Appendix 1: Notes on Cancer Registration in Canada and the Canadian Datas - This Battle Which I Must Fight: Cancer in Canada's Children and Teenagers

APPENDIX 1

NOTES ON CANCER REGISTRATION IN CANADA AND THE CANADIAN DATAS

Registration of cancer takes place in all provinces and territories of Canada. It is endorsed by both federal and provincial/territorial legislation.(13) Each province or territory operates a registry that attempts to identify all cases of cancer that occur in persons with a permanent residence in the province or territory. Computer records are then sent to Statistics Canada where the records for all of Canada are combined and edited for errors and duplications.

Some children are diagnosed and/or treated in a province neighbouring theirs if the medical facilities there are closer to their home. Registries in provinces west of Quebec routinely notify the province of residence when a cancer diagnosis in a child from outside the province is reported to them. Children from Ontario and the Atlantic provinces, diagnosed and treated in Quebec, and children from Atlantic provinces other than Nova Scotia diagnosed and treated in Halifax have not been reported routinely to the registry in the province where the child lives. New Brunswick estimates that two-thirds of its children and teenagers with cancer are diagnosed and treated in Halifax or Quebec. Underreporting may largely account for the apparently lower incidence rates for New Brunswick. This was corrected partially by using data from the Izaak Walton Killam Children's Hospital in Halifax.

Information on deaths from cancer is obtained for the whole country from the National Mortality Data Base maintained by Statistics Canada. This data base is a compilation of records from provincial/territorial departments of vital statistics. Both the incidence and mortality data bases are used to produce national statistics on cancer.(91) They are also used by Health Canada for detailed analyses.

The data presented in this document were obtained from these national systems and the incidence data were improved by a Health Canada special initiative. The data span the years 1985 to 1992. Because not all registries were able to supply data to 1992, values for these provinces were projected. Projected values for 1992 were made for New Brunswick, Ontario, Quebec and the Northwest Territories, and values for 1991 were projected for New Brunswick. Projections were made separately for the major categories (leukemia and lymphoma) and for their subcategories. Therefore, the numbers in the subcategories do not add to the totals in the major categories.

The cancer registries encode the specific type of cancer using the International Classification of Diseases (ICD). This classification is less suitable for childhood cancers than an alternate system called the Manchester or Marsden Classification. A description of the 12 major categories in the Manchester/Marsden Classification is presented in the Table Legend of Appendix 2. The Manchester/Marsden Classification includes some conditions that are benign according to the International Classification. Because these conditions are not routinely recorded by cancer registries, they are not included in the data in this book. These conditions include some benign brain tumours as well as some of the Langerhans cell histiocytoses (histiocytosis-X). Only acute histiocytosis-X (Letterer Siwe disease) is included in the numbers in this book and it is classified into the category of other reticuloendothelial neoplasms.

Mortality data do not include information on the morphology (cell type) of cancers other than the hematologic malignancies. Because the Manchester/Marsden classification requires this information to classify many childhood cancers, the mortality data cannot be categorized as finely as the incidence data. Consequently, comparisons of incidence and mortality data are restricted to the categories available for mortality. Also, where incidence and mortality rates are compared between children and adults (Figure 6), the data are for the years 1985 to 1991, as incidence data for adults were not available for 1992.

The incidence file, enhanced by the Health Canada initiative, indicates that 4647 cases of cancer were diagnosed between 1985 and 1988, whereas the survival analysis was conducted on 4409 cancers diagnosed in the years 1985 to 1988. This apparent discrepancy is the result of using an earlier source file from Statistics Canada that was linked to the National Mortality Data Base for the survival analysis. Follow-up for the survival analysis was to the end of 1991.

Finally, it is acknowledged that there are a few anomalies in the data. Some of these might be rectified by further data cleaning efforts. More likely, they represent erroneous or incomplete information in the source files from which the registries draw their information, and, therefore, they cannot be corrected. An example of an error that might be traced and corrected is the 14 epithelial cancers reported in infants aged less than one year. Most, if not all of these are likely erroneous and either represent common epithelial cancers of adults whose birth date was incorrectly recorded, or contain incorrect morphology with a more common pediatric site. An example of the consequence of information that is inadequate for accurate coding are the majority of cases in the other leukemia subgroup. Most of these are acute leukemia, not otherwise specified and leukemia, not otherwise specified. These and some others would likely fall into the two main subgroups (acute lymphoblastic and acute nonlymphoblastic), had further information been available.

Further details about cancer registration and the processing and analysis of the data are contained in a companion technical document.

[Previous] [Table of Contents] [Next]