparing the test data with normal values, recording any significant deviation (symptom) and, finally, attributing the deviation to a particular clinical picture (deductive medical decision phase) … Even if only one of the last three steps is lacking, there is no diagnostic method but at best a method of data acquisition or data processing that can be used in a diagnostic method" (Reasons 3.3; emphasis added).
In the light of the requirement of being "practised on the human or animal body", and referring to the need to construe Article 52(4) EPC narrowly as an exclusion clause, the board took the view that both examination (measurement of actual value) and establishing the symptoms on the basis of the examination results (ie the deviation measured from the norm) had to be carried out on a living human or animal body (Reasons 4.1). Article 52(4) EPC thus presupposed that both the measured value and the deviation from a norm that had to be regarded as a symptom were directly discernible on the body itself (Reasons 4.2 and 4.3). The condition thus ascertained had of itself to demonstrate the pathological deviation (Reasons 4.3.2).
The application underlying T 385/86 concerned a method for the non-invasive determination of chemical and/or physical conditions, ie temperature and pH value, inside a whole, intact, living animal or human body using local magnetic resonance (LMR). The claimed method resulted in a measured value readable in the high-resolution resonance spectrum that appeared on a screen or plotter page in the final stage of the diagnostic apparatus.
Applying the principles it had developed, the board said the claimed method was not a "diagnostic method" within the meaning of Article 52(4) EPC, because it did not contain all the steps involved in reaching a diagnosis. In the board's view, it was actually a patentable method of measurement, because the measurements it produced did not directly indicate the diagnosis. The board conceded that the measurements implicitly contained information which was of use in making a diagnosis; but what mattered in its view was that the presence or absence of a particular disease was not explicitly discernible from the measured values alone. In the case at issue it was first necessary to identify a significant deviation from a reference value regarded as normal, ie non-pathological, and to attribute it to a particular clinical picture. In the board's view, the claimed method of measuring a variable would become a diagnostic method only if these differentiation and comparison operations were incorporated into the claim, regardless of whether they were performed by a doctor or by computer (Reasons 3.4.1 and 3.4.2).
The board further considered that the "practised on the body" requirement was not met. The measured value was visible on a data carrier detached from the body only after further technical measures which took place outside the body. Any further step did not require the patient's presence. Moreover, the deviation from a norm that had to be regarded as a symptom was also not directly discernible on the body itself. The board felt it was not sufficient simply that an investigation into the state of a human or animal body was carried out for medical purposes. The condition ascertained had of itself to demonstrate the pathological deviation. A measurement was an absolute value which only revealed any irregularity when compared with a norm. It was only the comparison and the explicit indication of how great the deviation had to be to be characteristic of a particular disease or group of diseases that made the measuring method a diagnostic one (Reasons 4.2 and 4.3).
The view taken in T 385/86 regarding the notion of a diagnostic method was consistent with earlier decisions taken by Technical Board of Appeal 3.2.1. In T 61/83, T 208/83, T 18/84 and T 45/84 (all unpublished) the board stated that a "diagnostic method" within the meaning of Article 52(4) EPC had to contain not only the method of examination which yielded the basis for diagnosis but also a pointer to the diagnosis forming the result. In T 45/84 in particular the board had said "that a method is a diagnostic method only if it results in a concrete diagnostic result, and that moreover ... neither the diagnostic result in itself nor the method of examination which produces the basis for it is synonymous with the diagnostic method. Thus we may speak of a diagnostic method only if both conditions are met" (Reasons 2).
T 385/86's interpretation of "diagnostic method" within the meaning of Article 52(4) EPC as a method containing all the steps involved in reaching a medical diagnosis has been confirmed in a number of other decisions.
In Technical Board of Appeal 3.4.1's T 83/87 (unpublished), the invention related to a method for measuring sugar content in the presence of interfering foreign substances by means of a sensor. The board's conclusion was that the measurement obtained by means of an implanted blood sugar sensor could be utilised in reaching a diagnosis, but did not directly produce a diagnosis in the sense of identifying a pathological condition. The claimed method merely provided interim results and was therefore not to be regarded as a diagnostic method (Reasons 3.2).
In T 400/87 (unpublished), Technical Board 3.4.1, likewise citing T 385/86, held that a nuclear magnetic resonance (NMR) method which could be applied to the human body did not come under the Article 52(4) EPC exclusion from patentability. It said the method comprised only the examination and data gathering phase of a diagnosis. Methods which provided only interim results were not diagnostic methods, even if they could be utilised in reaching a diagnosis. Furthermore, a possible deviation from the norm was discernible only from the diagrams, not from the body itself (Reasons 3.1 and 3.3).
The application that was the subject of Board 3.5.1's decision T 775/92 (unpublished) claimed a multiple-step method for determining bone densities for the evaluation of an X-ray photograph of a bone. The board thought the expression "evaluation of an X-ray photograph" appeared so vague and general that it could also cover a diagnostic application. It could be interpreted to mean that the bone density distributions determined by means of the claimed steps were evaluated by a doctor, for example, who compared them with model distributions in order to find out a patient's condition in respect of ageing or bone diseases. Such an evaluation not only provided interim results, but localised a deviation to a particular clinical picture and thereupon allowed the doctor to initiate medical treatment. The board concluded that the method was a diagnostic method which came under the Article 52(4) EPC exclusion from patentability (Reasons 10).
The application considered in T 530/93 (unpublished) related to a method for producing images of a human heart with an NMR imaging system. In its decision, Technical Board 3.4.2 expressly referred to established case law, under which methods comprising only the data gathering phase of a diagnosis and providing only interim results that required a further step before they could be attributed to a particular clinical picture were not diagnostic methods within the meaning of Article 52(4) EPC (Reasons 4).
Similarly, Technical Board 3.2.6 in T 1165/97 (unpublished) found that the claimed method of using a vaginal discharge collector was not a diagnostic method within the meaning of Article 52(4) EPC. It said the exclusion from patentability applied only to methods of which the results made it directly possible to decide on a particular course of medical treatment. Hence the method claimed had to contain all the steps involved in reaching a medical diagnosis. That was not the case with the method being examined, because it did not even involve data gathering (Reasons 4.3).
The claims constituting the subject-matter of Technical Board 3.3.4's T 629/98 (unpublished) were in the form of a "second medical use" and concerned a preparation to be administered to a patient. The board therefore examined whether the preparation was intended for use in a method excluded from patentability pursuant to Article 52(4) EPC. On the basis of the criteria established in T 385/86 it confirmed that the application was a diagnostic method, on the grounds that it made it possible to determine the presence or absence of lung malignancies in a patient to whom the substance was administered. Thus it was a diagnostic method practised on the human body which provided an immediate clinical picture (Reasons 3).
The principles set out in T 385/86 were also applied by Technical Board 3.3.4 in T 1038/00 (unpublished). The method at issue was to be carried out on a sample, so the board concluded that the claimed method was not "practised on the human or animal body" (Reasons 6).
The criteria established in T 385/86 were also the basis for Technical Board 3.2.2's T 807/98 (unpublished). The application at issue claimed "a process for detecting a series of abnormal dysfunction-conditioned events among a multiplicity of normal events in an electrophysiological signal … of an organ exhibiting a dysfunction". The process was characterised by measurement and comparison operations. It provided among other things for the comparison of measured values with a defined threshold value and a defined theoretical value. In its Reasons the board examined in the light of T 385/86 whether a diagnostic method was involved. It observed that the earlier decision had denied the presence of a diagnostic method on the grounds that the examination data obtained had not been compared with normal values and no significant deviation had been recorded. By contrast, in the case at issue there had been such a comparison, and pathological deviations had been detected from a corresponding output signal. In T 385/86 a quantitative value had simply been measured, whereas in the case in point the quantitative values had been processed to form a corresponding signal representing the qualitative information as to whether or not a dysfunction, eg an arrhythmia, was present. As soon as such a heart dysfunction was present, an output signal was issued; if there was no dysfunction, no signal was issued. The absence of the signal meant the presence of a dysfunction could definitely be ruled out. The board expressly deemed that the negative finding that a particular disease could be ruled out constituted a diagnosis (Reasons 2.2). The board further concluded that the assumption of a diagnostic method was consistent with T 964/99, as the method was practised on the human or animal body and referred to a diagnosis. Moreover, for the detection of an arrhythmia, the definition of a threshold value and the definition of a theoretical value were essential steps which were diagnostic in nature and had to be regarded as fundamental diagnostic activities, as a doctor bore ultimate responsibility for them (Reasons 2.3).
Apart from being regularly confirmed in case law, T 385/86's narrow interpretation of the term "diagnostic method" within the meaning of Article 52(4) EPC has also established itself in Office practice. The Guidelines for Examination in the European Patent Office accordingly stipulate in C-IV, 4.3 (October 2001 text): "Diagnostic methods likewise do not cover all methods related to diagnosis. Methods for obtaining information (data, physical quantities) from the living human or animal body are not excluded by Art. 52(4) EPC, if the information obtained merely provides intermediate results which on their own do not enable a decision to be made on the treatment necessary. Generally such methods include X-ray investigations, NMR studies, and blood pressure measurements (see T 385/86, OJ 8/1988, 308)."
2. T 964/99, OJ EPO 2002, 4 ff – CYGNUS, INC
T 964/99 of Technical Board of Appeal 3.4.1 is not consistent with the principles underlying T 385/86 and the decisions based thereon. The application to which this decision related claimed a non-invasive method of sampling a substance from a human or animal body and analysing the concentration of the substance. In particular it comprised the following steps:
"(a) placing at least one sampling chamber at a collection site on a surface tissue of the human or animal body,
(b) extracting the substance or substance metabolite through the surface tissue into the sampling chamber …" by means of an iontophoretic current,
"(c) analysing the sampling chamber for the concentration of the substance or a substance metabolite",
and a step designed to reverse the reactions thus caused.
An auxiliary request directed to the sampling of glucose or a glucose metabolite did not include a step for "analysing the sampling chamber for the concentration of glucose or a glucose metabolite".
In its Reasons the board began by considering T 385/86, observing that it would go against the spirit of Article 52(4) EPC to interpret its provisions in such a way that "manual procedures" of physical examination essential for making a diagnosis and executed by a medical practitioner did not constitute an exception to patentability (Reasons 3.5). It also noted "that the restrictive interpretation of the patent exemption for diagnostic methods adopted by T 385/86 amounts to setting a different standard for diagnostic methods than that established for methods of surgery or therapy, the latter being excluded from patent protection if they comprise only a single step of a surgical or therapeutic nature" (Reasons 3.6).
On the basis of a semantic analysis of the words "diagnosis" and "diagnostic" and their equivalents in the other two official languages, the board concluded that Article 52(4) EPC was meant to exclude from patent protection all methods practised on the human or animal body which related to diagnosis or were of value for the purposes of diagnosis (Reasons 4.4).
The board noted that all method claims in the application at issue comprised the step of sampling a substance from a living human or animal body. The specific embodiments disclosed in the description concerned activities exercised in the course of a medical treatment of patients and served particularly for diagnostic purposes (Reasons 5.1; emphasis added).
The board held that the taking of a body sample for the purpose of a medical examination belonged to a fundamental diagnostic activity, regardless of the technical means used, be it a spatula, a syringe or an iontophoretic current. Hence the claimed step of sampling a substance related to diagnosis and in that context constituted an essential diagnostic measure practised on the living human or animal body. So the subject-matter of the method claims had to be considered a diagnostic method within the meaning of Article 52(4) EPC (Reasons 5.2; some emphasis added).
In arriving at this judgment the board expressly deemed it immaterial that the claimed methods could be performed by a patient himself and that their execution would not have a significant impact on the body or involve a serious health risk. What was decisive was the fact that all method claims on file comprised the step of taking a body sample for the purpose of diagnosis and that such a step was to be regarded as an essential activity pertaining to diagnosis and practised on the living body (Reasons 6.1; emphasis added).
Citing NMR procedures which, although performed on a living human or animal body, only define steps which concern the technical operation of exciting and detecting resonance signals, the board in its decision denied the involvement of a step which would have to be attributed to basic medical activities exercised on the human or animal body. The claimed steps concerned only the internal operation of a technical device. They thus fell without exception within the competence and under the exclusive control of the technician skilled in NMR technology, so the process could be regarded as patentable, "even if it generates and detects physical signals on a living body and its results may be evaluated for diagnostic purposes". The method claims could be considered as containing only steps which concerned the control and internal operation of a technical device in the form of a tomographic NMR machine, so that no specific step of diagnostic character could be recognised (Reasons 6.2).
In contrast, in the application at issue, the crucial step of diagnostic character was the extraction of a body substance for diagnostic purposes, which was to be considered as constituting an elementary diagnostic activity performed under the ultimate responsibility of a physician (Reasons 6.2, end; emphasis added).
T 964/99 had been preceded by a series of decisions based on T 385/86, some of them already mentioned under the previous heading. Among these earlier decisions there are however also a few which seemed to adopt the principles established in T 385/86 but, viewed in retrospect in the light of T 964/99, already showed first signs of the different understanding of the term "diagnostic methods" subsequently applied in T 964/99.
In T 655/92 (OJ EPO 1998, 17), Technical Board of Appeal 3.3.2 examined whether an in vivo NMR imaging technique characterised by the parenteral administration of contrast agents came under the Article 52(4) EPC exclusion. According to the board, the method involved was an invasive technique. In addition, intravenous administration was not devoid of any risk of side-effects, some severe, which meant that just a few drops had to be injected first to determine whether any undesirable side reaction appeared. This could only be determined by medical staff capable of recognising the earliest symptoms of the undesirable reaction and taking any appropriate counter-measures. The diagnostic process considered in its totality thus comprised at least one step essential for the desired diagnostic result, which could not fall under the exclusive responsibility of the technician skilled in NMR technology. While for a process whose steps as a whole were non-medical but technical it was legitimate not to derive a diagnostic character from its final diagnostic purpose, that did not apply to a process for a diagnostic purpose which was to be implemented in its essential steps by medical staff or under the responsibility of a doctor (Reasons 5.3). The fact that the specialists responsible for administering the contrast agent and for actually making the diagnosis could be two different people showed that a diagnostic character within the meaning of Article 52(4) EPC could be recognised in consideration of the medical character of some steps and independently of the final diagnostic activity, which was not part of the claimed process (loc. cit.).
The application to which Technical Board 3.2.2's T 329/94 related claimed a stimulant-based blood extraction assistance method for facilitating sustained venous flow through a human limb towards a venous blood extraction point. In this context the board found that protection had to be refused for a blood extraction method considered per se, since otherwise a great many medical activities would be hampered. "As a matter of fact, withdrawal of blood would fall under the exclusion of Art. 52(4) EPC three times if it could be regarded either
- as a step of a method for treatment by therapy, for example in connection with a dialysis method or with a retransfusion of blood after cleaning, or
- as a step of a method for treatment by surgery when considering that taking blood requires the use of surgical instruments and the operation is performed on the structure of the organism … or
- as a step of a diagnostic method, for example with a view to a blood analysis for finding out the cause of a disease" (Reasons 4).
The board's view was that in determining whether the claimed features constituted medical or diagnostic therapy, the purpose of the claimed subject-matter had to be defined in accordance with the patent application, as understood in the light of the description and the drawings. It mattered little whether the measure was performed by a medical practitioner or another person having medical knowledge or under the supervision of such a person. This sole criterion was not sufficient to decide whether the method step was objectionable under Article 52(4) EPC, though the competence of the practitioner might be, at first sight, a useful indication. Much more important were the purpose and inevitable effect of the feature under consideration (Reasons 5). In the board's view in the case in point, the purpose of the claimed method was merely of a technical nature, with the sole aim of improving the efficiency of taking blood from a donor. The method was clearly distinguishable from a therapeutic or diagnostic effect (Reasons 8). It was therefore not excluded from patentability under Article 52(4) EPC.
T 606/96 (unpublished) of Technical Board 3.3.4 related to a multi-step method for selecting at least one monoclonal antibody component for use in preparing a patient-specific substance for use in in vivo cancer detection or therapy. According to the board, a multi-step process was considered to relate to a method for treatment/diagnosis of the human body if it comprised at least one such step. In the board's view in the case in point, none of the steps was such a step (Reasons 3).
In T 310/99 (unpublished), a decision taken after T 964/99, Technical Board 3.3.8 also applied the principles set out in T 964/99. The case in point featured a claim for a method which provided an indication of the presence of foetal Down syndrome. As the claimed method was practised on a blood sample, not on the human body, and as none of the claims contained a sampling step, the board said that no "diagnostic method" was involved (Reasons 13). Furthermore, according to the board, the method could undoubtedly be carried out by a laboratory assistant without requiring the actual intervention of a physician (Reasons 14). The board thought it irrelevant that the claimed method might be preceded or followed by steps performed by a physician (Reasons 15). It also held the decision in T 385/86 to be inapplicable to the case at issue, as it had been based on a different set of circumstances ("Decision T 385/86 … relates to a different factual framework, the claims examined relating to a medical diagnosis in which not a sample of a body fluid but a whole, intact, living animal or human body is examined (using magnetic resonance). Consequently, decision T 385/86 … is not applicable to the present case"; Reasons 16).
III. The divergence between the principles adopted in T 385/86 and T 964/99 for interpreting the notion of "diagnostic methods"
In each decision the technical board in question worked on the assumption that the policy behind the exclusion of the methods referred to in Article 52(4) EPC from patent protection was that those who carry out such methods as part of the medical treatment of humans or animals should not be inhibited by patents (see T 385/86, Reasons 3.2; T 964/99, Reasons 3.1). In applying this policy, however, the two boards differed in the breadth of their interpretation of the expression "diagnostic methods practised on the human or animal body" within the meaning of Article 52(4) EPC.
1. The term "diagnostic method"
The technical board in T 385/86 construed the term "diagnostic method" very narrowly. The only methods to be excluded from patent protection as diagnostic methods were those whose results immediately made it possible to decide on a particular course of medical treatment. That in turn was only the case if the claimed method contained all steps involved in reaching a medical diagnosis, viz. examination, recording any significant deviation from the normal value, and attributing that deviation to a particular clinical picture. That meant that methods providing only interim results were not diagnostic methods, even if the results could be utilised in making a diagnosis (see T 385/86, Reasons 3.2; see also T 83/87, Reasons 3.2; T 400/87, Reasons 3.1; T 775/92, Reasons 10; T 530/93, Reasons 4; T 1165/97, Reasons 4.3). If the result of the claimed measures was a quantitative value of an isolated physical variable, the decisive issue was whether the measured value on its own made the nature of the disease immediately clear and thus immediately indicated the diagnosis (Reasons 3.4; see also T 629/98, Reasons 3). In T 807/98 it was further ruled that a diagnosis was not only a specific positive identification of a disease, but also the negative finding that a particular disease could be ruled out (Reasons 2.2).
The consequence of this narrow interpretation is that methods containing only one step involved in making a medical diagnosis are not excluded from patentability under Article 52(4) EPC. That applies even if the one step claimed is an essential element of a diagnosis.
The board in T 964/99 expressly observed that adopting the principles set out in T 385/86 would exclude from patentability procedures indicating a more or less complete diagnosis as the result of the fully automated operation of technical devices. On the other hand, however, this would lead to the conclusion that typical diagnostic procedures practised on the human body, such as percussion, auscultation or palpation, could in principle be patentable because they did not constitute a complete diagnosis and also did not fall within the two other medical categories of surgery and therapy referred to in Article 52(4) EPC. However, the board considered that it would go against the spirit of Article 52(4) EPC to interpret its provisions in such a way that "manual procedures" of physical examination essential for making a diagnosis and executed by a medical practitioner did not constitute an exception to patentability (see T 964/99, Reasons 3.5).
In the literature, too, the view has been taken that the interpretation set out in T 385/86 has resulted in a practical dissolution of the exclusion of diagnostic methods from patentability (see Moufang, Medizinische Verfahren im Patentrecht, GRUR Int. 1992, 10 ff, 23; Methods of medical treatment under patent law, IIC 1993, 18 ff; 46 f).
Departing from the interpretation set out in T 385/86, the board in T 964/99 therefore expressly took the view that the expression "diagnostic methods practised on the human or animal body" or the equivalent expressions in the other two official languages should not be considered to relate only to methods containing all the steps involved in reaching a medical diagnosis (see Headnote 1 and Reasons 4.1; emphasis added). In particular, the French text of Article 52(4) EPC did not favour an interpretation limiting the exclusion to methods encompassing all steps required for reaching a medical diagnosis (Reasons 4.2).
Instead, the board was of the opinion that Article 52(4) EPC was meant to exclude from patent protection all methods practised on the human or animal body which related to diagnosis or were of value for the purposes of diagnosis (Reasons 4.4). Consequently, all that was needed to justify exclusion under Article 52(4) EPC was that the claimed method comprised one step which served diagnostic purposes or related to diagnosis and was to be regarded as an essential activity pertaining to diagnosis and practised on the living human or animal body (Reasons 5.1, 5.2 and 6.1; some emphasis added).
As the board in T 964/99 expressly noted, the requirement in T 385/86 that for a method to be diagnostic it had to contain all the steps involved in reaching a medical diagnosis amounted to setting a different standard for diagnostic methods than for methods of surgery or therapy (Reasons 3.6). The latter are excluded from patent protection if they comprise only a single step of a surgical or therapeutic nature (for surgical methods see T 775/97, Reasons 2.5; T 1005/98, unpublished, Reasons 2.4; T 35/99, Reasons 8; T 182/90, OJ EPO 1994, 641, Reasons 2.5.1; for therapeutic methods see T 606/96, Reasons 3; T 820/92, OJ EPO 1995, 113, Reasons 5.5; T 82/93, Reasons 1.4). Accordingly, the presence of a single surgical step in a multi-step method for treatment of the human or animal body confers a surgical character on the method (see T 182/90, Reasons 2.5.1). The policy behind this broad interpretation of surgical and therapeutic methods is that, for a given patient, the optimum or only available treatment could not be administered if even a single part or step thereof were covered by patent protection (see T 35/99, Reasons 7).
In that respect it seems justified to put diagnostic methods on the same footing as surgical and therapeutic methods. A medical practitioner would be equally unable to administer diagnostic treatment to his patient regardless of existing patents if even a single essential step were covered by patent protection.
The general observation to be made is that the expression "diagnostic methods" is construed far more narrowly in T 385/86 than in T 964/99. Consequently, depending on which of the approaches set out in the two decisions is followed, assessment of a claimed method's patentability can lead to different results. If, for example, the interpretive principles set out in T 385/86 were to be applied to the method claimed in T 964/99, there would probably be no "diagnostic method" within the meaning of Article 52(4) EPC, and that provision at least would be no bar to the granting of a European patent. In the method claimed in the main request, which concerned sampling and analysing the substance, the step of attributing the measured deviation to a particular clinical picture would be missing. The method claimed in the auxiliary request, which concerned only sampling a substance or substance metabolite, would also lack the examination phase. On the basis of the results obtained it would not be "immediately ... possible to decide on a particular course of medical treatment" (T 385/86, Reasons 3.2). Furthermore, the method claimed in T 964/99 would probably be held not to constitute a diagnostic method according to the criteria laid down in T 385/86 on the grounds that the extracted bodily fluid is analysed outside the body, ie not on the human or animal body (see also III.2 below).
(a) Diagnostic purpose/relation to diagnosis
According to T 385/86, for a claimed method to qualify as a diagnostic method its results must immediately make it possible to decide on a particular course of medical treatment. For that to be the case, it must contain all the steps involved in reaching a medical diagnosis. If it provides only interim results, and consequently lacks differentiation and comparison operations enabling a pathological deviation to be identified, it is not a "diagnostic method" within the meaning of Article 52(4) EPC. In the board's express view, this applies even if the interim results can be utilised in reaching a diagnosis (thus also T 61/83, T 208/83, T 18/84, T 45/84, T 83/87, T 400/87, T 775/92, T 530/93 and T 1165/97).
Thus the board in T 385/86 equated the expression "diagnostic method" with "diagnosis", in that detecting, distinguishing and identifying a pathological condition and attributing the deviation to a clinical picture have to be essential elements of such a method.
As a result of this narrow interpretation, a diagnostic method denied patent protection under Article 52(4) EPC could be converted to a potentially patentable measurement method essentially by omitting the comparison operation from the claim. With this approach, the issue of whether the measurement method serves diagnostic purposes seems irrelevant.
In contrast, according to T 964/99, Article 52(4) EPC may apply to methods comprising even one step which serves diagnostic purposes or relates to diagnosis and is to be regarded as an essential activity pertaining to diagnosis and practised on the living human or animal body (see also T 606/96, Reasons 3).
However, T 964/99 does not specifically examine whether the step's diagnostic purpose or relation to diagnosis has to be clear from the claims themselves or whether it suffices for this to be clear, explicitly or implicitly, from the application documents as a whole. In the latter case, there remains the question of whether a "diagnostic method" may also be present if the application documents explicitly or implicitly also disclose a non-diagnostic purpose along with the diagnostic one, in other words if the method is not exclusively usable for diagnostic purposes, but also in other industrial fields. The following typical scenarios might arise:
The claim itself could simply define the steps, without more closely defining the purpose of the examination, with the description explicitly listing both diagnostic and non-diagnostic applications. It would also be conceivable that either diagnostic or non-diagnostic applications are discussed and that the other type of application inevitably occurs to the skilled person reading the application, such that the claim in its breadth is supported by the description.
For example, a claim could be directed to a method of optical tomography for examining tissue structures with irregular surface geometry, and the description's examples of applications for the method could expressly include both human tumour diagnosis and remote food inspection. Another example would be a method for determining tube patency (freedom from obstruction), which according to the description can be used to examine transport paths (vascular bundles) in plants, but which a skilled person knows is also suitable for diagnostic applications such as examining blood vessel patency in the living human body for the purpose of diagnosing heart disease. Another case would be where a step which determines whether a diagnostic method is involved is not included in the claim but necessarily takes place beforehand. If administering a substance associated with health risks, eg a contrast agent, is regarded as such a step, a claim of this type might read: "Method for taking a picture of the stomach, to which a contrast agent has been administered …".
In the application underlying T 964/99, which claimed a method of sampling a substance from a body and analysing that substance's concentration, no express reference to a diagnostic purpose is made in the claims. Thus for a method to be deemed diagnostic within the meaning of Article 52(4) EPC the relation to diagnosis does not seem to have to arise directly from the claims. On the other hand, in the board's view all the specific embodiments of the claimed method included in the description concerned activities exercised in the course of a medical treatment of patients and served particularly for diagnostic purposes (see Reasons 5.1). To that extent in the case in point the relation to diagnosis was at least clear from the disclosure of the application as a whole.
A further point that must be clarified is how to assess the situation when no express relation to diagnosis is apparent from the application documents as a whole, but in the skilled person's eyes the claimed method can also be utilised in the field of diagnostics. T 329/94 seems to support the requirement for an explicit relation by saying that, in determining whether the claimed method is a diagnostic method, the purpose of the claimed subject-matter is to be defined in accordance with the patent application, as understood in the light of the description and drawings (Reasons 5; some emphasis added). That would mean that a method that also serves a diagnostic purpose which is not explicitly disclosed in the application but is simply implicitly apparent to the skilled person does not constitute a "diagnostic method" within the meaning of Article 52(4) EPC.
(b) The need for a doctor's involvement
The board in T 385/86, applying the principles it had developed, concluded that the claimed method was not a diagnostic method excluded from patentability pursuant to Article 52(4) EPC because it did not contain all steps involved in reaching a medical diagnosis. Yet it still asked whether, "working on the assumption that the first sentence of Article 52(4) EPC is intended to prevent a doctor from being hampered in the practice of medicine by patent legislation", the claimed method, although not containing all the steps involved in reaching a diagnosis, could still not be considered susceptible of industrial application because it could "only be carried out by a doctor in the exercise of his healing skills" (Reasons 3.5).
Four factors persuaded the board that in the case in point a person skilled in nuclear spin resonance spectroscopy could implement the measures claimed in a commercial laboratory environment without specialist medical knowledge or skills: the effect on living matter was of a technical nature, the interaction of the technical measures with the body did not constitute invasion of the living substance, it did not lead to any permanent changes in the body matter, and it did not leave any harmful side-effects (Reasons 3.5.1; emphasis added). The claimed steps did not include any measures having the character of medical treatment or requiring a doctor to carry them out. In fact it was a technician who, using the method claimed, was able, quite independently, to produce a working basis for the doctor's subsequent activity of diagnosis. Since the examination phase did not require the presence of a doctor who, moreover, had at his disposal a piece of equipment detached from the living body, the measures claimed were susceptible of industrial application, just like laboratory tests carried out on blood or tissue samples taken from the body. A method involving interaction with the human or animal body was susceptible of industrial application if it could be used with the desired result by a technician without specialist medical knowledge and skills (Reasons 3.5.2; emphasis added; see also T 400/87, Reasons 3.2).
Thus the conclusion to be drawn from T 385/86 is that a method comprising at least one step that only a doctor can perform can also be regarded as a diagnostic method. This conclusion is based on a number of criteria: whether the method involves invasion of the living substance; whether it leads to permanent changes in the body matter; or whether it leaves harmful side-effects. Typical steps that today at least only the doctor himself can perform are checking that a patient is suited to a particular diagnostic method and selecting the diagnostic method. Others are examination techniques that are not yet fully automated, such as catheter insertion, biopsies or palpation (where not yet automated), or those (such as auscultation) where the doctor reaches diagnostically relevant conclusions during the examination itself on the basis of his training and experience. Finally, the actual making of the diagnosis in the light of the examination results is itself a step performed exclusively by a doctor, provided that this too has not been automated (for example with the absence of a given technical signal ruling out the presence of a specific disease, see T 807/98).
In cases where not all steps involved in reaching a diagnosis are claimed, it seems on the basis of T 385/86 that the presence of a diagnostic method should be denied even if all steps can be carried out by a technician without medical knowledge or skills or by the patient himself. In keeping with this decision, even the possibility of a technician applying the method appears to rule out application of the Article 52(4) exclusion. The board would thus tend to a narrow interpretation of Article 52(4) EPC, as it repeatedly stressed at other points in the decision (see Reasons 3.2 and 4.1).
According to T 964/99, Article 52(4) EPC is intended to apply if the claimed method comprises a step that is taken for diagnostic purposes and is to be regarded as an essential activity pertaining to diagnosis and practised on the living human or animal body. In arriving at this judgment the board expressly deemed it immaterial that the claimed steps could be performed by the patient himself and that their execution would not have a significant impact on the body or involve a serious health risk (Reasons 6.1). That means that in keeping with T 964/99 methods would have to be regarded as "diagnostic methods" within the meaning of Article 52(4) EPC even if they could be applied by someone with no medical training. By contrast, in T 385/86 the fact that all steps could be performed by people other than doctors meant that the claimed measures did not qualify as a diagnostic method excluded from patentability (see Reasons 3.5.2).
Apart from methods which are "diagnostic" because they contain all the steps involved in making a diagnosis, the only other methods exempt from patent legislation according to T 385/86 are those in which at least one step has to be performed by a doctor. Thus according to T 385/86 the crucial factor in judging whether a "diagnostic method" is present is that the personal involvement of a doctor is required when the method is applied. According to T 964/99, on the other hand, the fact that the method can be carried out by the patient does not stop it qualifying as a "diagnostic method". In accordance with this decision, the area exempt from patent legislation does not seem to be confined to methods which have to be carried out with the involvement of a doctor.
The notion that in assessing whether a method is diagnostic it is immaterial who carries it out would be in keeping with the definition of "therapeutic methods" within the meaning of Article 52(4) EPC. There too the exclusion may apply even if the method can be carried out not only by a doctor but also by someone with no medical training (see T 116/85, OJ EPO 1989, 13, Reasons 4.3; with regard to surgical methods see also T 182/90, Reasons 2.2; T 24/91, OJ EPO 1995, 512, Reasons 2.4).
So according to T 964/99 it does not seem primarily to matter who carries out the claimed method. What really seems significant is that one of the steps is performed for diagnostic purposes and is to be regarded as an essential activity pertaining to diagnosis. In that context the board looks for the presence of a specific step of diagnostic character (Reasons 6.2; emphasis added). In the case in point it deemed the crucial step of a diagnostic nature to be the extraction of a body substance for diagnostic purposes, which was to be considered as constituting an elementary diagnostic activity performed under the ultimate responsibility of a physician (loc. cit.; emphasis added). Earlier the board had referred to the taking of a body sample for the purpose of a medical examination as a "fundamental diagnostic activity", regardless of the technical means used (Reasons 5.2). This approach has since been confirmed in T 807/98. There the crucial factor was deemed to be that the claimed method comprised essential steps which were of a diagnostic character and were to be regarded as fundamental diagnostic activities, because they were performed under the ultimate responsibility of a doctor (Reasons 2.3).
Thus while T 385/86 calls for a further test to establish whether there is at least one step that has to be performed by the doctor himself, T 964/99 instead makes the character of the activity the deciding factor, and the personal presence of the doctor when the method is carried out appears not to be an essential precondition. It seems at least to be sufficient that one step is performed under the responsibility of a doctor. The need for a doctor to be personally involved had already been removed by T 655/92, where the board saw the significant factor as being that the method contained at least one step that could only be implemented by medical staff or under the responsibility of a doctor (Reasons 5.3; emphasis added).
The question that arises in this context is when a doctor can be assumed to have to bear responsibility. One indication might be risks to the patient's health arising from application of the method (see T 655/92, Reasons 5.3; see also the similar approach for surgical methods, T 24/91, Reasons 2.4). The use of technical devices on humans would entail thinking about the values that had to be set, such as the intensity of radiation dosages, possibly defined individually for each patient, or about the part of the body to which to apply the method. Thus in T 807/98, for example, defining the threshold value and the theoretical value on the basis of which a heart dysfunction was detectable was considered to be a step that could only be carried out under a doctor's instructions (see Reasons 2.1).
The decision in T 964/99 might possibly be construed to mean that a "diagnostic method" within the meaning of Article 52(4) EPC may be present even if a doctor does not necessarily have to bear responsibility for any of the steps. That is because the board had also noted that the claimed methods could be performed by the patient himself and that their execution would not have a significant impact on the body or involve a serious health risk (Reasons 6.1). One of the embodiments of the method claimed in T 964/99 mentioned in the description was the measurement of blood glucose levels, which could be performed by diabetes patients for self-monitoring purposes.
In this broad interpretation, the crucial criterion seems to be whether a specific step of diagnostic character is included. As a rule, a step appears to have a diagnostic character if a doctor has to perform it in person or bears responsibility for it. Even if neither of these alternatives applies, a step could still have a diagnostic character.
In the light of T 310/99, however, it is not certain that such an interpretation is valid. In that decision it is not just a question of who is involved in implementing the method. As the claimed activities could "undoubtedly be carried out by a laboratory assistant without requiring the actual intervention of a physician", the method was not a "diagnostic method" (Reasons 14). No test was performed to determine whether a step of diagnostic character was nonetheless present.
It therefore still seems necessary to clarify whether an activity's qualification as having a diagnostic character depends on who is involved or whether that is purely circumstantial to the extent that if the claimed method has to be performed by a physician or under his responsibility, it will normally come under the Article 52(4) exclusion (re surgical methods see T 24/91, Reasons 2.4).
2. The criterion of being "practised on the human or animal body"
In T 385/86, following the narrow interpretation of Article 52(4) EPC, the criterion that the diagnostic method has to be "practised on the human or animal body" was construed to mean that all the steps characterising a diagnostic method had to be performed on the human or animal body itself. In the board's view that meant that "where the facts show a particular subject-matter to be covered by the first sentence of Article 52(4) EPC, both examination (measurement of actual value) and establishing the symptoms on the basis of the examination results - hence the deviation measured from the norm - must be carried out on a living human or animal body" (Reasons 4.1). Consequently, both the actual values measured and the deviation from a norm that had to be regarded as a symptom had to be directly readable from parts of the body or directly discernible on the body itself (Reasons 4.2 and 4.3, emphasis added; see also T 400/87, Reasons 3.3; T 1038/00, Reasons 6).
Typical examples of such methods carried out directly on the body mentioned by the board in T 385/86 were: an allergy test in which the abnormal deviation could be detected from a change to the skin; a method for determining the patency of a body duct whereby liquid was injected into the uterus with a catheter and the pressure build-up in the uterus was observed; a method in which scarlet-fever spots were directly observed or photographed; and an endoscopic examination carried out to ascertain liver damage (Reasons 4.3.1).
The method at issue by contrast resulted in a measured value only visible outside the body in the high-resolution resonance spectrum that appeared on a screen or plotter page in the final stage of the diagnostic apparatus. The measurement data thus obtained was not directly readable from parts of the body, but on a data carrier detached from the body only after further technical measures which took place outside the body. Hence the presence of the body was not required for these further measures, and the board deemed that the "practised on the body" criterion was not satisfied (Reasons 4.2).
Moreover, in the board's view another reason why this criterion was not satisfied was that in the case in point the pathological deviation was not directly discernible on the body itself. The board said it was not sufficient simply that an investigation into the state of a human or animal body was carried out for medical purposes. The condition ascertained had of itself to demonstrate the pathological deviation. Only a comparison and the explicit indication of how great the deviation had to be to be characteristic of a particular disease made the measuring method a diagnostic one. Thus a radiographic examination with X-rays did not make the internal condition discernible until after the X-ray quanta had been converted into visible fluorescent light outside the body and displayed on a screen. Even then, a pathological condition could only be ascertained when the density structure had been compared with normal values (Reasons 4.3.2).
Under the T 385/86 approach, the "on the human or animal body" criterion would presumably likewise not be satisfied by ultrasound tissue imaging, electrical impedance tomography, electrocardiography or electro/magneto-encephalography. In these cases, too, the examination results are not directly readable on the body itself.
According to T 385/86, to satisfy the "on the body" criterion it is essential that both the measurement of values and the identification of a pathological deviation are performed on the living human or animal body itself. All the steps defining a "diagnostic method" must therefore be performed directly on the body itself. Conversely that means that the "on the body" criterion is not satisfied if even a part of the diagnostic method is implemented outside the body being examined. In such a case the exclusion from patentability under Article 52(4) EPC would not apply.
According to T 385/86, the result of the body's interaction with diagnostic examination equipment seems to have to be directly readable on the body. In the eyes of the board in this decision, the intensity or quality of the interaction seems not to be significant in terms of the "on the body" criterion.
If the principles adopted in T 385/86 were applied in assessing the method that was the subject-matter of T 964/99, the criterion of being "practised on the human or animal body" would presumably not have been satisfied. The latter case involved extracting a substance by iontophoresis through the surface tissue of a human or animal body into a sampling chamber and analysing the sampling chamber outside the body to determine the concentration of the substance.
In T 964/99, however, the board observed that the medical art knew of a broad spectrum of diagnostic methods applied by the medical practitioner, ranging from general observations of the appearance of a patient and purely manual interventions to diagnostic techniques utilising sophisticated physical instruments and chemical or biochemical tools. Diagnostic methods could be classified in two categories: those which were practised on the living body and those whose performance took place outside the body. Only methods "practised on the human or animal body" were excluded from patent protection under Article 52(4) EPC, whereas for instance extra-corporal laboratory tests were patentable (Reasons 4.3).
On the issue of when the "practised on the body" criterion is satisfied it may be noted that in T 964/99 a sample was taken from a living human or animal body. In its decision the board further cited percussion, auscultation and palpation as examples of "typical diagnostic procedures practised on the human body" (Reasons 3.5). This might therefore allow the conclusion that the "practised on the body" criterion is in any case satisfied if direct contact with the body is involved.
Yet one might ask whether, to satisfy the "practised on the body" criterion, it might also be sufficient for there to be some other kind of interaction with the living body. One might for example envisage non-invasive procedures that could be performed for measurement and analysis purposes using radiation and that could form the basis for a diagnosis. T 964/99 does not concern itself further with the quality or intensity of an interaction satisfying the "practised on the body" criterion. According to the wording of Article 52(4) EPC ("practised on the human or animal body"), the mere presence of the human or animal body might potentially suffice, so even an assessment of the appearance of the human or animal body could be subsumed under it. This is the interpretation that Board 3.5.1 seems to have applied in T 775/92, as it also classified a remote interaction with the body as a diagnostic method (Reasons 10).
The method that constituted the subject-matter of T 964/99 also included steps that were to be performed both "on the human or animal body" (placing the sampling chamber on the surface of the body, extracting the substance through the surface tissue) and outside the body (analysing the sampling chamber for the concentration of the substance by technical means). The step of taking a sample of a substance from a living human or animal body for the purpose of a medical examination was classified as a "fundamental diagnostic activity" by the board, which regarded this step as an "essential activity pertaining to diagnosis and practised on the living body" (Reasons 5.2 and 6.1). Since the claimed method included that step, it was to be regarded as a "diagnostic method" within the meaning of Article 52(4) EPC.
Thus it is clear from this decision that not all steps need to be performed on the body in order for a method to come under the exclusion from patentability of diagnostic methods practised on the human or animal body under Article 52(4) EPC. In fact it seems to suffice that one such step is performed on the human or animal body. Such an interpretation also seems to be consistent with established case law on surgical and therapeutic methods.
In T 964/99 the step that had "diagnostic character" was also the one that was "practised on the human or animal body". This logically raises the issue of whether this association always has to exist or whether under some circumstances a "diagnostic method practised on the human or animal body" may also be present if, in a multi-step process, the step practised on the body is not the step that relates to diagnosis and constitutes an essential diagnostic activity. In the light of T 807/98, it seems that the step having a "diagnostic character" may also take place outside the body. In that decision, defining the threshold value and defining the theoretical value were held to constitute "essential steps that have diagnostic character and are to be regarded as a fundamental diagnostic activity", as the ultimate responsibility for them rested with a doctor (Reasons 2.3). Defining these values involves activities which are not performed directly on the living human or animal body, but actually on the instrument that is used.
IV. Divergent decisions of two boards of appeal within the meaning of Article 112(1)(b) EPC
According to its wording, Article 112(1)(b) EPC may be applied if "two Boards of Appeal have given different decisions on [a] question".
In the present case the divergent decisions T 385/86 and T 964/99 originate from the same technical board of appeal in terms of the business distribution scheme, ie Board 3.4.1. Nevertheless, they are to be construed as divergent decisions within the meaning of Article 112(1)(b) EPC for the following reasons:
T 385/86 has been followed by other technical boards of appeal in T 775/92, T 530/93, T 1165/97, T 629/98 and T 807/98. These decisions all cite the earlier T 385/86 and adopt the arguments used there. Consequently T 964/99, which has also been followed by another board (in T 310/99), also diverges from decisions of other boards in terms of the business distribution scheme. The divergence has primarily been demonstrated on the basis of T 385/86 and T 964/99, as these two decisions have set out differing approaches to the exclusion from patentability which may lead to different results in a specific case, such as that underlying T 964/99. Furthermore, Technical Board of Appeal 3.4.1 in its decision in T 964/99 expressly distanced itself from the approach adopted in T 385/86 (see Reasons 4.1 in particular).
Furthermore, Technical Board 3.4.1 took its decisions in T 385/86 and T 964/99 in completely different compositions. In the case of divergent decisions of the Legal Board of Appeal, the Enlarged Board admitted a referral by the President in its opinion in G 4/98 (OJ EPO 2001, 131, Reasons 1.2). The decisive factor there was that the effect of Article 112 EPC would be unduly restricted if the President's power of referral were to be defined by a restrictive reading of the terms "two Boards of Appeal" based on organisational structure, as in that case no referrals would be possible with regard to the Legal Board of Appeal, which is one organisational unit only. In the situation at that time the Enlarged Board admitted the referral, as the conflicting decisions had been taken by the Legal Board in different compositions. In the light of the meaning and purpose of Article 112 EPC, in cases where a technical board has taken divergent decisions in totally different compositions there could be a closely comparable need to clarify which interpretation of the law is to prevail. Accordingly, the term "Board of Appeal" in Article 112(1)(b) EPC could be taken to mean more than just a board in terms of the business distribution scheme (see also Schulte, Patentgesetz mit EPÜ, 6th edition 2001, Annex to Section 73, Article 112 EPC, point 42; Günzel in: Benkard, EPÜ, 2000, Article 112, point 5).
The idea behind Article 112(1)(b) EPC is that the President of the EPO should be able to refer a point of law to the Enlarged Board of Appeal if two divergent board decisions have been taken and produce legal uncertainty which needs to be clarified, for example on account of their differing interpretations of the law. In that context it ought to be immaterial whether the decisions have been taken by two different boards in terms of the business distribution scheme or by two boards in different compositions or even by boards in partly or entirely identical compositions (similarly Schulte, Patentgesetz mit EPÜ, 6th edition 2001, Annex to Section 73, Article 112 EPC, point 42; on making a referral's admissibility conditional on the existence of two conflicting decisions and not on whether the decisions were taken by two boards with different organisational names, see also Joos, in: Singer/Stauder, European Patent Convention, 2nd edition 2000, Article 112, point 30). According to its wording, the very purpose of Article 112 EPC is to ensure uniform application of the law (see also Moser, Münchner Gemeinschaftskommentar, volume 20, Article 112, point 28; Paterson, The European Patent System, 2001, 4-175, 187). If board of appeal decisions set out different principles which lead to different results when applied to specific cases, there is a major need on the part of the Office and the public for clarifying which interpretation of the law should prevail. In such cases the need for legal certainty and clarity exists regardless of the board composition in which the differing decisions were taken.
The present case is seemingly not about the continuing development of case law, but about divergent jurisprudence. In T 964/99 the board explicitly stated that it would not apply the principle adopted in T 385/86 (and decisions confirming it); it wished to apply a different principle. It wholly distanced itself from T 385/86's interpretation of the term "diagnostic method" within the meaning of Article 52(4) EPC by using, in Headnote 1 and Reasons 4.1 and 4.2 of T 964/99, almost the same wording as in the definition of the term in T 385/86 in order to give a negative definition. In addition, the board gave examples of procedures that would not be excluded from patentability under Article 52(4) EPC if the principles set out in T 385/86 were adopted, which in its view would go against the spirit of that provision (see T 964/99, Reasons 3.5). The board thus made it quite clear that the term "diagnostic method" was specifically not meant to be understood as stated by the board in T 385/86, and it deliberately dissociated itself from the interpretive principles set out in T 385/86. This goes beyond the bounds of further development of existing case law. Moreover, the divergence between T 385/86 and T 964/99 is clear from the later ruling in T 807/98. In that decision Board of Appeal 3.2.2, in examining whether a diagnostic method excluded from patent protection by Article 52(4) EPC was involved, applied the principles behind the two earlier decisions in parallel (see Reasons 2.2 and 2.3). Such a parallel approach would not have been necessary in a normal case of further development of case law.
Hence T 385/86 (and the decisions which take its line of argument) and T 964/99 are divergent. For reasons of legal certainty and clarity a referral by the President thus appears necessary, regardless of the fact that the differing approaches were adopted by the same board of appeal in terms of the business distribution scheme.
V. Ensuring uniform application of the law and importance of the point of law
According to Article 112 EPC, a referral by the President of the EPO is admissible only in order to ensure uniform application of the law or if an important point of law arises.
Application of the interpretive principles developed in T 385/86 on the one hand and T 964/99 on the other might in a specific case lead to the same result; but it is equally likely that choosing one approach or the other in a specific case might result in a different conclusion over patentability under Article 52(4) EPC. For example, the board in T 964/99 found that a diagnostic method within the meaning of Article 52(4) EPC was involved. Yet if the principles of interpretation set out in T 385/86 were applied to the method claimed in T 964/99, there would be no "diagnostic method" (see also III.1 and III.2 above). So the result would be different. Other examples where applying the two approaches would yield different results are cited in T 964/99 itself (see Reasons 3.5). Thus depending on which approach is adopted, a specific case might involve a "diagnostic method" within the meaning of Article 52(4) EPC and hence an exclusion from patentability, or it might not. Ultimately, therefore, a patent's chances of being granted may depend on which interpretive approach is chosen.
As already mentioned in II.1 above, the Guidelines for Examination in the European Patent Office follow the interpretation set out in T 385/86. Hence Office practice essentially reflects this understanding of the terminology, although examiners may depart from the Guidelines in exceptional cases (see Guidelines, General introduction, 1.2). Thus in the light of the interpretation adopted in T 964/99 and confirmed by subsequent rulings, from the Office's viewpoint there is an urgent need to clarify which approach to the interpretation of the term "diagnostic method" within the meaning of Article 52(4) EPC is to be followed.
For applicants, too, it is crucially important to know which approach is going to be adopted, above all when they are formulating their applications. They need to know which examination criteria will apply to them, as their chances of having their patents granted may ultimately depend on which interpretive approach is followed.
Patent proprietors are equally in need of such clarification, as they may have to reckon with losing their patents in appeal or opposition proceedings. The choice of interpretive approach may determine the chances of success for an opposition or an appeal.
Clarification is likewise in the interests of opponents, who need to be able to assess the cost risk associated with an opposition or with an appeal against its rejection.
Thus clarification by the Enlarged Board of Appeal of how the expression "diagnostic methods practised on the human or animal body" is to be construed within the meaning of Article 52(4) EPC appears to be necessary in order to ensure uniform application of the law.
The issue of which interpretation to follow is also an important point of law. The purpose of the exclusion from patentability under Article 52(4) EPC is to deny patent protection to methods which serve medical purposes, so that no one can be hampered in the practice of medicine by patent legislation. Hence the scope of this exclusion is materially dependent upon the interpretation of the term "diagnostic methods" as used in Article 52(4) EPC. The breadth of this interpretation seems to vary depending on whether the principles applied are those of T 385/86 or those of T 964/99. Accordingly, the board in T 964/99 also found that adopting the principles set out in T 385/86 would lead to the conclusion that typical diagnostic procedures practised on the human body, such as percussion, auscultation or palpation, could in principle be patentable because they did not constitute a complete diagnosis and certainly did not fall within the further medical categories of surgery and therapy referred to in Article 52(4) EPC. In the board's view, however, it would go against the spirit of Article 52(4) EPC to interpret its provisions in such a way that certain "manual procedures" of physical examination essential for making a diagnosis and executed by a medical practitioner did not constitute an exception to patentability (see Reasons 3.5). Hence the interpretation of the notion of diagnostic methods practised on the human or animal body determines the extent of the area exempt from patent legislation on socio-ethical grounds.
The Enlarged Board of Appeal is requested to indicate if it deems further information necessary.
The case is pending under case No. G 1/04.
Ingo KOBER
President