Norms and Standards in Epidemiology:
New PAHO List 6/67 for Tabulation of ICD-10 Mortality Data

The PAHO 6/61 List
Definition of the six broad groups of causes
Preparation of the ICD-10 Short List
Comparison of the PAHO 6/67 (ICD-10) and 6/61 (ICD-9) Lists
Differences between the 6/61and the 6/67 lists

 

When changing from one revision of the International Clasification of Diseases to another, a particular concern is the potential impact of the changes on health statistics (1). This concern is usually taken into account when preparing special tabulation lists, such as those for presentation of mortality data.

PAHO efforts underway concerning development of a short list for mortality tabulation using ICD-10 were mentioned in a previous article (2). In this regard, and as part of a process leading to development of a short ICD-10 list for PAHO use in tabulating mortality data, the Special Program for Health Analysis undertook a review of experience using the 6/61 list for tabulating ICD-9 mortality data (3,4). This was accompanied by a review of the special tabulation lists for mortality recommended in ICD-10 volume 1, as well as lists developed for their own use by the countries and by other international agencies (5). A provisional list was then prepared, submitted for scrutiny by data users and ICD experts, and used for trial tabulation of real data from several different countries of the Americas which are at different stages of development.

The PAHO 6/61 List
Development of the PAHO 6/61 list was conceptualized in 1987, when several countries of the Americas carried out research projects on health profiles/mortality analysis, with technical support from the Health Situation and Trend Assessment Program (now the Special Program for Health Analysis) and under the sponsorship of the PAHO Research Grants Program. The research projects focused on grouping of causes of deaths and preparing short lists for mortality analysis. Upon completion of the projects a Regional Meeting on Guidelines and Procedures for Mortality Analysis was held, in February 1988, and subsequently a summary of the reports of several research projects were published in the PAHO Epidemiological Bulletin (6,7,8).

At the Regional meeting the need to be able to analyze causes of death according to different epidemiological criteria was discussed, and also the fact that the groups of causes of death which are used in the analyses generally reflect the epidemiological criteria applied. Different analytical objectives usually require different groupings of causes, and short lists differ depending, for example, on whether the purpose of the analysis is to generate knowledge, to identify leading causes of death, or to define priorities and orient actions at political and technical levels. It was acknowledged that several lists are probably needed for use at national level, and the recommendation was made that a short list be developed by PAHO for Regional use, to facilitate comparisons between countries.

First used in the 1992 edition of Health Statistics from the Americas, the 6/61 list is used by PAHO for summary presentation of mortality data. Designed specifically for mortality data coded according to the ICD-9, its main purpose was to facilitate inter-country comparisons, and thereby to assist PAHO in undertaking mortality analysis from a Regional perspective. A broad process of consultation with selected analysts and experts in the Region was followed in its preparation, with the hope that the list, as finally adopted, would be useful at country level as well. Ultimately, it was intended to be a PAHO contribution to the countries as a tool for mortality analysis-- a short list that could serve the needs of countries as well as those of PAHO.

Of necessity, however, the 6/61 list reflects the data restrictions imposed by the PAHO mortality questionnaire. This was the instrument used by PAHO for many years for collecting ICD-9 data from the countries of the Region, and is essentially a condensed version of the ICD-9 Basic Tabulation List (9,10). In recent times the questionnaire has gradually been replaced by diskette transmission of data at the level of individual deaths. However, the list of cause groups on the questionnaire, called the "A9 list", is still the maximum level of detail available in the PAHO mortality database.

As shown on Table 1, the 6/61 list consists of six broad groups of causes and 61 more detailed groups nested within those six. Deaths due to signs, symptoms and ill-defined conditions are presented in a separate category that precedes the 6/61 cause groups.

The six broad groups are defined as follows:
- Group l, "Communicable diseases," comprises all infectious and parasitic diseases, i.e. all the categories in ICD-9 chapter I, and in addition, meningitis, acute respiratory infections, and pneumonia and influenza. It should be noted that deaths due to AIDS are not included in this group. A few countries use ICD-9 codes 042-044 for AIDS, and report AIDS deaths to PAHO under these codes. However, a WHO Expert Group recommended use of codes 279.5 and 279.6, and as a result, upon inclusion in the PAHO database all AIDS deaths are (re)assigned to A9 category 18.9, a residual grouping of ICD-9 categories that includes code 279. Most countries of the Americas use ICD-9 codes 279.5 and 279.6 for AIDS deaths, but some use 279.1 or another subcategory of 279.

- Group 2, "Neoplasms," comprises all of ICD-9 chapter II, i.e. malignant neoplasms, benign neoplasms, carcinoma in situ, neoplasms of uncertain behavior and those of unspecified nature.

- Group 3, "Diseases of circulatory system," comprises all of ICD-9 chapter VII, i.e. acute rheumatic fever, chronic rheumatic heart disease, hypertensive disease, ischemic heart disease, diseases of pulmonary circulation and other forms of heart disease, cerebrovascular disease, and "other" diseases of the circulatory system.

- Group 4, "Certain conditions originating in the perinatal period," comprises all categories in chapter XV, i.e. maternal conditions and obstetric complications affecting the fetus or newborn, slow fetal growth, fetal malnutrition and immaturity, birth trauma, hypoxia, asphyxia, other respiratory conditions of fetus or newborn, infections specific to the perinatal period, and other and ill-defined conditions originating in the perinatal period.

- Group 5, "External causes," comprises all of chapter XVII, Code E, i.e., all accidents, suicide, homicide, injury due to legal intervention and operations of war, and injury undetermined whether accidentally or purposely inflicted.

- Group 6, "All other diseases," comprises all other defined causes not included in groups 1 to 4. It is important to note that Group 6, as a whole, is not a residual category; rather, it contains several subgroups that have major public health importance but do not belong in any of groups 1 to 4. Examples are: diabetes mellitus, congenital anomalies, and complications of pregnancy, childbirth and the puerperium.


Preparation of the ICD-10 Short List
In preparing a PAHO short list for tabulation of general mortality under ICD-10 it was clear that several other ICD-10 lists would need to be prepared as well. For example, a second ICD-9 list is already in use for tabulating infant and child mortality, and an ICD-10 version is needed (11). Yet another list is needed for ascertainment of leading causes of death.

The new ICD-10 short list would not be used for every purpose, but it had to be a list that PAHO could recommend to countries for adoption. It also had to serve a fundamental purpose which the 6/61 list already addressed since it was created: as a point of entry to any mortality analysis. The ICD-10 short list was intended, as was the 6/61, to provide a panoramic view of causes of mortality in any country. It was to be compiled at the earliest stages of any mortality analysis undertaken, and it would place causes of death in proper perspective and weight relative to each other. Once this was accomplished, special tabulations for particular analytical purposes would follow.

Several criteria were established in preparing the ICD-10 short list. First, the six broad groups of the 6/61 list would be retained. Second, the groups contained within the broad six would be as similar as possible to those of the 6/61 list, although exact line-by-line equivalences were not imperative and, moreover, might not always be possible in light of changes between ICD Revisions. Third, groups could be dropped or new ones included, based on experience with real data from the countries. Finally, it was assumed that ICD-10 mortality data will be provided by countries to PAHO at 4-character level, hence no restrictions would be placed on the level of detail of the ICD codes making up the groups.

A provisional 6/65 ICD-10 list was circulated to the WHO Collaborating Centers for Classification of Diseases in the Region of the Americas. These are the Brazilian Center for Classification of Diseases (for the ICD in Portuguese), the Venezuelan Center for Classification of Diseases (for the ICD in Spanish), and the North American Center, serving Canada and the United States and located at the U.S. National Center for Health Statistics. A number of experts in different countries of the Americas were also consulted, as well as PAHO staff at Headquarters and in the field. After a final internal review which took into account the feedback received, supplemented by data-based considerations, the 6/67 list was agreed upon and is presented in Table 2. A detailed analysis of the differences between the 6/61 and 6/67 lists follows.


Comparison of the PAHO 6/67 (ICD-10) and 6/61 (ICD-9) Lists

As was the case for the 6/61 list, the 6/67 list was prepared for mortality tabulation. However, the 6/61 list has also been used for tabulation of morbidity, and although less than ideal this will probably continue to happen. In anticipation of this, ICD-10 codes that should not be used for underlying cause of death, which are contained in various chapters, were not excluded. Nonetheless, the two entire chapters (XIX and XXI) that should not be used for coding underlying cause of death were omitted from the 6/67 list.

The 6/67 list, as planned, contains six broad groups of causes; however, it was not possible to attain exact equivalence between all the broad groups of both lists. Two broad groups of the 6/67 list are completely equivalent to the corresponding broad groups of the 6/61 list: 2.00 Neoplasms, and 5.00 External causes. Broad groups 1.00 Communicable diseases, 3.00 Diseases of circulatory system, 4.00 Certain conditions originating in the perinatal period, and 6.00 All other diseases are different as to content. Differences are largely minor, except for a major change that should be highlighted: a new group, 1.07 HIV disease, was established under broad group 1.00 of the 6/67 list. By contrast, list 6/61 placed it in group 6.14 Residual of all other diseases, and PAHO tabulations place AIDS deaths in 6/61 group 6.14, following WHO recommendations that this cause be coded as a subcategory of ICD-9 code 279.

Based on experience using the 6/61 list, several groups that usually had a very small number of deaths were excluded from the 6/67 list. However, several new groups were added, and the ICD-10 list grew to 67 groups nested within the six broad groups. As an intended result, residual groups should contain fewer deaths under list 6/67 than under the 6/61.

The names of the broad groups remained the same, while the names of several groups nested within the broad six were changed slightly in order to maintain consistency with terms used in ICD-10, or to reflect the changed content of the group. Within each broad group, the groups are placed in increasing ICD-10 code sequence.


Those groups for which there is no exact equivalence between the 6/61 and 6/67 lists are discussed below.

a. Certain vector-borne diseases and rabies (6/67, 1.03)
Group 1.03 of the 6/67 list includes dengue, hence it is not equivalent to group 1.04 Certain vector-borne diseases of list 6/61, which does not. On list 6/61, dengue is part of group 1.09 Other infectious and parasitic diseases. "Rabies" was added to the name of the group on the 6/67 list, in recognition of the fact that it is not a vector-borne disease. Rabies as a cause of death is included in this group under both lists.

b. Certain diseases preventable by immunization (6/67, 1.04)
Group 1.04 of the 6/67 list contains the following diseases, in addition to those included in group 1.03 of list 6/61: neonatal tetanus, obstetrical tetanus, rubella (except congenital), viral hepatitis B, and mumps. On list 6/61 these diseases are included in groups 1.09 Other infectious and parasitic diseases, 4.06 Other conditions originating in the perinatal period, and 6.12 Complications of pregnancy, childbirth and the puerperium.

c. Meningitis (6/67, 1.05)
Group 1.05 of the 6/67 list contains all clinical forms of meningococcal disease (such as meningococcal meningitis and meningococcemia) and also viral meningitis, in addition to bacterial meningitis and meningitis of unspecified cause. On the 6/61 list, group 1.07 Meningitis contains only bacterial meningitis and meningitis of unspecified cause.

d. Syphilis and other venereal diseases (6/61, 1.06)
This group does not appear separately on the 6/67 list; rather, it is part of group 1.09 Other infectious and parasitic diseases.

e. HIV disease (AIDS) (6/67, 1.07)
This 6/67 group is a new one. On the 6/61 list, this disease is part of group 6.14 Residual of all other diseases.

f. Other infectious and parasitic diseases (6/67, 1.09)
As a result of the changes mentioned above, this 6/67 group is not equivalent to 6/61 group 1.09.

g. Malignant neoplasm of uterus (6/67, 2.07, 2.08, 2.09)
The three components of list 6/61 group 2.07 Malignant neoplasm of cervix uteri, and of body and unspecified parts of uterus appear separately on list 6/67. These are 6/67 groups 2.07 Malignant neoplasm of cervix uteri, 2.08 Malignant neoplasm of corpus uteri, and 2.09 Malignant neoplasm of uterus, part unspecified.

h. Malignant neoplasm of lymphoid, hematopoietic and related tissue (6/67, 2.12 and 2.13)
List 6/61 group 2.10 Malignant neoplasm of lymphatic and hematopoietic tissue was separated into two groups on the 6/67 list: 2.12 Leukemia and 2.13 Malignant neoplasm of lymphoid, other hematopoietic and related tissue.

i. Pulmonary heart disease, diseases of pulmonary circulation and other forms of heart disease (6/67, 3.04)
This 6/67 group is not equivalent to any on the 6/61 list. Group 3.04 Diseases of pulmonary circulation and other forms of heart disease, of the 6/61 list, was disaggregated into three 6/67 groups. One of them is 3.04 Pulmonary heart disease, diseases of pulmonary circulation and other forms of heart disease; the other two are 3.05 Cardiac arrest and 3.06 Heart failure, which are mentioned below.

j. Cardiac arrest (6/67, 3.05)
This group was not identified separately on the 6/61 list; it was included on list 6/67 to offer users the possibility of interpreting it as an ill-defined cause or keeping it within group 3.00 Diseases of the circulatory system.

k. Heart failure (6/67, 3.06)
This group was not identified separately on the 6/61 list, where this cause of death is part of group 3.04 Diseases of pulmonary circulation and other forms of heart disease.

l. All other diseases of the circulatory system (6/67, 3.09)
This residual group of the 6/67 list is not equivalent to the residual 6/61 group 3.07 Other diseases of circulatory system, because the 6/67 group excludes polyarteritis nodosa. ICD-10 places this cause of death in Chapter XIII Diseases of the musculoskeletal system and connective tissue, which is a part of 6/67 group 6.14 Remainder of diseases.

m. Bacterial sepsis of newborn (6/67, 4.05)
On the 6/61 list this is not a separate group; rather, it is included in 4.06 Remainder of certain conditions originating in the perinatal period.

n. Hemolytic disease of fetus or newborn (6/61, 4.05)
This 6/61 group does not appear as a separate group on the 6/67 list. Instead, it is included in the residual 6/67 group 4.06 Remainder of certain conditions originating in the perinatal period.

o. Remainder of certain conditions originating in the perinatal period (6/67, 4.06)
As a result of the changes mentioned above, this 6/67 group is not equivalent to 6/61 group 4.06.

p. Land transport accidents (6/67, 5.01)
Under ICD-9, motor vehicle traffic accidents are grouped in E810-E819 and appear as group 5.01 of the 6/61 list, whereas under ICD-10 they are scattered among the transport accidents, which are organized quite differently. Under ICD-10 the first axis is the type of vehicle, followed by two other axes: type of accident and person involved, which are not always placed in the same sequence. Use of the group Land transport accidents, which corresponds to an entire range of ICD-10 codes (V01-V89) provides a good approximation to motor vehicle traffic accidents.

q. Other and unspecified transport accidents (6/67, 5.02)
This group refers to water transport accidents, air and space transport accidents, and other and unspecified transport accidents. It is not exactly equivalent to any group on the 6/61 list.

r. Falls (6/67, 5.03)
This 6/67 group excludes fracture, cause unspecified, which was part of 6/61 group 5.05 Accidental falls. Under ICD-10, fracture, cause unspecified does not have an individual code but is assigned to residual category X59 Exposure to unspecified factor. Hence this group is not equivalent to 6/61 group 5.05.

s. Misadventures during medical care, abnormal reactions and late complications, and drugs and medicaments causing adverse effects in therapeutic use (6/61, 5.04)
This 6/61 group does not appear on the 6/67 list as a separate group. Rather, it is part of the residual 6/67 group 5.10 All other accidents.

t. Accidental threats to breathing (6/67, 5.06)
This group did not appear on the 6/61 list.

u. Exposure to electric current (6/67, 5.07)
This group did not appear on the 6/61 list.

v. Accidents caused by machinery, and by cutting and piercing instruments (6/61, 5.08)
This 6/61 group does not appear on the 6/67 list as a separate group. Rather, it is part of the residual 6/67 group 5.10 All other accidents.

w. All other accidents (6/67, 5.10)
This residual 6/67 group is not equivalent to the 6/61 residual group 5.10 Other accidents, including late effects. In view of the fact that different accidental cause groups were established in the 6/61 and 6/67 lists, the residuals are also different.

x. Assault (homicide) (6/67, 5.12)
This 6/67 group is not equivalent to group 5.12 of the 6/61 list, in that the 6/67 group excludes deaths due to legal intervention and operations of war, whereas the 6/61 group includes them.

y. All other external causes (6/67, 5.14)
This group does not appear separately on list 6/61. It includes legal intervention and operations of war, as well as sequelae of external causes.

z. Nutritional deficiencies and nutritional anemias (6/67, 6.02)
This group differs from 6/61 group 6.02 Nutritional deficiencies and anemias, which includes anemias that are not nutritional in nature (hemolytic and aplastic, for example).

aa. Chronic lower respiratory diseases (6/67, 6.05)
This group is not exactly equivalent to any on the 6/61 list. Group 6.05 of the 6/61 list contains only chronic and unspecified bronchitis, emphysema and asthma, whereas group 6.05 of the 6/67 list includes all chronic obstructive pulmonary diseases.

bb. Remainder of diseases of the respiratory system (6/67, 6.06)
This group is not equivalent to any on the 6/61 list. Jointly with 6/67 groups 6.05 Chronic lower respiratory diseases and 1.08 Acute respiratory infections, the three groups make up all of ICD-10 Chapter X Diseases of the respiratory system.

cc. Pregnancy, childbirth and the puerperium (6/67, 6.12)
This 6/67 group is not exactly equivalent to 6/61 group 6.12, partly because the 6/67 list places obstetrical tetanus in group 1.04 Certain diseases preventable by immunization. As is the case with ICD-9, under ICD-10 several causes of death which may be maternal are excluded from the chapter. These include HIV disease; injury, poisoning and certain other consequences of external causes; and mental and behavioral disorders associated with the puerperium.

dd. Remainder of all other diseases (6/67, 6.14)
As a result of the changes mentioned above, this 6/67 group is not equivalent to 6/61 group 6.14 Residual.

NOTES:
(1) PAHO. Revisions of the International Classification of Diseases (ICD-9 and ICD-10): Impact on Health Statistics. Epidemiological Bulletin. Vol. 17, No. 2. Washington, D.C., July 1996.
(2) PAHO. Implementation of the International Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10). Epidemiological Bulletin. Vol. 18, No. 1. Washington, D.C., March 1997.
(3) PAHO. Health Statistics from the Americas, 1992 edition. Washington, D.C., 1992.
(4) PAHO. Health Statistics from the Americas, 1995 edition. Washington, D.C., 1995.
(5) WHO. International Statistical Classification of Diseases and Related Health Problems, Tenth Revision. Vol. 1, pp. 1207-1214. Geneva, 1992.
(6) PAHO. Regional Meeting on Guidelines and Procedures for Mortality Analysis. Epidemiological Bulletin. Vol. 9, No. 2. Washington, D.C., 1988.
(7) PAHO. Health Profiles, Brazil, 1984. Epidemiological Bulletin. Vol. 9, No. 2. Washington, D.C., 1988.
(8) PAHO. Health Profiles, Argentina, 1980-1982. Epidemiological Bulletin. Vol. 9, No. 3. Washington, D.C., 1988.
(9) WHO. International Statistical Classification of Diseases, 1975 Revision. Vol. 1, pp. 746-755. Geneva, 1977.
(10) PAHO. Health Statistics from the Americas, 1998 edition, pp. 464-466. Washington, D.C., 1998.
(11) PAHO. Health Statistics from the Americas, 1998 edition, p. 459. Washington, D.C., 1998.