Elementary concepts of medicine: XI. Illness in a community: morbidity, epidemiology

In clinical medicine the concern is with individual clients one at a time, and more or less independently (despite ‘family medicine’), while in community medicine the concern is with a ‘community’– a defined population – that the physician cares for, in any given instance of care action either this client population as a whole or some subpopulation of this overall population . Whereas the clinician's individual client in the context of any given encounter either does or does not have a particular illness, the counterpart of this in community medicine is a particular level of morbidity in the community at a given time. Our medical dictionaries define the concept of morbidity as: ‘a diseased condition or state; the incidence or prevalence of a disease or of all diseases in a population’ (Dorland 1994); or ‘1. A diseased state. 2. The ratio of sick to well. 3. The frequency of the appearance of complications following a surgical procedure or other treatment’ (Stedman 1995). An epidemiologic dictionary (Last 1995), in turn, specifies: ‘Any departure, subjective or objective, from a state of physiological or psychological well-being.’ These definitions are grossly at variance not only with one another but also with what we meant by morbidity above. While ‘diseased state’ is obviously a clinical concept, we just do not see it as clinical terminology to diagnose ‘morbidity’ in a patient any more than to prevent or certify ‘mortality’ of the patient. We see morbidity to be singularly a concept pertaining to populations, or alternatively, to series of instances of something. It is occurrence of illness in a population or a series, inherently in the quantitative meaning of the ‘level’, frequency of this. It is expressed in terms of a rate of occurrence, indeed as a rate of ‘incidence or prevalence’ but not as a matter of ‘the ratio of sick to well’. And just as a clinician never diagnoses the presence of a totally unspecified illness, so a practitioner of community medicine never quantifies the occurrence rate for ‘all diseases’ in the client population or a subpopulation of this. The definition in the dictionary of epidemiology is, remarkably, even more grossly at variance with the genuine concept of morbidity. In terms of the concepts in the discipline expressly concerned with human populations in general – not epidemiology but demography – there are fundamentally two types of population. The duality is based on how the population's membership is defined. One type of population is a cohort, a closed population. Central to its definition is an event. Membership begins as of the occurrence of this event; and it lasts forever thereafter. Thus there is no exit from membership in a cohort, not even by death; the population is ‘closed’ in this meaning. The other type of population is dynamic, open (to exit). Its definition is based on a state. Membership begins as of the beginning of this state, and it lasts for the duration of this state. The client population in community medicine, just as whatever ‘catchment population’ in clinical medicine, is generally an open population (e.g. with exits by death, for example). All populations are ineluctably in ‘motion’ over time, and therefore an important alternative to consideration of a population as such over time is to consider a series of person-moments from a population, each instance in the series representing a given person at a particular point in time. A morbidity survey of a client population – in ‘community diagnosis’– is a matter of documenting morbidity in a series of ‘arrested’ person-moments from the time course of this population-in-motion. Similarly, a clinician diagnosing the presence of a particular illness ‘in the patient’ does not generally diagnose it in the person without reference to time but, rather, as of a particular point in time – in a particular patient ‘encounter’, in a particular instance of serial encounters, as of a particular person-moment. Prevalence is the occurrence concept that pertains to states of illness. Prevalence rate (empirical) pertains to a series of person-moments. For the series of instances the rate expresses the proportion of the instances such that the illness state is present. Most illnesses, by far, represent a state of ill-health, spanning the time period from the inception of the illness to its termination (in outcome, or in death from a ‘competing cause’); but it is commonplace to speak about the incidence of such illnesses – as though at issue were an event, incidence being the occurrence concept pertaining not to states but to events. To be understood here is that ‘incidence’ in this context refers to the occurrence not of the illness state per se but of its inception – with this event commonly operationalized as the event of its detection (rule-in diagnosis). With highly fatal illness, such as cancer of the pancreas or lungs, the incidence of death from it – the cause-specific mortality rate – serves as a measure of the incidence rate of the illness itself, meaning incidence of its inception/diagnosis. Incidence rates, including those of death from particular illnesses, in community medicine are typically of the form of incidence density. This pertains, different from prevalence rate (above), to an aggregate of population-time: if a population of size N was followed from time t1 to t2, this follow-up covered the aggregate N × (t2 − t1) of population-time. And if a total of n events occurred in the course of this follow-up, then the incidence density was n/N × (t2 − t1). Now, a mortality rate of this form, common in community medicine, must not be confused with the clinically highly relevant case-fatality rate. The latter is the proportion of cases of a particular illness such that the outcome is death from this illness. The concept may or may not be conditional on absence of deaths from other, ‘competing’ causes. For a rate (empirical) of this type, like for the proportion-type rate (of prevalence) above, the referent is not a population but a series of person-moments. It is for a series of instances of the illness that the empirical case-fatality rate expresses the proportion in which fatal outcome turned out to be ‘in store’. Dictionary of epidemiology (Last 1995) gives an untenable definition of case-fatality rate, one that is also at variance with epidemiologic usage of the term. The topic of the occurrence of illness in a population – morbidity – naturally brings up the concept of epidemiology, as it indeed already has above. Its definition in one of our medical dictionaries (Dorland 1994) is: the science concerned with the study of the factors determining and influencing the frequency and distribution of disease, injury, and other health-related events and their causes in a defined human population for the purpose of establishing programs to prevent and control their development and spread. Also, the sum of knowledge gained in such a study. The other two dictionaries (Stedman 1995; Last 1995) give similar definitions, though with ‘the study of’ instead of ‘the science concerned with’ as the opening note specifying the proximate genus. The intended meaning is likely the same, as sciences commonly are defined in terms of what they are ‘the study of’, as indeed is the case in that definition of epidemiology as a ‘science’. Application of Occam's razor – and some other scholarly considerations – to the definition above clarifies it to mean that epidemiology is: The science of morbidity in man. (Definition A) Inherent in this concise form of the definition above is that epidemiology is about the frequency of illness occurrence in human populations, including how this frequency depends on its various determinants, causal and other. This does not mean that epidemiology is about those determinants per se but rather, that it is about occurrence in relation (causal or acausal) to them – occurrence relations in this meaning. A science is not properly defined, even in part, by the purpose of its inquiries; but inherent in the concept of science is indeed not only the aggregate of inquiries into its subject-matter but also ‘the sum of knowledge’ gained by it. To gain perspective on this definition, let us ascertain the definitions of cardiology and morphology in medical dictionaries. Cardiology: ‘the study of the heart and its functions’ (Dorland 1994); or ‘the medical specialty concerned with the diagnosis and treatment of heart disease’ (Stedman 1995). And morphology: ‘the science of the forms and structures of organisms; the form and structure of a particular organism, organ, or part’ (Dorland 1994); or ‘the science concerned with the configuration or the structure of animals and plants’ (Stedman 1995). As for cardiology, then, one dictionary presents it as a study/science, the other as a medical specialty; but in point of fact, the term ‘cardiology’ has a dual meaning: that of a medical specialty (of practice) on one hand, that of a science on the other. It is therefore to be defined in the ‘1. . . . 2. . . . ’ format. ‘Morphology’, by contrast, has a singular meaning; but contrary to those definitions, morphology is not a science. Rather, it constitutes one of the many ‘formal objects’ of interest in the context of any given science-defining ‘material object’, the heart, for example. Material object distinguishes cardiology, as a science, from neuroscience, for example; but both involve morphologic issues, ones of the ‘form and structure’ aspect of the material object. Epidemiology, too, is a dual concept, but not akin to cardiology. The duality echoes one of the two meanings of cardiology and the genuine meaning of morphology. To us epidemiology is: 1. The specialty of medicine (of community medicine, its practice) concerned with morbidity (in a community client). 2. The morbidity aspect of an illness. (Definition B) The misguided idea that epidemiology is a science parallels not only the misconception that morphology is a science (cf. above) but also, more importantly, the fallacy that the practice of modern medicine is science. The latter fallacy is particularly prevalent in respect to epidemiologic practice, in which the pursuit of community gnosis is commonly thought of as a matter of ‘research’. (cf. ‘survey research’ in statistics – a non-topic in science.) Study, in the abstract, of the occurrence aspect of a cardiac illness is obviously a matter of science (Miettinen 2001a), but in cardiology rather than in the ‘science’ of epidemiology, just as study of the morphologic aspect of the heart falls in the science of cardiology, not in the non-existent science of morphology. Epidemiology, like morphology, is not a science because it does not have a coherent material object, different from, say, cardiology. ‘Clinical epidemiology’ is, arguably at least, a contradiction in terms. Clinical practice is oriented to illness in the individual, not morbidity in a population; and while its knowledge base (in gnosis) is about frequencies (Miettinen 1998, 2001b) and thus epidemiologic in nature, clinical deployment of this knowledge is not epidemiology. Just as clinical practice depends on epidemiologic knowledge, so epidemiologic practice depends on clinical knowledge – without the latter adducing the concept of community clinical medicine. In recent decades, the terms ‘epidemiology’ and ‘clinical epidemiology’ have come to use also in reference to the respective disciplines of methodology in health-related occurrence/morbidity – statistical – research in man. The now-prevailing ideas in these disciplines are quite questionable (Miettinen 1999, 2002, 2003). We prefer to think of theory – concepts and principles – of medicine, of its practice first and then, subordinate to this, of directly practice-relevant research (Miettinen 1998, 2001b).

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.