The International Classification of Diseases (ICD) is the descendant of a series of events dating back to the early seventeenth century and the work of John Graunt.
The annual London Bills of Mortality had been established early in the sixteenth century, initially listing only the numbers of burials "as a sort of an early warning system against the onset of bubonic plague." This was not trivial information; the earliest London epidemic of the disease recorded in the "Bills" occurred in 1563 and killed between 20 and 25 percent of the population. By early in the seventeenth century, much additional information had been added, including causes of death. Sometime near the middle of the seventeenth century, John Graunt (1620–1674), a merchant, felt that the Bills contained a wealth of information that was not being used. He tabulated and studied the thirty-two years of data from the annual Bills from 1629 through 1660, and in 1662 he published Natural and Political Observations Made upon the Bills of Mortality. The volume used the Mortality Bills' list of eighty-one causes of death, and is considered the forerunner of today's international mortality classifications.
The next noteworthy step in the history of the classification occurred with the establishment of the General Register Office of England and Wales in 1837 and the appointment of William Farr (1807–1883) as its first statistician. Farr lobbied for an improved classification, and the first International Statistical Congress (ISC), Brussels, 1853, asked Farr and Dr. Marc d'Espine of Geneva to prepare the necessary list of categories. Farr brought back a list based in general on anatomical site, while d'Espine brought a list based on the nature of the disease. Farr's scheme prevailed.
The International Statistical Institute succeeded ISC and in 1891 charged a committee headed by Dr. Jacques Bertillon (1851–1922) to prepare a classification of causes of death. The committee's list was accepted in 1893 at the institute's meeting in Chicago. In 1898, the American Public Health Association recommended the adoption of Bertillon's list in the United States.
In 1900, the French government convened the first International Conference for the Revision of the Bertillon or International Classification of Causes of Death. Delegates from twenty-six countries attended. The list that was adopted had 179 groups of causes of death and an abridged classification of thirty-five groups. This was the first of the ICDs ("International Classification of Diseases"), the initialism that has been applied to the series since 1955 despite slightly modified titles and expanding scope of content. In 1946, the United Nations gave responsibility for ICD to the World Health Organization (WHO), which issued the sixth and succeeding revisions. ICD-6 (1948) included a comprehensive list for morbidity as well as mortality statistics, and saw the establishment of national committees on health and vital statistics throughout the world and increasing worldwide coordination of health statistical activities. The current revision is the International Statistical Classification of Diseases and Related Health Problems, 10th Revision (ICD-10).
Over the years, the number of categories in ICD has grown to about thirteen thousand, in response to the increasing variety of uses to which it has been put—mortality, followed by morbidity, hospital indexing and statistics, reimbursement, public policy, and others. It has become a multi-purpose classification. Yet, by attempting to accommodate each succeeding user it has become less satisfactory for any. Actually, ICD's fundamental purpose of international exchange of mortality and morbidity statistics is for most purposes still answered by the use of "short lists" of one to two hundred broad categories, into which the greater detail for the categories of interest is collapsed. The views of health are thus clearer. Further, many developing nations cannot collect data in much greater detail. The short lists are given at the back of ICD as "Special Tabulation Lists" for mortality and morbidity.
ICD-10 states that "… in the interests of international comparability, no changes should be made in the content (as indicated by the titles) of the three-character categories and the four-character sub-categories of the Tenth Revision … except as authorized by WHO…. WHO should be promptlynotified about the intention to produce translations and adaptations or other ICD -related classifications." In an effort to enforce this position,
ICD-10 was the first of the ICD revisions to be copyrighted.
In the United States in the mid-1950s, ICD-6 (1948) and ICD-7 (1958), with minor subdivisions and modifications, began to be used in hospitals for clinical purposes, initially the indexing of medical records and compilation of hospital statistics. Later, the categories were used for reimbursement and a variety of other demands for information on diagnosed and their treatment. The early modifications were published in 1959 by the U.S. Public Health Service (Publication 719) as International Classification of Diseases Adapted for Hospital Records and Operation Classification, the "Disease Index," which was revised in 1962.
In 1967, WHO published ICD-8. This was followed by competing clinical modifications in the United States, the Hospital Adaptation of ICDA (H-ICDA), published by the Commission on Professional and Hospital Activities (CPHA), a nonprofit corporation, and the U.S. government's Eighth Revision, International Classification of Diseases, Adapted for Use in the United States (ICDA-8). Each of these volumes was used in about half of the country's hospitals.
When ICD-9 was published by WHO in 1977, it again failed to meet the United States' widening clinical needs, and another clinical modification was created jointly by the U.S. National Center for Health Statistics and the Council on Clinical Classifications (a consortium of physician organizations and CPHA): International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM). While ICD-9 had about seven thousand categories, ICD-9-CM has about twelve thousand, the increase being the result of the clinical demand for greater specificity, a demand accommodated primarily by subdivisions that permitted collapsing the detail into the seven thousand of ICD-9's categories when necessary for international statistical purposes.
WHO in 1989 prepared the International Statistical Classification of Diseases and Related Health Problems, 10th revision (ICD-10), which was published in 1992. ICD-10 was put into use for the classification of death certificate information in the United States in 1999, but as of 2000, ICD-9-CM was still being used for all other disease classification purposes.
Although ICD-10 included most of the clinical modifications from the United States' ICD-9-CM, the United States again prepared a modification, ICD-10-CM with about sixty thousand categories. ICD-10-CM has not, at this writing (2000), been put into use. In view of the WHO copyright of ICD-10, the United States had to obtain special permission to create ICD-10-CM.
Neither ICD itself nor its clinical versions is a nomenclature (a list of approved terms) for diseases, although it is sometimes mistakenly referred to as such. All of the versions are classifications— sets of categories into which to place "all" diseases, about one hundred thousand of which are given in the alphabetic index to ICD-9-CM. Detailed description of the organizing principles behind the groupings in ICD is beyond the scope of this essay. It can be noted, however, that in some sections of ICD, grouping is by cause (etiology) as with "Certain Infectious and Parasitic Diseases." The chapter entitled "External Causes of Morbidity and Mortality" does not classify diseases at all. Elsewhere, the results of the external causes can be classified under "Injury, Poisoning, and Certain Other Consequences of External Causes." In some sections, grouping is by physiological systems, such as respiratory and circulatory. Obstetrical conditions form a group. In ICD-10, the classificatory territory has expanded to include a chapter entitled "Symptoms, Signs and Abnormal Clinical and Laboratory Findings, Not Elsewhere Classified" and one entitled "Factors Influencing Health Status and Contact with Health Services." The latter two are not within the usual definition of diseases.
ICD is not truly a classification of diseases (even in those chapters that deal with diagnoses). It is actually a way to group individuals who exhibit the diagnoses—or other "objects" in its universe, such as "need for immunization," rather than just the diagnoses themselves. For this purpose, other attributes of the individual often must be taken into account. For example, "pneumonitis" is, by itself, a diagnosis, but to be classified in ICD, the classifier must put it into one of three different categories, depending on whether it is due to inhalation of food, oil, or some other solid. In some generations of the ICD series, acute myocardial infarction was put in one group if the patient had hypertension and in a different group if not. This distinction was based on an obsolete data management process called "combination coding" (combination coding has been used liberally in ICD-10-CM and is largely responsible for the great increase in categories in that volume). A diagnosis, without further information about the patient, often cannot be properly classified in ICD.
ICD, like all classifications, has its categories arranged in a sequence logical to its author. The categories are often divided into subcategories and even sub-subcategories. For convenience in arranging these categories, they are numbered sequentially, with decimal subdivisions indicating subcategories. It is quite natural to substitute the sorting number of a category for its label—to use the sorting number as a code for the category. If ICD were never modified to reflect changes in knowledge and appearance of new diseases, this "category coding" would present no problems. But consider genetic disorders, which were given 32 categories in the sixth revision, a number which had grown to 709 in the tenth revision (written in 1989, before the human genome developments of the 1990s).
Since ICD itself and its clinical modifications will forever have to respond to these and other influences, the category codes keep changing their meanings, making it impossible to know if a given code means the same thing today as it did yesterday. As a result, longitudinal studies are often impossible, because they must be based on the codes—the exact diagnoses, the "diagnostic entities," which once put into a given category cannot be retrieved. This problem, the inherent impossibility of decoding ICD codes, can only be solved by an information system that tags each code with an identifier as to its source, in the same fashion that the number "0–9615255-2–5" only has meaning if it "tagged" with "ISBN," a tagging that identifies forever a specific book. ICD cannot solve this problem—only a properly designed information system can.
The authors of ICD-9 had "realized that the ICD alone could not cover all the information required and that only a 'family' of disease and health-related classifications would meet the different requirements in public health." They proposed that ICD should be "a 'core classification'…with a series of modules, some hierarchically related and others of a supplementary nature." The authors of ICD-10 followed up on this idea, and ICD-10 diagram of the concept is shown in Figure 1.
Assumptions about the relationships would seem to have been that the hierarchical classifications outside the core would be feeders to the core categories, basically a "parent-child" arrangement. One would then expect that the "specialty-based adaptations" would have greater detail but have the same organization and thus "fit into" the broader categories of the core, ICD. The major and oldest such classification is ICD-O, International Classification of Diseases for Oncology, which is in its second edition (1990). It was written after ICD-10 had been created but before its publication. ICD-O states, however:
There are basic differences between the structure of ICD-O and ICD. Chapter II (Neoplasms) of ICD is basically a topography code that takes into account the behavior of the neoplasm, i.e., malignant, benign, in situ, or uncertain whether malignant or benign … ICD-O has one set of four characters for topography based on [emphasis added] the malignant neoplasm section of ICD-10, and the behavior code, incorporated in the morphology field identifies whether the neoplasm is malignant, benign, etc….
The inference that may be drawn from this is that ICD-O has simply used ICD-10 topography categories as one "module" and has linked to each category a morphology module of its own, so that description of a tumor is a "topography-morphology pair" of codes. However, there are significant departures in the topography categories between the two volumes.
For example, Lymphocytic lymphoma of the stomach is coded C83.0 in ICD-10 but in ICD-O the topography would be coded Stomach C16.9 and the morphology M-9670/3.
In the case of the "supplementary" relationship, the separate classifications would be used to give added information related to core categories as appropriate. For example, when a patient's occupation would be useful to know, that information could be given by an occupation code accompanying the disease code. Many supplementary classifications have been developed by others, some but not all of them other arms of WHO. For example, WHO published a procedure code only once, on a trial basis in 1978: International Classification of Procedures in Medicine. The U.S. Health Care Financing Administration (HCFA) in the mid-1990s commissioned the creation of a procedure coding system which, although entitled ICD-10 Procedure Coding System (ICD-10-PCS), is not a classification and is in no way related to ICD-10.
One of the "family" shown in the diagram that could be considered supplementary is the International Nomenclature of Diseases (IND) series. This built on work begun by the Council for International Organizations of Medical Sciences (CIOMS) in 1970, and joined by WHO in 1975. The stated intention of IND was to provide a single "recommended" name for each "morbid entity." The names were to be, "as much as possible, specific (apply to only one disease), unambiguous, self-descriptive, simple, and based on cause." Each name carried with it a brief definition and list of synonyms, if any. Ten volumes were published in the series, which was suspended in 1992 for lack of funds. Subsequently, WHO has published five volumes described as having "diagnostic definitions."
(SEE ALSO: Classification of Disease; Graunt, John; Notifiable Diseases; World Health Organization)
Bibliography
American Association of Medical Record Librarians (1959). "Efficiency in Hospital Indexing of the Coding Systems of the International Statistical Classification and Standard Nomenclature of Diseases and Operations." Journal of the American Association of Medical Record Librarians 30:95–110, 111, 129.
Commission on Professional and Hospital Activities (1968). Hospital Adaptation of ICDA. Ann Arbor, MI: Author.
Graunt, J. (1662). Observations on the Bills of Mortality. London.
Israel, R. A. (1978). "The International Classification of Diseases: Two Hundred Years of Development." Public Health Report 93:150–152.
Kupka, K. (1978). "The International Classification of Diseases, Ninth Revision." WHO Chronicle 32:219–225.
Moriyama, I. M. (1975). Current Disease Classification and Implication for the Future. Geneva: World Health Organization.
Slee, V. N. (1978). "The International Classification of Diseases: Ninth Revision (ICD-9)." Annals of Internal Medicine 88(3):424–426.
Slee, V. N.; Slee, D. A.; and Schmidt, H. J. (2000). The Endangered Medical Record: Ensuring Its Integrity in the Age of Informatics. St. Paul, MN: Tringa Press.
U.S. Department of Commerce, Bureau of the Census(1931). Manual of the International List of Causes of Death, 4th Decennial Revision, 1929. Washington, DC:U.S. Government Printing Office.
—— (1940). Manual of the International List of Causes of Death, As Adopted for Use in the United States—Fifth Revision. Washington, DC: U.S. Government Printing Office.
—— (1962). International Classification of Diseases, Adapted: For Indexing Hospital Records by Diseases and Operations, PHS Publication 719, Revised. Washington, DC: U.S. Department of Health, Education, and Welfare.
—— (1968). Eighth Revision International Classification of Diseases, Adapted for Use in the United States (ICDA). Washington, DC: Author.
White, K. L. (1985). "Restructuring the International Classification of Diseases: Need for a New Paradigm." The Journal of Family Practice 21:17–20.
World Health Organization (1949). International Statistical Classification of Diseases, Injuries, and Causes of Death. Sixth Revision of the International Lists of Diseases and Causes of Death. Geneva: Author.
—— (1957). International Statistical Classification of Diseases, Injuries, and Causes of Death, 7th Revision, 1955. Geneva: Author.
—— (1967). International Statistical Classification of Diseases, Injuries, and Causes of Death, Eighth Revision(1965). Geneva: Author.
—— (1979). International Statistical Classification of Diseases, Injuries, and Causes of Death, 1975 Revision, (Ninth). Geneva: Author.
—— (1992). International Statistical Classification of Diseases and Related Health Problems, 10th Revision (ICD-10), 3 Volume Set. Geneva: Author.
— VERGIL SLEE
