Contact Info
Cancer Technical Notes
Data Sources
Cancer incidence data
Minnesota cancer incidence statistics are based on data from the Minnesota Cancer Reporting System (MCRS). The 1987 Minnesota Legislature established the MCRS (formerly the Minnesota Cancer Surveillance System) to assure that accurate, complete, and timely data would be available to inform planning and decision-making at the local, state, and national levels as well as to foster research into the causes of different cancers.
Starting in 1988, MCRS gathered data on nearly all microscopically confirmed malignant and in situ cancers diagnosed in Minnesota residents. In 2012, after a data collection rule change, data on both clinical and microscopically confirmed cancers were gathered. Cancers excluded from reporting include the most common forms of skin cancer (basal and squamous cell carcinomas) and in situ carcinomas of the cervix. These exclusions are consistent with guidelines for cancer registration practice in the U.S. (See Registry Methods and Standards below.)
MCRS authority and data protection
- For information on the history, statutory authority, and objectives of the Minnesota Department of Health’s statewide cancer registry please visit Legislative Authority for MCRS.
- Learn more about the Minnesota Government Data Practices Act.
Cancer mortality data
MCRS obtains and analyzes death certificates annually from the Minnesota Office of Vital Records (OVR), which processes and maintains these records. Analyzing deaths from cancer is necessary to completely describe the cancer burden, and to evaluate the progress made in treating and controlling cancer in Minnesota. Only the underlying cause of death is used in calculating cancer mortality rates. For more information about Minnesota death certificates, visit Minnesota Center for Health Statistics, Office of Vital Records.
Population estimates
National Cancer Institute (NCI) – U.S. Population Data
Collecting and processing cancer incidence and mortality data
Registry methods and standards
The North American Association of Central Cancer Registries (NAACCR) provides the data dictionary and standards governing data collection, coding, and processing used in member central cancer registries. For more information about NAACCR please visit the section “Central Registry Standards” on the North American Association of Central Cancer Registries (NAACCR).
Definitions for cancer incidence data
A diagnosis of cancer includes identifying and describing where in the body (site) the cancer started and the cell type (histology) of the tumor. International Classification of Diseases for Oncology (ICD-O) is the coding system that defines the site and histology for each cancer. The World Health Organization maintains the ICD-O coding rules. The current version of the ICD-O rules is ICD-O/WHO 2008. For more information on the ICD-O coding system, visit International Classification of Diseases for Oncology.
To analyze cancer data, ICD-O site and histology codes are grouped together using the National Cancer Institute’s SEER Program conventions and standards. (SEER Site Recode page and to read more about the Site Recode ICD-O-3/WHO 2008 Definition).
Definitions for cancer mortality data
Causes of death are coded using the World Health Organization’s International Classification of Diseases (ICD). The current version of the ICD is ICD-10, 2016, which can be viewed at ICD-10 Version 2016.
The NCI’s SEER program groups ICD causes of death codes together to analyze cancer mortality data. The site groupings account for changes in coding over time to facilitate reporting of long-term trends. To learn more about SEER’s Cause of Death Recode, please see the SEER Cause of Death Recode page.
Statistical methods
Analytic software
| Estimate or analysis conducted | Software used |
|---|---|
| Rates and counts | SEER*Stat |
| Analyzing trends | Joinpoint |
| Limited duration prevalence | SEER*Stat |
| Complete prevalence | Comprev |
Definitions for cancer statistics
Age-adjusted rate (or age-standardized rate)
An age-adjusted rate is a rate that has been adjusted (or standardized) for differences in age distributions between populations. Age-adjustment enables fair comparisons of rates when there are differences in the age structures of populations, for example, over different time periods (e.g., Minnesota rates from 1988-2012 versus 2015-2019) or for different geographic areas (e.g., Minnesota rates versus U.S. rates).
The age-adjusted rate is a weighted average of age-specific rates (see below) with the weights calculated from a standard population as the proportion of people in age-groups. The standard population is based on the 2000 U.S. population and age-adjusted rates are reported as per 100,000 people. To see how age-adjusted rates are calculated, go to the NCI’s Tutorial to Calculate Age-Adjusted Rates.
Age-specific rate
An age-specific rate is the rate of cancer for a specific age group. It is calculated as the number of cancers occurring in people of a particular age group during a specified period of time divided by the total number of people in that age group at that time period.
Annual percent change (APC)
The annual percent change in the age-adjusted rate each year over a specified period of time.
Average annual percent change (AAPC)
The average annual percent change for a defined period of time summarizes the annual percent change in age-adjusted rates over a period of multiple years. The AAPC is a weighted average of APCs.
Cancer mortality
Cancer mortality is the number of deaths from cancer in a specified time period.
Cancer mortality rate
A cancer mortality rate is the rate of deaths from a cancer (or group of cancers) in a defined population during a specified time period. It is calculated as the number of deaths from a cancer or group of cancers during a specified period of time divided by the total number of people in a defined population during that time.
Confidence intervals
Confidence intervals show the expected range of random variation in an estimated statistic. For a 95% confidence interval, we would expect the rate to fall within that range 95% of the time. Confidence intervals provide a measure of the uncertainty (or lack of precision) of a cancer statistic. An estimate with a wide confidence interval would not be considered very precise.
Cancer incidence
Incidence in the number of newly diagnosed cancers occurring in a specified time period.
Cancer incidence rate
Is calculated as the number of newly diagnosed cancer or group of cancers occurring in people in a defined population during a specified time divided by the total number of people in the population at risk for the cancer or group of cancers during that time. Rates are typically expressed as number of cancers per 100,000 population (or persons).
Unless otherwise noted, cancer incidence rates are reported by gender for males and females only. Gender is reported as a person’s sex at birth and includes male, female, transexual natal male and transexual natal female, not stated or unknown, and other (intersex, disorders of sexual development). Rates are not calculated for individuals who fall into gender categories other than male and female because of small numbers, individual privacy, and a lack of accurate census population data.
Prevalence of cancer
There are different definitions of cancer prevalence with different estimation methods. Prevalence counts include both people with newly diagnosed cancers and cancer survivors who were diagnosed during a specified period of time.
- Limited duration prevalence – is the estimated number of people alive on a specified day who had a diagnosis of the disease within a specified period of years.
- Complete prevalence – is the estimated number of people alive on a certain day who ever had a diagnosis of cancer, regardless of whether they are undergoing active treatment or are considered cured.
Relative cancer survival
Relative survival compares the survival of people with a given cancer (or group of cancers) to the survival of people who do not have cancer. It is calculated as the observed survival rate for people diagnosed with cancer divided by the expected survival rate for people with the same age and race/ethnicity who are in general population. A 100% 5-year relative survival rate for a given cancer means that people diagnosed with this cancer are as likely to live five more years as people of the same age-group and race/ethnicity from the general population.
Standard population
The current standard population used in calculating age-adjusted rates is the 2000 U.S. standard population.
State Community Health Services Advisory Committee (SCHSAC) geographic divisions
The State Community Health Services Advisory Committee (SCHSAC) advises the health commissioner and provides guidance on the development, maintenance, financing, and evaluation of community health services in Minnesota. SCHSAC recommendations influence public health policy, guidelines, and practice throughout Minnesota. SCHSAC regions represent Minnesota’s community health boards, whose representatives are members of SCHSAC. For more information about SCHSAC regions, go to State Community Health Services Advisory Committee (SCHSAC).
Unstable rate
An unstable rate is defined as one with a count less than 20 or a relative standard error (100 x SE/Rate) > 30%. If a rate was unstable only counts were included in a table. Unstable rates in the tables are denoted with "N/A" or a flag denoting an unstable rate.
Other terms
Please visit NCI Glossary of Statistical Terms.