Proposed Terms

Below the reader will find:

(A) First, a draft list of proposed definitions by Richard Scheuermann, Werner Ceusters, and Barry Smith for the meeting titled: Signs, Symptoms and Findings: First Steps Toward an Ontology of Clinical Phenotypes (September 3-4, 2008).

(B) Then, emails with comments pertaining to the proposed definitions from a few of those who participated in the meeting.

(C) Based in part upon these comments, a paper titled "Toward an Ontological Treatment of the Initiation, Realization, Recognition and Representation of Disease: Some Terminological Proposals" [1] was produced by Richard Scheuermann and Barry Smith and distrubuted to those who participated in the meeting for further comments. So, finally, the reader will find emails with comments pertaining to this paper from a few of those who participated in the meeting.

Draft List of Proposed Definitions

Diseases, Signs and Symptoms: Draft List of Proposed Definitions

We use ‘bodily feature’ to designate biological qualities, processes or structures of an organism such as blond hair, coughing, swelling.

We use ‘clinically normal’ to designate bodily features of a human being that are typically not associated with pain or other feelings of illness, with dysfunction, or with other indications of enhanced morbidity.

We use ‘homeostasis’ to designate the state in which the bodily processes of the organism are regulated in such a way as to (1) maintain bodily features within a certain homeostatic range and (2) respond successfully to departures from this range caused by external influences. During homeostasis the organism continually assesses its current state to determine if its bodily features fall within this range.

Homeostatic Range =def. The range of types of bodily features whose maintenance is continuously sought by an organism in the state of homeostasis.

Normal Homeostasis =def. Homeostasis of a type that is clinically normal for a human being of a given type and age in a given environment.

Abnormal Homeostasis =def. Homeostasis of a type that is not normal.

Disorder =def. A bodily structure in a human being that is clinically abnormal.

Etiological Process =def. A biological process in a human being that leads to a disorder.

Pathological Process =def. A biological process in a human being that is caused by a disorder.

Acute Pathological Process =def. A pathological process terminating with a resolution of the disorder and a return to normal homeostasis.

Acute Disorder =def. A disorder that leads to an acute pathological process.

Chronic Pathological Process =def. A pathological process that results from an adaptation on the part of the patient to a level of abnormal homeostasis.

Chronic Disorder =def. A disorder that, in the absence of intervention, would typically lead to a chronic pathological process.

Progressive Pathological Process =def. A pathological process that deviates increasingly from homeostasis in such a way that the re-establishment of homeostasis is precluded.

Progressive Disorder =def. A disorder that, in the absence of intervention, would lead to a progressive pathological process.

[2] Types of Pathological Process

Physical Examination =def. A sequence of acts of observing eliciting responses, and measuring the bodily features of a patient, occurring in the context of a clinical encounter.

Sign =def. A bodily feature of the patent that is observed in a physical examination and is hypothesized by the clinician to be a disorder or a manifestation of a disorder.

Symptom =def. A quality of the patient that is observed and can be observed only by the patient and is of the type that can be hypothesized by the patient as a manifestation of a disorder.

Laboratory Test =def. A laboratory assay that has as input a specimen derived from the patient, and as output a result that represents a quality of the patient.

Laboratory Finding =def. The representation of a quality of a patient that is the output of a laboratory test.

Clinical Finding =def. A representation of a bodily feature of a patient that is recorded by a clinician because the feature is hypothesized to be of clinical significance.

Clinical Phenotype =def. A constellation of those types of bodily features that are associated with a disorder at each stage of its development.

Clinical Picture =def. A representation of a clinical phenotype as instantiated in a given patient that is inferred from the constellation of laboratory and clinical findings available to the clinician about a given a patient at any given stage.

Diagnosis =def. The conclusion of an interpretive process that has as input a clinical picture of a given patient and as output an assertion to the effect that the patient has a disorder of such and such a type.

Email Comments Based Upon Draft List of Proposed Definitions

From: Xia, Ashley (NIH/NIAID) [E] [3] Sent: Tuesday, September 02, 2008 10:54 AM Subject: RE: Signs Symptoms and Findings Workshop: Definitions Proposed for Discussion

---

Hi Richard

I would like to suggest to add a concept of Homeostatic Profile. The current concept of Homeostatic Range is good for a single measure of a sign, symptom or finding. Homeostatic Profile is good for a collection of homeostatic ranges of a homeostatic state.

Ashley

---

From: Kent Spackman [4] Sent: Tuesday, September 02, 2008 2:00 PM Subject: RE: Signs Symptoms and Findings Workshop: Definitions Proposed for Discussion

---

I haven’t spent much time examining these proposed definitions, but thought I would give you my initial ‘off-the-cuff’ reaction in the form of questions (see attached). There may be good answers to some of these questions – in fact I hope there are. But I suspect I can generate a number of additional difficult questions on further reflection. The general ‘gist’ of the terms is easy to grasp, of course, but I think it would be unwise to underestimate the degree of difficulty of getting good definitions – and getting consensus about the meanings.

I would prefer to back up a step and try to answer two questions: 1) What are the fundamental types of things for which we need ontological categories? (Do we really need to differentiate “signs” from “symptoms”?) 2) What are the criteria by which we can judge whether we have good categories and good definitions? And I’d like to propose one important criterion. It is: the degree to which ordinary clinicians can understand and reproducibly apply the definitions.

Kent Spackman

---

From: Sivaram Arabandi [5] Sent: Tuesday, September 02, 2008 3:00 PM Subject: Re: Signs Symptoms and Findings Workshop: Definitions Proposed for Discussion

---

I can see that the discussion is already getting off the ground and want to add a couple of thoughts. The workshop and the definitions of terms here (like 'Normal Homeostasis', 'Disorder' etc) are focused with the context 'human beings'. However the terms themselves are equally applicable in the more general sense to all animals. It may be useful to provide a broader definition because of the interplay between humans and animals (infectious diseases and their accompanying signs and symptoms - eg. rabies) as well as translational research.

Sivaram

---

From: Anita Burgun-Parenthoine [6] Sent: Tuesday, September 02, 2008 4:14 PM Subject: Re: Signs Symptoms and Findings Workshop: Definitions Proposed for Discussion

---

Thanks for sharing these questions. I have two comments: - phenotypes (as well as observable entities) may be normal. It's interesting to mention that in MPO the synonyms for normal phenotype are 'viable' and 'fertile'.... - As Kent already said, some distinctions are difficult to get. For example, the distinction between chronic disorder and progressive disorder is difficult in practice as many chronic diseases end up with complications. Looking forward to the workshop, Anita

---

From: John Armstrong [7] Sent: Tuesday, September 02, 2008 5:34 PM Subject: RE: Signs Symptoms and Findings Workshop: Definitions Proposed for Discussion

---

Richard,

From the content of the document you distributed involving disorders, findings, signs, symptoms, and processes, I gather that we will be facing some difficult conceptual issues related to classifying subsumption relations for which there do not appear to be intuitive child-parent links (e.g. those relating findings and disorders.) This is a constant challenge we face at Lead Horse Technologies, whether we are browsing SNOMED-CT or developing and editing our own, proprietary ontologies. There are approaches aimed at solving this dilemma, used by us and others, but they often can involve labor intensive curation and constant editing. If it’s not too late, I’d like to propose that topics discussed at the workshop this week include the idea that dilemmas like the one described here may be approached through tying the curation of intraontological relations not to intelligent design but to evolution – that is, linking the curation of subsumption relationships to actual clinical enquiries received from practicing clinicians rather than to the efforts of ontology development professionals such SNOMED-CT editors. This would boil down to applying a wiki-approach to ontology evolution and it is one that we are working on at Lead Horse. Food for discussion, even if only over a glass of wine.

Thanks, and I’m looking forward to the workshop. John

---

From: Colombo Gianluca [8] Sent: Tuesday, September 02, 2008 7:34 PM Subject: R: Signs Symptoms and Findings Workshop: Definitions Proposed for Discussion

---

Dear all, it is a pleasure to contribute to the discussion.

We would like to bring up two general points, and then proceed to a more detailed discussion of the proposed definitions.

First of all, a remarkable aspect of the diagnostic activity is that certain observations are performed, in order to reconstruct the state of a patient; the state itself is not fully/directly observable (e.g. diabetes cannot be observed per se); the observables collected during the diagnostic activity are interpreted by the clinician, according to some explicit theory and/or his experience, in order to fully reconstruct such state. We therefore deem important to disinguish between what is observed and what is inferred/reconstructed. Also the methodology/istrumentation for the observation, and the criteria for the inference/reconstruction play an importat role. Second, looks like that the notion of disorder plays a pivotal role. A disorder must ncessarily be an anomaly of structure. What is the rationale behind this choice?

Best, Gianluca Colombo and Daniele Merico

HERE ARE OUR COMMENTS:

Signs, Symptoms and Findings: Draft List of Definitions Proposed for Discussion

We use ‘bodily feature’ to designate biological qualities, processes or structures of an organism such as blond hair, coughing, swelling.

We use ‘clinically normal’ to designate bodily features of a human being that are typically not associated with pain or other feelings of illness, with dysfunction, or with other indications of enhanced morbidity. How is dysfunction defined?

We use ‘homeostasis’ to designate the state in which the bodily processes of the organism are regulated in such a way as to (1) maintain bodily features within a certain homeostatic range and (2) respond successfully to departures from this range caused by external influences. During homeostasis the organism continually assesses its current state to determine if its bodily features fall within this range.

Homeostatic Range =def. The range of types of bodily features whose maintenance is continuously sought by an organism in the state of homeostasis.

Normal Homeostasis =def. Homeostasis of a type that is clinically normal for a human being of a given type and age in a given environment.

Abnormal Homeostasis =def. Homeostasis of a type that is not normal.

Disorder =def. A bodily structure in a human being that is clinically abnormal. 1) Why a structure, and not a quality or process? What is the rationale? 2) A disorder can be both a cause and an effect? From the poin of view of etiology, it aperas to be neuter.

Etiological Process =def. A biological process in a human being that leads to a disorder. Why introducing the notion of etiological process but not of etiological factor?

Pathological Process =def. A biological process in a human being that is caused by a disorder. 1) Why is a pathological process interesting? Is it the connection between the disorder and the symptoms/signs, or the patient’s malaise, or morbidity? 2) A biological process ensuing a disorder may be physiological rather than pathological (e.g. wound healing), if pathological implicitly means “associated with pain or other feelings of illness, with dysfunction, or with other indications of enhanced morbidity” (quoted from clinically normal).

Acute Pathological Process =def. A pathological process terminating with a resolution of the disorder and a return to normal homeostasis.

Acute Disorder =def. A disorder that leads to an acute pathological process.

Chronic Pathological Process =def. A pathological process that results from an adaptation on the part of the patient to a level of abnormal homeostasis.

Chronic Disorder =def. A disorder that, in the absence of intervention, would typically lead to a chronic pathological process.

Progressive Pathological Process =def. A pathological process that deviates increasingly from homeostasis in such a way that the re-establishment of homeostasis is precluded.

Progressive Disorder =def. A disorder that, in the absence of intervention, would lead to a progressive pathological process.

Physical Examination =def. A sequence of acts of observing eliciting responses, and measuring the bodily features of a patient, occurring in the context of a clinical encounter.

Sign =def. A bodily feature of the patent that is observed in a physical examination and is hypothesized by the clinician to be a disorder or a manifestation of a disorder.

Symptom =def. A quality of the patient that is observed and can be observed only by the patient and is of the type that can be hypothesized by the patient as a manifestation of a disorder.

Laboratory Test =def. A laboratory assay that has as input a specimen derived from the patient, and as output a result that represents a quality of the patient.

Laboratory Finding =def. The representation of a quality of a patient that is the output of a laboratory test.

Clinical Finding =def. A representation of a bodily feature of a patient that is recorded by a clinician because the feature is hypothesized to be of clinical significance.

Clinical Phenotype =def. A constellation of those types of bodily features that are associated with a disorder at each stage of its development. 1) Is there a relevance criterion in the association? Since the Clinical Phenotype refers to bodily feature, any feature may be included. 2) A bodily feature can be a structure; hence a structure can be a clinical phenotype; in other ontologies/works, it is usually preferred to define a phenotype as a property bore by a structure (f. Bard et al., Nature Reviews Genetics 2004).

Clinical Picture =def. A representation of a clinical phenotype as instantiated in a given patient that is inferred from the constellation of laboratory and clinical findings available to the clinician about a given a patient at any given stage. Why are Clinical Picture and Clinical Phenotype two distinct concepts? Why is only the former related to Laboratory/Clinical Findings and not simply Bodily Features? Does that imply being observable (directly, or by means of instrumentation) in antithesis to being reconstructed/inferred? Or is the difference related to significance according to the clinician? Indeed, Clinical Findings must be of clinical significance for the clinician (but not Laboratory Findings).

Diagnosis =def. The conclusion of an interpretive process that has as input a clinical picture of a given patient and as output an assertion to the effect that the patient has a disorder of such and such a type. If a disorder is neuter with respect to causes and effects, is a diagnosis neuter to causes and effects as well? Declining this definition on a specific example, is Diabetes Mellitus Type-I a diagnosis? Is Diabetes Mellitus Type-I a disorder? Or is rather the absence of insulin-producing beta cells of the pancreas the disorder? Or is it the absence of insulin the disorder? Or is it the abnormally high level of glucose in the blood? Glucose is a blood component, and thus is a bodily structure.

---

From: goldberg@buffalo.edu [9] Sent: Wednesday, September 03, 2008 8:44 AM Subject: Re: Signs Symptoms and Findings Workshop: Definitions Proposed for Discussion

---

Homeostatic and homeostasis are old and venerable terms and everyone more or less knows what they mean but they are misleading. None of the states they refer to are static in any way. Alternate terms that have been suggested are homeodynamic and homeokinetic.

---

From: Zhangzhi Hu [10] Sent: Wednesday, September 03, 2008 9:42 AM Subject: Re: Signs Symptoms and Findings Workshop: Definitions Proposed for Discussion

---

Just throw in my 2 cents:

- We should add "Etiology" besides "Etiological Process", just like there is Homeostasis as opposed to "homeostatic process" (already a GO term, GO:0042592)".

- We also should add Prognosis as opposed to Diagnosis.

- Perhaps we also consider Progression and Remission...

- Also there seems to be already a Symptoms Ontology, but not sure about its content:

http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?tp=&arnumber=1647635&isnumber=34552

Zhangzhi

---

From: Sivaram Arabandi sivaram.arabandi@gmail.com Sent: Fri 09/05/08 10:07 AM Subject: Fwd: Re: FYI: An attempt to integrate draft list of definitions into an existing medical record

---

I too found the workshop very stimulating and ended up doing a lot of background reading on my entire flight back to Cleveland! I thoroughly enjoyed the discussions and the consensus approach to resolving the challenges. Here are a few thoughts:1. I feel that 'clinical-finding' and those that are clinically significant i.e. 'clinically-significant-finding' are not one and the same - instead they have a subclass relationship.

2. Common to encouraging adoption of new products/technologies, be it in the patient-centric side or in the population-centric research/studies side, is the fact that the users need to have control of the data i.e. data portability. This is an area where using Ontologies to provide common, reusable models and RDF for data persistence provides the separation between the application layer and the data layer and give the user ownership of data. This is a problematic area for current EMRs and once a user/practice/hospital starts using a specific implementation of EMR, they are essentially locked down.

3. I particularly liked the idea of breaking down the concepts behind symptoms and signs into their component parts - what was found, who found it, how was it found etc.. to define them at a very granular level. These can then be used, globally or locally, to define what one considers as symptoms or signs in a flexible way. 4. One thing that I was not sure of was 'What are the boundaries for the Phenotype ontology?' - how far should we go in identifying the symptoms and signs and lab findings and associating them with the disorders/diseases? Is this the place to define 'these are the clinical findings associated with a particular disorder'?

Sivaram

---

On Sep 4, 2008, at 11:49 PM, Ogbuji, Chimezie wrote:

Thanks for the excellent workshop. I thoroughly enjoyed the discussions and found them very insightful. I was sufficiently encouraged and motivated that I spent the 2 hours in-flight re-factoring the Computer-based Patient Record ontology I've been slowly developing, over a period of over a year, to incorporate the terms and definitions from the draft list distributed to the participants. The exercise was was very insightful for me and in the end, the combination of the definitions and the conversations we had about them helped fill certain holes that I have had for some time in the ontology. The updated OWL file is attached for anyone interested. I defined a namespace for the terms that I re-used from the term list. In the absence of a better option, I used the workshop URL as the basis for this namespace: http://bioontology.org/wiki/index.php/DallasWorkshop

Below are the terms I incorporated, the others either already had an existing term in the ontology from the prior version or I ran into difficulty incorporating them: pheno:bodily-feature pheno:aggregate-bodily-feature pheno:clinical-phenotype pheno:laboratory-test-finding pheno:abnormal-homeostasis pheno:normal-homeostatis

Embedded within the OWL assertions were a few editorial comments that identify some interesting issues I ran into during the re-factoring process: For pheno:bodily-feature, the criteria that restricts instances of this defined class to only include those that are 'of' an organism was not axiomatized since, I was not sure what relation, specifically, what role restriction, would be relevant to represent this criteria. For pheno:clinical-phenotype, the criteria that there is an association with a disorder was not axiomatized since I was not sure what relation would be relevant as a general, permanent association between a constellation of phenotypes and a disorder. For clinical-finding, the criteria regarding the clinician's hypothesis of clinical significance is also not axiomatized. This seemed (to me) part of a more difficult problem of capturing modal logic, which would seem necessary to represent beliefs. I have also included a screen shot of the classes from within Protege to show where they fell in the larger framework into which they were re-factored. Thanks again, for a wonderful workshop!

Chimezie (chee-meh) Ogbuji

Email Comments Based Upon "Toward an Ontological Treatment..." Paper

---

From: "Hogan, William R" <hoganwr@upmc.edu> Sent: 10/14/2008 Subject: RE: Follow-up Information Regarding the Dallas Workshop on Clinical Phenotypes

---

Bill: 1. Under examples of acquired genetic disease, you cite malignant colon cancer. However, a physical object like a tumor is not a disposition. Thus, I would have thought malignant colon cancer to be an acquired genetic disorder, not an acquired genetic disease, the disease being the disposition toward unregulated cell division and the lumps of cells that result being another disorder that can lead to other dispositions (e.g., disposition to abnormal colonic transport).

Barry's Response: We (actually me, since I didn't check this with Richard yet) view the cancer as the disposition, the disorder is whatever it is in the cell(s) in virtue of which they have this specific disposition toward unregulated cell division of this specific sort; the tumor is another disorder which is a consequence of the realization of the disease. Not every disorder serves as physical base for a disease; the tumor in question does not (typically) do so. (Does it ever?)

Bill: 2. I am not sure that your definition of infection rules out normal flora, colonization, or carrier status. I know that you inserted the word "pathogenic", but the situation I am concerned about is that normal flora (at least in humans) are a common source of bacterial infections.

Barry's Response: E.G., normal flora invade the weakened lung of smokers; but then they are not normal -- because they are in the wrong place.

Bill: One common usage of 'pathogen' is to refer to types of organisms some of whose instances cause infectious diseases in humans.

Barry: One common usage of 'drug' is to refer to types of chemical substances some of whose instances are packaged, sold and administered to patients. Penicillin is a drug in this sense. But it is a loose sense, which a good ontology needs to guard itself against. E.g. because there are non-drug penicillin varieties (wild type penicillin expandase?)

Bill: Thus, the definition of "infection" as currently stated would apply to every human, and I don't find such a definition particularly helpful.

Barry: because every human has many pathogens within his or her interior? If so, I think we are safe, since we say that there is infection only if the pathogen contributes to a disorder, and a disorder (we say) has to be clinically signifant.

Bill: I think that the terms pathogenic/pathogen therefore may require some explication (I don't know whether you have a page limit), and obviously care will be required not to be circular (i.e., can't define pathogen as organism that causes infection or infectious disease).

Barry: We are bending over backwards not to be circular.

Bill: Or rather the definition of infection could be modified slightly.

Attempt 1: pathogen = organism that disrupts normal homeostasis in its host organism, or leads to abnormal homeostasis. Not sure this one covers latency such as with Zoster virus. Or perhaps such latency does not qualify as infection because the virus is not pathogenic at that time. Pathogenicity according to this definition would be a quality of an instance of an organism that it may acquire/lose.

Barry: To talk of pathogens is to talk of organisms playing certain roles. To make this clear replace this with:

Attempt 1*: pathogen = organism that is within a host organism and either is disrupting normal homeostasis or ADD CLAUSE REGARDING LATENCY.

Bill:

Attempt 2: pathogen = organism with strong disposition to cause morbidity in a particular host. As opposed to weak disposition? I think attempt #1 with the idea that the latency of zoster viruses in ganglion cells does not equal infection is better.

Attempt 3: redefine infection: "...presence of a pathogenic organism -- in either a clinically abnormal quantity or location or both -- within the host organism..." It has to be clinically abnormal because for example, any bacteria in the urine of a pregnant woman requires treatment, but you only need to treat >100K CFUs (colony-forming units) in the urine of a non-pregnant woman. In this case, the Zoster latency qualifies as infection (which is my preference).

Barry: I believe the issue here is covered by our present definitions.

Bill: 3. Definition of sign: I am somewhat uncomfortable with its restriction to the physical exam (unless you define physical examination as including any imaging study). For example, see the following link for an example of an "echocardiographic sign": http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=481116

Barry: If Richard agrees I would be happy to expand 'sign' to allow image-associated signs

Bill: 4. Definition of symptom: I have difficulty with the "can be observed only by the patient" aspect of it. For example, we may say the patient has a symptom (which you define as a type of quality) of fever or rash, and the clinician can subsequently observe those things (qualities) also.

Barry: I would be happy to drop this.

Bill: 5. Definition of physical exam: perhaps say "observing and measuring" as opposed to just "measuring".

Barry: Fine with me.

Bill: 6. Definition of clinical finding: I would include laboratory finding in the scope of this definition.

Barry: Surely the laboratory findings can at least include much more, e.g., at the molecular level, that would not be classed as clinical?

Bill: 7. Possibly missing term - physical exam finding: representation of a sign (as sign is currently defined), or representation or output of a physical examination.

Barry: Fine with me.

Bill: 8. Clinical manifestation: I would add at the end of the definition "...including the signs and symptoms themselves."

Barry: Logically redundant, in fact, but I suppose I could go along with it.

Bill: 9. Clinical phenotype: I would change "bodily qualities" to "bodily features" in the definition. It seems to encompass both structures and their qualities, which is what the defined class of "bodily feature" as stated earlier in the paper comprehends.

Barry: We hope to clean up this whole aspect in the next version.

Bill: 10. Discussion: the diagnostic process is an iterative one. The clinician is forming hypotheses during history taking, testing them during additional history taking, forming new hypotheses a result, testing those hypotheses during physical exam, forming new hypotheses as a result, etc. It is such a canon of how medicine is taught that exclusion of the iterative nature of the diagnostic process from this paper may alienate clinicians. However, I understand that including it in this paper is not straightforward and perhaps not even necessary to the purposes of the paper. However, simply paying homage to the iterative nature of the diagnostic process--including that the response to a particular treatment is often itself a diagnostic sign--in a single sentence may be sufficient to avoid rejection of the rest of the paper due to its exclusion.

Barry: I agree.