Canadian Cancer Registry (CCR)

Status:
Active
Frequency:
Annual
Record number:
3207

This is an administrative database that collects information on cancer incidence from all provincial and territorial cancer registries in Canada.

Detailed information for 2010

Data release - October 4, 2012 (first in a series of releases. Please refer to left sidebar under the heading "The Daily")

Description

In 1992, the person-oriented Canadian Cancer Registry (CCR) evolved from the event-oriented National Cancer Incidence Reporting System established in 1969. The CCR is an administrative database. Beginning with cases diagnosed in 1992, cancer incidence data collected by provincial and territorial cancer registries (PTCRs) have been reported to the CCR, which is maintained by Statistics Canada.

The CCR is a collaboration between the 13 Canadian PTCRs and the Health Statistics Division of Statistics Canada, where the data are maintained. Ultimate authority and responsibility for the completeness and the quality of the data reside with the provinces and territories. The data that comes into the CCR describes both the individual with the cancer, and the characteristics of the cancer.

The primary objective of the CCR is to provide a national database of information that can be used to produce standardized and comparable statistics for cancer incidence and survival data for each primary type of cancer. This information is used to assist and support health planners and decision-makers at all levels of government to: identify risk factors for cancer; plan, monitor and evaluate a broad range of cancer control programs (e.g., screening); and, conduct research in health services and economics.

The CCR is a dynamic database of all Canadian residents, alive or dead, who have been diagnosed with cancer since 1992. The CCR is a patient-based system that records the type and number (incidence) of primary cancers diagnosed for each person until death. Subsequent primary cancers occurring for patients who are already in the database are linked to their existing patient information. The advantage of this system is that longitudinal data is available for each cancer patient. The patient data is regularly linked to mortality data (death clearance) to optimize the accuracy of date, cause, and place of death fields in the CCR and to identify potential primary cancers not currently registered in the CCR. Since patients' records remain active on the CCR until confirmation of their death, survival rates for the various forms of cancer can be calculated.

In addition, the CCR regularly employs internal record linkage to identify and eliminate duplicate patient and tumour records. Conflicting information and other problems with the data uncovered by CCR processes are resolved through consultation with the reporting provinces and territories.

For more details, see "History of the Canadian Cancer Registry" in the Documentation section.

Reference period:
Calendar year
Collection period:
Fifteen months after the end of the reference period.

Subjects

  • Diseases and health conditions
  • Health
  • Health care services
  • Life expectancy and deaths

Data sources and methodology

Target population

The conceptual universe of the Canadian Cancer Registry database is persons whose usual place of residence is Canada or who are non-permanent residents. The residence at diagnosis is generally the place of usual residence, as stated by the patient or, as defined by the Census, "the dwelling in Canada where a person lives most of the time", regardless of where they are when diagnosed. For patients diagnosed with multiple primary tumours over time, the address may be different for each primary tumour.

Sampling

This survey is a census.

Data are collected for all units of the target population, therefore no sampling is done.

Data sources

Data are extracted from administrative files.

The Canadian Cancer Registry (CCR) is comprised of administrative data. Each provincial and territorial cancer registry (PTCR) supplies data on cancer patients and tumours in a standard format, by encrypted file through a secure electronic transfer process. Each year, approximately 180,000 new primary tumour records are loaded into the CCR database which is maintained at Statistics Canada. Subsequent changes to registry data due to errors or omissions found in the data are also transmitted to Statistics Canada. Each PTCR can send in records to add, update, and delete records. The following three software modules are used to build and maintain the data in the CCR database: Core Edit; Internal Record Linkage; and Death Clearance. Internal record linkages and death clearances are performed regularly on the complete CCR.

The Canadian Council of Cancer Registries (CCCR) recommends that the following tumours should be reported to the CCR:

- All primary, malignant tumours (topographies C00.0-C80.9 and behaviour codes of 3);

- All carcinoma in situ/intraepithelial/noninfiltrating/noninvasive tumours (behaviour codes of 2); except cervix (topographies C53.0-C53.9) and prostate (topography C61.9);

- All borderline malignancies (behaviour codes of 1);

- Primary, benign tumours of the meninges, brain, spinal cord, cranial nerves and other parts of the central nervous system (topographies C70.0-C72.9 with behaviour codes of 0);

- Primary, benign tumours of the pituitary gland, craniopharyngeal duct and pineal gland (topographies C75.1, C75.2, C75.3 with behaviour codes of 0, for 2007 and later);

- EXCEPT certain non-melanoma skin cancers (topographies C44.0-C44.9 with behaviour codes of 1, 2, 3).

See the Canadian Cancer Registry System Guide for details of scope inclusions and exclusions. To obtain copies of the Canadian Cancer Registry System Guide, please contact Client Services 613-951-1746 (hd-ds@statcan.gc.ca), Health Statistics Division.

Patient records include nominal, demographic and mortality information about the person. Tumour records contain information about the characteristics of the tumour and its diagnosis. The list of data elements on the Patient record and the Tumour record is provided in the Canadian Cancer Registry System Guide.

Error detection

Each cancer registry supplies information to the Canadian Cancer Registry (CCR) for each new patient and tumour in a standard format. Validity edits verify that each field contains valid entries. Correlation edits check the compatibility of different data elements within a record and verify that relationships between patient and tumour records make sense. These verifications are done as data are loaded into the CCR database and any records failing edits are rejected and returned to the provincial or territorial cancer registries for verification and/or correction.

The CCR also performs internal record linkage to detect duplicate records, and a death clearance process to optimize the accuracy of death-related information and maximize completeness of case-ascertainment. These procedures are done regularly. Conflicting information and data issues uncovered by these processes are resolved through consultation with the reporting provinces and territories. The CCR also has edits that ensure the accuracy and consistency of data in all provincial and territorial cancer registries.

Imputation

Statistics Canada does not impute missing data on cancer records.

Quality evaluation

Statistics Canada produces Data Quality Reports that provide detailed feedback to the provincial and territorial cancer registries (PTCRs) on the quality of data submitted each year. The Report provides data quality indicators for all patient and tumour records reported in a given data year.

CCR Data Quality Reports are used by PTCRs to monitor the quality of their information in the CCR and meet standards for acceptance in international publications such as Cancer Incidence in Five Continents and Cancer Incidence in North America.

Before the availability of cancer statistics is announced in The Daily, the publication data files are reviewed and approved for official release by each PTCR.

Disclosure control

Statistics Canada is prohibited by law from releasing any information it collects which could identify any person, business, or organization, unless consent has been given by the respondent or as permitted by the Statistics Act. Various confidentiality rules are applied to all data prior to release or publication, such as suppression of confidential variables in microdata and suppression, random rounding or controlled rounding of tabular data.

In regards to random rounding of tabular data, the Canadian Cancer Registry (CCR) rounds each cell count, independently of other cells, to an adjacent multiple of 5 using an unbiased random rounding scheme. Specifically, numbers ending in 0 or 5 are not rounded; numbers ending in 1 or 6 are rounded up with a probability of 0.20 and down with a probability of 0.80; numbers ending in 2 or 7 are rounded up and down with probabilities of 0.40 and 0.60, respectively; numbers ending in 3 or 8 are rounded up and down with probabilities of 0.60 and 0.40, respectively; and, numbers ending in 4 or 9 are rounded up and down with probabilities of 0.80 and 0.20, respectively. Thus, counts are more likely to be rounded to their nearest multiple of five but table additivity will be maintained only by chance. Controlled rounding to an adjacent multiple of 5 is more complex to implement than random rounding because linear programming techniques are used to round table entries while attempting to maintain table additivity.

When calculating crude or specific rates and confidence intervals, the CCR uses the rounded count. Otherwise, the actual count can be deciphered relatively easily by using publicly available population estimates. Because age-standardized rates are more complex to calculate, involving the combining of data across multiple age strata, the CCR uses the actual age-specific counts to calculate the age-standardized rate and confidence intervals. However, the overall count of events contributing to the age-standardized rate is rounded when presented. For the special circumstance where the rounded overall count of events contributing to the age-standardized rate is zero, the actual age-standardized rate and corresponding confidence intervals are suppressed to maintain the ambiguity of zeros.

Data accuracy

Completeness

Since each Canadian province and territory has a legislated responsibility for cancer collection and control, case ascertainment of malignant tumours is considered relatively good. Each provincial and territorial cancer registry (PTCR) identifies tumours in its population by combining information from sources such as: cancer clinic files, radiotherapy and haematology reports; records from in-patient hospital stays, out-patient clinics, and private hospitals; pathology and other laboratory/autopsy reports; radiology and screening program reports; reports from physicians in private practice; medical billing and hospital discharge administrative databases; and, reports on cancer deaths from Vital Statistics registrars. Some PTCRs experience problems in submitting data to Statistics Canada on a timely basis or updating the CCR to reflect the current content of their PTCR.

Designed studies are used by the PTCRs to assess the completeness of case ascertainment (check for under and over-registration of cases). Re-abstraction is used to evaluate accuracy and completeness of reported items.

To avoid duplication, the CCR accepts tumour records only when the reporting PTCR is the same as the province or territory of residence at the time of tumour diagnosis. Each PTCR is required to return records for residents of other jurisdictions to the appropriate PTCR. Under-coverage remains a stronger concern than over-coverage because of the following reasons: some PTCRs do not use, or have had periods in the past where they have not used, death certificates as a source of tumour incidence; differing definitions among the PTCRs of what constitutes a reportable or malignant tumour; difficulty in diagnosing certain tumours because of their location (or site) in the body; differences among the PTCRs in coding practices, data entry or processing procedures; and, failure to report tumours diagnosed and treated in a province, territory or country outside of the province or territory of residence.

For determining multiple primaries with tumours diagnosed as of January 1, 2007 onward, the CCR adopted the Multiple Primary and Histology Coding Rules from the Surveillance Epidemiology and End Results (SEER) program of the National Cancer Institute (NCI) in the United States. Differences in incidence between the years 2006 and 2007 may be partly attributable to this change. As well, due to technical issues in updating systems to adhere to these new rules, not all registries were able to report according to the new requirements for 2007. This may result in inconsistencies between provinces and territories.

Documentation