On the inadequacy of the concept of disease

Medical Hypotheses 9:

ON THE INADEQUACY OF THE

169-177, 1982

CONCEPT

OF

DISEASE

C.Zajicek.The Hubert H.Humphrey Center for Experimental Medicine and Cancer Research.The Hebrew University-Hadassah Medical School P.O.B. 1172.Jerusalem,Israel 91010. ABSTRACT A framework for a critical examniation of the disease concept is presented.The human organism is regarded as a complex system upon which certain hierarchies are imposed.The cell is regarded in this framework as the lowest hierarchy.Cells are aggregated into tissues and tissues into organs.Organs are assembled into organ systems constituting the human organism.The elementary indivisible unit in each hierarchy is called the HOLON. When viewed from the hierarchy below it encompasses a continuum.The holon is defined by its attributes.Its STATE represents the magnitudes of its attributes.The state of the organism is defined by the states of its holons.For each attribute a criterion level (or state) defines its ILLNESS STATE. ILLNESS is defined by the set of attributes crossing the criterion level.These attributes are divided into two two sets: Attributes requiring treatment by the physician and those not requiring intervention by him.The first is defined as DISEASE, while the second as the VIS MEDICATRIX NATURAE. THERAPY is the restoration of "ill" attributes below their criterion levels.The process of selecting the DISEASE attribute set from all organismic attributes is called DIAGNOSIS. The correctness of a diagnosis is testable by treatment resu1ts.A successful treatment corroborates the diagnosis and vice versa.Since an intractable disease is also untreatable,its diagnosis is not testable (or falsifiable) in Popper's sense.The framework provides also means to test the relevancy of cancer theories to human cancer.

Philosophy of Medicine, System's theory, Cancer, Neoplasia.

INTRODUCTION The concept of DISEASE occupies in Medicine a fundamental role.Its definition however is still vague : a "Morbid condition of the body" (l).The renewed interest in the definition of disease and the nature of the diagnostic process was initiated by the desire to apply computers to clinical diagnosis.These attempts calling for the construction of a logical framework for Medicine, were only partially successful.Feinstein applied Boolean

169

Algebra to define disease (2).Sneath and Sokal introduced Taxonomy to medicine (31, and Card tried to define a Calculus for medicine (4).These and other efforts did not succeed in disease uniquely,and the defining computer,when interrogated generally probable list of provides a diagnoses.Only in restricted disease sets is a conclusive automatic diagnosis feasible (5). The main obstacle for creating a logical Medicine resides in the ambiguous disease concept.Engle and Davis (6) noted that *At the present time there is concept of disease . . . In diagnosing no unified different diseases.we often different types of fundamental criteria”, and according to use entirely Scadding (7j.a “unified concept of disease is virtually impossible and the first step of dispelling confusion is to recognize that the concept “a Disease” is logically complex”. And indeed Medicine continues being an art,utilizing rational logic padded with intuition,a mixture which had served it adequately for ages.Yet the same kind of logic had to be banned from Physics in order to pave the way for *Relativity theory”. The transition from the intuitively sound Euc1idean.t.o the non intuitive yet strictly logical non-Euclidean universe was accompanied by the rejection of intuition.Could it be that a similar fate awaits Medicine, which in order to become a science “logically complex” disease concept for a non has to trade its intuitive intuitive yet logically simple one? The tools for such a transition already exist and will be applied here for the creation of a logical framework in such as Disease ,Diagnosis which concepts and Cure may uniquely be defined,while the inadequacy of the “logically complex” disease concept is clearly illustrated. THE HIERARCHIALSTRUCTURE OF THE HUMAN ORGANISM The organism is regarded here as a complex system upon which certain (8).The cell is regarded in this framework as the hierarchies are imposed into tissues and tissues into lowest hierarchy.Cells are aggregated organs.Organs are assembled into organ systems constituting the human organism.The latter is regarded here as the highest hierarchy.A detailed description of this framework has been provided previously (8).In each unit is defined: when viewed from the hierarchy hierarchy an indivisible below it encompasses another continuum.This unit has been named by Koestler as HOLON(9).The attributes of each holon are regarded as its coordinates so that a holon with n-attributes may be regarded as a point in an n-dimensional represented by a vector with n-comp0nents.A set of holons defines a space STRUCTURE,while the change of a structure or part of it with time is defined as a PROCESS(or FUNCTION). In the cell hierarchy for instance we may distinguish between three cell cells.All three cell and functional types : Stem cells,transit-progenitors originate in the first-Each stem cell division generates two cells,one types of which generally replaces the stem cell to remain such.The other continues transit-progenitors gradually to multiply.generating a cell clone of differentiationg.This genealogy when taken as a whole is defined as tissue (81. THE EVOLUTIONOF MEDICINE When viewed holons,structures traversing along

from

below,the

and processes,yet

an n-dimensional

organism encompasses it is from above holon trajectory.This

170

continuum of a regarded as a holon was the first object

of ancient medicine when the first diseases had been defined.Among all the structures and processes the physician probably observed in his patient,some or even death .These were the first were associated with a change of conduct recognized diseases.From there on medical thinking was shaped by the development of means to detect structural and functional deviations in the patient. The evolution of the disease concept may thus be conveniently expressed in the the realm of our hierarchies.At each stage, disease definition extended over deeper and deeper hierarchies,involving more and more holons.The disease was defined initially in EDEMA for instance terms of the highest hierarchy.When it was recognized that edema may accompany ailments of different organs (or organ systems) these holons were included in its definition e.g. Cardiac-edema or Kidney-edema. Gradually the organ system concept took over a leading role in disease definition,while EDEMAreceived the label of a “Sign”.The distinction between allergic and uremic belonging the the Tissues or edemas,introduced into its definition arguments Cell hierarchies.When it was recognized that allergy is associated with antibodies or gamma globulins,elements of the Sub-Cellular hierarchy had to be introduced into the disease definition as well. Scientific progress thus continuously moulds the disease concept,infusing from deeper and deeper hierarchies.The it more and more structures into studied also by other scientific disciplines e.g. Physics or latter are Chemistry.so that when introduced into Medicine,these structures were actually wrapped in concepts originating in them.These concepts were never Each continues truly incorporated into a consistent theory of Medicine. contributing to the logical complexity of the disease existing independently, concept. It seems of advantage to classify medical knowledge describing the into two categories: ILLNESS and DISEASE.The first includes all the organism necessary to DESCRIBE the patient’s condition or STATE while information DISEASE is regarded here solely as a conceptual tool to RECOGNIZEthe patient’s STATE.The concept of ILLNESS will be studied first, and only then,the concept of DISEASE. THE STATE OF THE ORGANISM Since the organism is extremely complex,only a small subset of its In each case attributes is generally studied. it may be represented by The various attributes fluctuate different holons and holon attributes. continuously,and the observed attribute magnitude is defined here as its ATTRIBUTE-STATE.The HOLON-STATE is defined by the states of attributes,and the HIERARCHY-STATE,by its holon states and so on. In order to define the STATE OF THE ORGANISMthe holons and their attributes which are to represent it have to be selected first.Once their magnitudes are determined the over all state is known.The state of the organism at a given moment is defined as its CONFIGURATION. Since the selection of the attribute set to represent an organism is utterly arbitrary,two states are COMPARABLE only if being defined by the same attributes or coordinates. hierarchy Medicine distinguishes between two basic In the highest states.Health and Illness.The first represents a reference state in relation is defined.Since in our framework Health and Illness are to which Illness defined in terms of holon attributes we ought to define a measure evolution of an illness more and more distinguishing between them.During the

171

attributes deviate from the healthy state.The deviation will now be expressed by a DISTANCEMEASURE‘d’. Suppose the organism’s system is defined by one attribute,its temperature ‘T’.A one holon system with one coordinate.T0=36.8 degrees, will be defined as a reference state.In order to measure illness progression in terms of the measured temperature,‘d’ will be defined as an absolute difference between two temperature measurements.Suppose the patient’s temperature rises to dlz0.7 deg.= T1=37.5 deg., we shall say that our system reached the distance from the reference state.In another patient T2= 38 deg. and 136.8-37.51 d2z1.2 deg.= here 136.8-38.01. ‘d’ serves several purposes.First it quantitates morbidity.The second patient for instance is by 0.5 deg. more morbid than the first. *d’ expresses also the SIMILARITY degree between two states.Let d3=0.5=138.0-37.51 be the distance between two patient Is states.Sicce d3

=Tc is regarded as ill.Generally for each attribute patient exhibitting indicating the illness state has to be defined.The c(i) a(i) .a criterion ‘d’ may now be applied to express the distance of an attribute a(i) distance smaller than c(i) ;a(i)
illness stage.The Koplik spot in Measles for instance precedes the skin rash so that the gingiva attribute is a more sensitive illness indicator than the skin attribute.In each hierarchy we could now rank attributes according to their sensitivity to indicate illness.Obviously the better controlled an attribute,the less sensitive it is. a drug mild blood Suppose ingested daily induces a red cell destruction.The increased blood loss is adequately replenished by the bone marrow so that the erythrocyte count remains normal.Bilirubin excess is easily cleared by the liver.Both attributes thus do not cross their respective criterion states,anemia and bilirubinemia, and the patient appears “healthy”. On the other hand clearance of radioactive chromium tagged red blood cells,clearly indicates a system change.Even if hemolysis intensifies as long as it is compensated by the bone marrow the red blood cell will remain constant .Anemia ensues only when the bone marrow fails to replenish the missing blood, and that may be relatively late in terms of the progressing illness .Anemia is thus a less sensitive illness indicator than chromium tagging. ILLNESS AND DISEASE The restoration of an attribute below its criterion level is defined here as TREATMENT.Weshall distinguish in an illness between attributes that do and do not require treatment by the physician.A broken arm for instance is manifested by many attribute changes e.g. humerus discontinuity,pain,sweating and palpitation. It suffices however to repair the first in order to restore the other below their criterion levels.Among all the attributes defining an illness we shall distinguish a subset called DISEASE,including only those attributes requiring intervention by the physician.The BROKENARMdisease is defined by one attribute,HUMERUS DISCONTINUITY,while the BROKENARMillness includes the disease as changes well as not requiring all treatmect,accompanying it.The process of selecting the DISEASE ATTRRIBUTEset is defined here as DIAGNOSIS. While examining an ill patient the physician assesses the atributes requiring active treatment‘this set is defined as the patient’s Disease.It contains only attributes which will and may be actively treated. An attribute which by the available means is untreatable is NOT INCLUDEDin the DISEASE set.If the diagnosis is correct, treatment will ultimately restore all ill attributes to their pre-illness however treatment state.If has failed,obviously the diagnosis was wrong and the physician has to redefine the DISEASE set.By selecting only attributes requiring intervention,the physician implicitly assumes that all attributes excluded from the disease set will be restored by the organism.This self healing capacity of the organism is regarded here as the modern representation of the “vis medicatrix naturae” (lO),and the set of all the NONDISEASE attributes in an illness is defined herewith as the VIS HEDICATRIXset.The logical complexity of today’s disease concept is attributed here to the failure of modern Medicine to distinguish between Disease and Illness (or Illness state),not realizing that the first is essentially an approximation of the latter. DIAGNOSIS The act of representing an illness by a disease is defined here as Diagnosis.Diagnosis formulation should be regarded as equivalent to the formulation of a hypothesis, and since each disease indicates the attributes 173

requiring treatment,this hypothesis is testable or FALSIFIABLE by the treatment resu1ts.A successful treatment corroborates the hypothesis,while its failure rejects it.We may now apply Popper’s reasoning (11) in distinguishing between falsifiable and falsifiable diseases (or non hypotheses1.A BROKENARMwould represent the first,while an incurable illness the second.Suppose the broken arm of a patient fails to heal with the usual means.Treatment failure clearly rejects the diagnosis.During reexamciation a silent metastasis may have been detected.Since this attribute requires treatment,the attribute METASTASIS has to be included in the disease definition: HUMERUSFRACTURE ACCOMPANIEDBY A METASTASISand indeed if a local x-ray treatment contributes to fracture healing,the diagnosis is further that another patient with a broken arm verified.Let us suppose carried an undetected silent metastasis which however did not prevent bone reunion, the initial diagnosis HUMERUS FRACTUREwill still be correct.Even if with systemic it occurs in a patient metastatic spread,as long as the fracture heals properly, the presence and absence of a metstasis there is irrelevant to the diagnosis. Such an approach reminds one of the problem oriented patient record (12) according to which the attributes requiring treatment are grouped into Problems which have to be corrected.Weed’s method however still distinguishes between Problems and Diagnosis,while here the Problem is the Diagnosis. It may seem difficult to accept treatment failure as the sole indicator for a wrong diagnosis.Many diseases are still refractory to treatment and life may not be prolonged at will.In some diseases e.g. sickle cell anemia,in spite of the known molecular defficiency underlying it, the disease is still untreatable,so that it appears as if sickle cell anemia is diagnosable even if untreatable.The last statement however does not apply to the above defined DISEASE concept which is based solely on treatable attributes.The mere understanding of the molecular deficiency,contributes only to the ILLNESS specification.In order to be included in the DISEASE definition,it has to be also treatable. The duality by which the VIS MEDICATRIXis defined,first as a set of attributes and then as a healing factor,is deeply rooted in Medicine.Whenever examines his patient and evaluates his reserves,he implicitly a physician measures the potency of the VIS. By isolating the attributes requiring treatment (and thus defining a Disease) he actually leaves the other attributes to her.By diagnosing a SIMPLE FRACTUREin a patient harbouring systemic metastasis.the physician implicitly assumes that the VIS will take care of all the attributes influencing the healing fracture.Since in the organism all attributes interact.the metastases obviously affect the healing process somehow.As long as the fracture heals properly the VIS is assumed here to contribute sucessfully to this process.If on the other hand the SIMPLE FRACTURE did not heal because the patient was diabetic,the diagnosis is wrong for two reasons.First because the Disease set did not include the DIABETES attributes,and second,because the VIS has been wrongly assigned to correct an attribute she is not capable of correcting.Thus in order to properly diagnose a patient,the physician has to assess the VIS potency,an evaluation to be regarded here as the modern interpretation of Hippocratic PROGNOSIS. In the present context PROGNOSIS is not a “forecast of course of a disease” (lj).but the means to assess the chances of a patient to withstand illness (or a noxa) (10). The severer an illness the less reliable the VIS.More and more attributes require active treatment and the Prognosis becomes poorer.In an intractable illness there is hardly an attribute which 174

may be left to the VIS, so that the disease is defined on a nearly infinite attribute require intetvention,a task which is set.All attributes active beyond the physician’s power.And since treatment may not be realized,the diagnosis of an intractable disease is not testable.Practically this would mean that before losing hope,the physician should keeping reexamining his diagnosis (or hypothesis1 searching for mistakes (or uncorrected attributes). THE PROGRESSION OF ILLNESS Since every illness evolves.its attributes continuously change.Some cross the criterion level while others return to their initial state.Illness progression shapes also the Disease set which may change at any given moment.The physician has continuously to update his disease set leaving the rest to the vis.In many diseases e.g. the BROKENARM ,the disease set remains invariant in time and it appears as if one diagnosis suffices to define illness along its entire course.Cenerally the fewer attributes contribute to the disease definition,the greater the likelihood of it presenting itself as invariant.The apparent constancy of the observed attributes depends further upon the scale in which the attribute changes are expressed.@lNeuralgiall or “Angina” for instance are measured by the nominal scale consisting of two levels,they may either exist or not.so that the one attribute disease hierarchy NEURALGIA is essentially constant.By incorporating deeper attributes,the disease concept had gradually adopted different measurement scales. Ancient medicine was dominated by the nominal scale.Later,diseases which could be ranked into categories e.g. hypo- or hyper-tension or stupor and coma,were measured by the ordinal scale,while most modern disease attributes ,generally involving physical and chemical units,are measured by the interval scale. The three scales shaped many of the medical theories. Some may be regarded as fossil remnants of ancient medical concepts,yet unlike true fossils remaining confined in hidden strata.the medical fossils permeate modern medicine from genuine modern medical and are hard to distinguish concepts.Take for instance the attribute couples “normal-pathological**or, “health-disease” originating in the “nominall’ medical era,when diseases were conceived as invariant .When applied to attributes measured by the interval scale *these concepts lose any meaning.And indeed they are foreign to the exact sciences.Who ever heard of a “pathological” hydrogen bond or a llnormal*l gravity? It is true that in many diseases these concepts are extremely convenient ,yet in others they turn out to be dangerous.Psychiatry for instance continues to classify individuals into normal and disturbed, a distinction which in some countries led to monstrous consequences. This conceptual difficulty initiated a widespread search for the definition of “normality” on the interval scale.Take for instance the density distribution of the red blood cell count in the healthy males. It exhibits an inverted bell shape known as Gaussian,which unfortunatrelly has recieved the adjective 11norma181.Themean of this density curve is 5 million cells per cubic m.m.Suppose a patient exhibits a value of 5.5 million cells per cubic m.m.,how should he be regarded,ill or healthy (or normal)? The dilemma was thought to be solved by deciding that if the observed value exceeds 5.0 + 2 standard deviations, the patient is said to be ill, while if the observed value is is less than the criterion. he is regarded as healthy. It seems best that concepts e.g. normal be banned once for ever.The state of the above patient does not depend upon his position in the Gaussian curve,but how the whether he needs active Disease set is defined.i.e. treatment or 175

if all the other system attributes are below their criterion not.Obviously if suffering from a terminal car level,he may not need any treatment.yet erythrocyte count no matter which position in the pulmonale. any excessive Gaussian curve it occupies, needs treatment. Standard deviations,or percentiles are extremely useful for selecting high a procedure which is only distantly related to the above risk individuals, Disease definition. Effect should also be regarded as fossil remanants from the Cause acd infection CAUSED by fftumor ,rubor ,calor and era.Yesterday “Nominal sealen it TRIGGERSit.Inflammation is triggered also by functio laesa” yet today chemical or physical injury as well as by ischemia.Yesterday we distinguished between tubercle.foreign body and Aschoff granulomata,all assumed to be CAUSED by specific pathogens.To day they represent the same mechanism EVOKED of triggers.In many diseases “Cause-Effect” arguments by a whole spectrum still suffice,but they ought always to be viewed as rough approximation to the complexity of the organism in which all attrributes continuously interact (10). The failure of modern medicine to distinguish between Disease and Illness sense is highlighted by the concepts of a LOCALor isolated in the present disease.Medicine still regards the HUMERUS FRACTUREas strictly local.Such an in this mono-attribute disease is obviously sufficient since it is attitude corroborated by the treatment sucess,yet may be extremely dangerous in a non treatable disease e.g. Cancer in which neoplasia is also regarded as strictly is obviously not supported by modern cancer treatment local.This hypothesis of the local neoplasm fails to cure it (14).and yet the since excision parasite” constitutes the corner stone of concept of the “localized Oncology.Since treatment does not corroborate the assumption of cancer being a local phenomenon,the Disease set obviously lacks some attributes which be treated and is therefore improperly defined. Being ought to intractable.Cancer has to be defined on an infinite attribute set so that it (15).In other words instead of focussing is essentially a systemic disease upon the local neoplasm, Oncology ought to reexamine all systemic attributes for the neccessary clues. ON THE RELEVANT MODELFOR HUMANCANCER In a previously published study,hierarchy theory served to assess the relevancy of experimental models to human cancer (81. The present arguments provide different means for a similar purpose.Suppose a phenomenon observed is assumed to reproduce human cancer.In order to compare both in an animal systems they ought to be defined by the same attributes.The relevancy of a theory to human cancer is assessable if and only if both are defined by equal the Crown-Gall tumor of the plant attributes.Many modern cancer models e.g. fail to meet even this elementary requirement.More precisely, they do meet it if defined on a too restricted attribute set.If defined *for instance *only by one attribute TUMOR.the Crown Gall disease and cancer represent the same phenomenon. theories were formulated during the “highest hierarchy” era Many cancer of Medicine defined above as the %ominalll,when the disease name “Cancer” or was coined.These fossil theories continued to evolve,changing shapes by crab concepts the Mutation lower hierarchy Somatic adopting e.g. theory,representing a modern offspring of the “evil crab” concept sprinkled In the present context in order to check a hypothesis with Darwinian ideas. all the attributes it has to be treated as a DISEASE whose set includes 176

mentioned by the theory.If for instance one attribute is not treatable the theory is not testable.The somatic mutation belongs to this category. Its inclusion in the Cancer Disease set does not contribute to its cure, so that in its proposed form the theory is non falsifiable. Nearly all modern cancer theories suffer from the same inadequacy.The virus theory of cancer started as a falsifiable one.It was assumed that if cancer was induced by a virus,and the latter isolated,it could be cured by a vaccice.The theory could thus be tested by the treatment results.In its modern form,known as the Oncogene theory,it is no more testable. The theory proposes that each cell actually harbours oncogene sequences.which become activated to transform a cell into a malignant one. In the way it is formulated,the oncogene theory may contribute to the Cancer ILLNESS definition yet is not testable as such in the realm of clinical medicine .In order to be testable it has to be defined in form of a DISEASE set and tested by its curative potency. The immunological theory of cancer also started as a testable hypothesis aiming to detect a tumor antigen,with the aid of which the host could be immunized against his neoplasm.Such an antigen has never been found so that the Disease set IMMUNOLOGICAL CANCER was defined by an untreatable attribute and could not be tested.Its modern version known as Immunological Surveillance theory is also not testable.Like the Oncogene theory it may contribute to the Cancer ILLNESS definiton,not to the Cancer DISEASE. These arguments reiterate in an orderly manner the known fact that all modern cancer theories have unfortunatelly failed and are no better than any other fringe theory labelled by the medical community as quackery. REFERENCES 1. 2. 3.

4. 5. 6.

Fouler HW,Fowler FC,Eds. The concise oxford dictionary.Oxford University press p.349 1967. Feinstein AR, Clinical Judgenent.Williams L Wilkis Co. Baltimore 1967. Sneath PHA,Sokal R. Numarical Taxonomy. WH Freeman and Co. San Francisco 1973. Card WJ Computer assisted diagnosis and pattern recognition. Proc.R.Soc. London B 184 : 421,1973. DeDombal FT. Computer aided diagnosis of abdomonal pain.In: Biomedical Computing Ed Perkins WJ .University Park Press Baltimore p.35. 1977. Engle RL,Davis BJ. Medical diagnosis past and future. Arch.Ict.Med.

7.

112:512,1963. Scadding JG. 2:877,1967.

a.

Zajicek

G.

Diagnosis: On

the

The

clinician

relevant model for

and human

the

computer.Lancet

cancer.

Med.Hypoth.

7:il39,1981.

9. 10.

Koestler A. Beyond atomism and holism,the concept of the Persp.Biol.Med. 13:131.1970. Zajicek G. The fundamental principle of medicine. Med.Hypoth.

holon. 6:665.

1980.

11. 12. 13. 14. 15.

Popper K. Unended Quest. Fontana-Collins London p.131 1976. LL. Medical records,medical education and patient care. Western reserve University Press Cleveland, 1969. Fouler HW,Fowler FG. ibid p.976. Zajicek G. ‘The risk of cancer therapy. Med.Hypoth. 6:727,1980. Zajicek G. Cancer as a systemic disease. Med.Hypoth. 4:193.1978. Weed

177

Case