SEER*Stat MP-SIR Exercise 1a: Introduction to the Multiple Primary - Standardized Incidence Ratio Tables

Use the Multiple Primary Standardized Incidence Ratios (MP-SIR) session to conduct an analysis examining multiple subsequent cancers. That is, create an observed to expected (O/E) ratio based on the observed secondary events for the cohort using a SEER 9 registry rate file (to calculate expected). The cohort is defined by the selection criteria for an index record in conjunction with the rate file.

Create a table by sex, latency, and subsequent cancer site for subjects who have a first primary cancer (sequence number 0 or 1) meeting the following criteria:

  • cancer of the brain or nervous system,
  • not diagnosed by death certificate or autopsy,
  • in the SEER 9 registries (1973-2013).

Key Points

  • Several referent rate files are provided with SEER*Stat. You can also use SEER*Stat to create referent rate files. The rate files provided by SEER*Stat are stratified by race, sex, age (5 year intervals), calendar year (5 year intervals), and event variable (cancer site). Two files are currently distributed for use with each MP-SIR and MP-SMR database. One file is provided for each of the following race categories.
    • Race Combinations:
      • White/Unknown, Black, and Other
        Should be used when analyzing all races combined.
      • White, Black, and Other (not including Unknown)
        Should be used for race-specific analyses.
  • Starting with the November 2015 data submission, referent rates for SIRs are only provided using Site recode B ICD-O-3/WHO 2008.
  • By default, any cohort records that do not correspond to a race or sex in the selected rate file will be excluded from the analysis. The user can choose to use available expected rates for these records. For example, rather than excluding unknown sex, the user may choose to use either the male or female rates.
  • The SEER*Stat MP-SIR session has an "Historic Compatibility Option" on the Output Tab. This option is used for matching analyses pre-dating the release of MP-SIR. In the normal course of research, this option should not be used.
  • Care must be taken when generating SIR statistics on sex-specific sites. This exercise illustrates how some sex-specific statistics are correct while others could be inappropriate.

Step 1:  Create a New MP-SIR Session

  • Start SEER*Stat.
  • From the File menu select New > MP-SIR Session or use the MP-SIR button on the toolbar.

Step 2:  Select a Database (Data Tab)

  • It is extremely important that you select the database as the first step. The other choices you will make in this session will be based on variables in the selected database. The correct database must be selected in order to see the correct list of rate files and variables in selection statements, table statements, and the dictionary editor.
  • On the Data Tab select "Incidence - SEER 9 Regs Research Data, Nov 2015 Sub (1973-2013) <Katrina/Rita Population Adjustment>"

Step 3:  Choose the Referent Rate File (Rates Tab)

  • Move to the Rates Tab.
  • Select "SEER 9 1973-2013 (Nov 2015 sub), Race (WU/B/O), Event: Site recode B ICD-O-3/WHO 2008" from the list of available rate files.

Learn More...

  • The Rate File Details box contains information about the highlighted rate file, such as the database on which it is based and the variables it contains.
  • You can use a SEER*Stat Rate session to create your own referent rate files for use in MP-SIR sessions. See the MP-SIR Rates Tab in the SEER*Stat help system for more information.

Step 4:  Defining the Analysis Cohort (Selection Tab)

  • The analysis cohort is defined as the people with a primary cancer matching the index record criteria. You will use the Selection Tab to define the index record criteria.
  • Use the default Standard Index Record Selections to select malignant behavior and to exclude people diagnosed by death certificate or autopsy only.
  • To define the cancer site for the index record, add the following selection statement in the Additional Index Record Selections box:
    {Site and Morphology.Site recode B ICD-O-3/WHO 2008} = ' Brain and Other Nervous System'
  • Use the default Multiple Primary Selection to include the first primary only.

Step 5:  Set the Parameters (Parameters Tab)

  • Use the default settings for all parameters in the Study Time box. Because you are using a SEER database, the Exposure Date and Start Date are always the person's "Date of diagnosis recode", which is the date of diagnosis for their index record (as defined on the Selection Tab).
  • In the Latency (months and/or years) field, select "1y,5y,10y" from the drop down box. Leave the other fields blank in the Cut-Points for Time-Dependent Variables section.

Step 6:  Setting the Events (Events Tab)

The settings on the Events Tab define how events are identified in your analysis. Events are occurrences of interest in an individual's cancer history, such as development of a cancer, or death. In this exercise, we will use the default settings on the Events Tab.

Step 7:  Choose Your Statistic (Statistic Tab)

  • In the Statistic box, make sure that SIR Tables is selected.
  • Use the default settings in the Confidence Intervals box.

Step 8:  Set Table Variables (Table Tab)

  • When SIR Tables have been selected (Statistic Tab), the controls on the Table Tab include two lists of variables.
    • Values for the Event and Time Dependent Variables are not stored in the database, but are generated by SEER*Stat during execution.
    • Cohort Stratification Variables include standard, user-defined, and merged variables from the dictionary.
  • The Problem Statement for this exercise specifies that the output should be stratified by sex, latency, and subsequent cancer site.
  • From the Event and Time Dependent Variables box, set "Selected Events" as a Row variable and "Latency" as a Column variable.
  • From the Cohort Stratification Variables box, set "Sex" as a Page variable.

Step 9:  Edit the Output Settings (Output Tab)

  • Enter the following title on the Output Tab:
    SEER 9, 1973-2013, MP-SIR Multiple Event Analysis
    Cohort: Malignant Brain and Other Nervous System, First Primaries Only
    By Sex, Latency, and Selected Event
    MP-SIR Exercise 1a
  • Check the box to Display All Calculated Statistics In Output Matrix

Learn More...

  • There are 12 statistics types available in an MP-SIR session. By default, only two are shown in the output matrix unless the option to display all is selected. If you click on the Set Default button when the check box is checked, the default when you run any MP-SIR session will be to display all calculated statistics.

Step 10:  Execute SEER*Stat

  • Use the Execute button or select Execute from the Session menu to execute the session. (Execute Offline is a 3rd option available and has been explained in previous exercises.)
  • A dialog will display the progress of the job. When the job completes a new window will open containing the output table or matrix.

Step 11:  The Results Matrix

  • Use the Save As command on the File menu to save the matrix. Enter "MPSIR Exercise 1a" as the filename. SEER*Stat will assign the "smm" extension to indicate that this is a "SEER*Stat MP-SIR Matrix" file.
  • Compare your results to this SEER*Stat matrix file: Exercise Matrix 1a Results.
  • The Sex variable set on the Table Tab is a page variable so the results matrix consists of multiple pages of output. Use the drop down list on the toolbar to select a different page to view.
  • All 12 MP-SIR statistics types are displayed based on the selection you made on the Output Tab. Use the Matrix Options to hide any of the statistics.
  • The MP-SIR Results Matrix section of the help system contains more information about the SEER*Stat matrix and its features.