Software Products for Collaborative Staging Data Collection System Version 02.03.02
Release history of Collaborative Stage version 2
Collaborative Stage Version 02.03.02 Released 12/2/2010 Timeline for Implementation CSv2 version 02.03.02 must be used to code all cases diagnosed on or after January 1, 2011. Once version 02.03.02 is implemented in a registry, it should be used to code all newly abstracted cases diagnosed from 2004 forward. (2004 was the first year for which Collaborative Stage was collected.) Therefore, the item CS Version Input Original [NAACCR item number 2935] must be 020302 or greater for cases diagnosed in 2011 or abstracted using version 02.03.02. Once version 02.03.02 is implemented in a registry, and existing CS data have been converted according to specifications provided, all cases with CS data coded in prior versions should have all output stage fields re-derived using the algorithm of the new version 02.03.02. Therefore, all cases with CS data will be expected to have CS Version Derived [NAACCR item number 2936] of 020302 after implementation and conversion. Contents of Release Collaborative Stage Data Collection System version 02.03.02 includes updates to the schemas resulting from the table proofreading and validation project. The last two digits of the version number are 02 because numbers 00 and 01 have been designated to mark cases undergoing conversion to version 02.03.02. Instructions for applying the version numbers 020300 and 020301 are provided in the conversion documentation portion of the release materials. This release contains the following changes: MyelomaPlasmaCellDisorder: A new schema for Multiple Myeloma and Plasma Cell Disorders has been added to the CS schemas. These cases were formerly part of the HemeRetic schema. Updates to Schemas: The CSv2 schemas have undergone an extensive proofreading and validation process. This has resulted in a number of changes throughout the system. These changes have provided a more consistent and standardized methodology of presenting the schema tables. The scope of work for this validation process included:
Site-Specific Factors: Site-specific factors have been added or modified for the following schemas:
Schema Discriminator (SSF 25): When a schema discriminator is required to identify the proper schema, the DLL will require a valid SSF25 value. For those schemas that have a defined schema discriminator (SSF 25), any blanks which were allowed in CSv0202 have been changed to 981 or 982. SSF 25 describes when it is needed to arrive at the correct schema. For schemas where SSF 25 is defined, if site and histology can already determine the correct schema, the DLL ignores what is in SSF 25 and selects the correct schema. The DLL will only return an error message if the code is an invalid value for SSF 25. Any edits based on invalid codes will also return an error if blanks (or other invalid codes) are in SSF 25. Other edits may be written to prevent valid values other than 981 and 982 from being used when the schema can be determined based on site/histology. Schema Discriminator (SSF 25) and EDITS: If site and histology define the schema, the DLL does not check the value of SSF 25 for consistency, such as checking SSF 25 versus the site code. Only if the SSF 25 code is invalid will the DLL report it as an error for SSF 25. There will be an edit provided within EDITS to check for the consistency of SSF 25 with site and histology. CSv2_Production_Release_Version_02_03_02_Release_Date_2010_12_07.zip The contents of the zip file include the following in separate folders:
Implementing Version 02.03.02 If you have not yet implemented CS version 02.00.01 (released in January 2010), 02.01.00 (released in February 2010), or 02.02.00 (released in April 2010), you must do so before implementing this new release. All conversion specifications and implementation instructions published for the initial version 2 releases must be followed prior to conversion to v020302. CDC has provided Northcon12 and a DLL that will perform the conversions and derivations from CS version 02.0x to CS version 02.03. See the section “Program to Convert from Collaborative Stage Version 02.02 to Version 02.03” below. Testing the CS Algorithm - tools to ensure that your software program calculates stage correctly. Test-O-Matic (4450K ZIP) – is an interactive Windows program that is useful for exploring Collaborative Staging tables and deriving stage. For more detailed information, please refer to CSv1 Test-o-Matic description. Conversion Specifications V.02.02 to V.02.03 The conversion specifications, document with release notes and rules-based conversions, and spreadsheets with schema/code-specific instructions, comprise the guide for release 02.03.02. No other implementation guides are provided for this release. Three updates to the initial conversion specifications have been issued and are published here. Changes to the conversion specifications and release notes in the PDF file are cumulative through the May 12, 2011 update. An update log is included at the beginning of the document, and changes in the text are highlighted accordingly. Changes to the spreadsheets are noted in red, and the red lettering is cumulative in the April 19, 2011 update. No changes to the spreadsheet were required for the May 12, 2011 update. Changes noted in the February 15, 2011 specifications: Review criteria for MyelomaPlasmaCellDisorder, Testis, and GIST and NET schemas modified; rules supplied for converting blank values in most site-specific factors to 988; conversion from CSv1 to CSv2 before conversion to CSv3 emphasized; prostate conversion and possible review before conversion clarified; generation of lists for manual review by conversion program emphasized.
Changes noted in the April 19, 2011 specifications: Rules supplied for converting obsolete values mapping to ERROR in SSF25 for EsophagusGEJunction/Stomach and LacrimalGland/LacrimalSac; review described for converted code 015 in SSF25 for LacrimalGland/LacrimalSac.
Changes noted in the May 12, 2011 specifications: Clarifications have been added to emphasize that the schema names utilized in the document and associated spreadsheet are based on definitions and schema names for CSv0203. Also, the specifications which refer to the fifth spreadsheet tab (Obsolete Reviewed) were enhanced.
A conversion program to convert cases from CSv0202 to CSv0203 is now available (see the next sections). NOTE: Facilities that have created their own conversion programs for v0202 to v0203 based on the specifications posted from December 2010 through April 2011 will need to apply the revised conversion specifications as of May 12, 2011 to their data. The following document details considerations that these facilities should note when updating their data according to the May 12, 2011 specifications. Program to Convert from Collaborative Stage Version 01.xx to Version 02.02 CDC has released Northcon12, a freestanding Windows-based utility program that converts a file of cancer registry records in NAACCR11 record layout format to the NAACCR 12 record layout. Conversion of Collaborative Stage (CS) data from version 1 to version 2 is embedded in this program. The CS conversion is based on the published specifications and has been extensively tested. When the conversion is run, the user has the option to derive the Collaborative Stage version 2 stage outputs on appropriate records. For more information or to download the program, see the CDC-NPCR Web page at: http://www.cdc.gov/cancer/npcr/tools/registryplus/up_download.htm. Instructions for accessing the program’s underlying dynamic link libraries (DLLs) are available on within the download package. These can be used by anyone who wishes to perform the conversions using another program. Program to Convert from Collaborative Stage Version 02.02 to 02.03 CDC provides a conversion program and function library to convert data from NAACCR 12.0 to NAACCR 12.1. The conversion of Collaborative Stage (CS) data field-by-field from version 0202xx to 0203xx is embedded in the program. The program, Northcon121.exe, uses a new Windows Dynamic Link Library function library (DLL), nconv121.dll, to do its work., Instructions for using the DLL are included in the setup package. Download from the CDC-NPCR Web page, http://www.cdc.gov/cancer/npcr/tools/registryplus/up_download.htm. Support Files and Test Files CS Defaults version 020302 Dated 20110218.xls (120K XLS) An Excel spreadsheet which provides the default values for the CSv0203 data fields for Death Certificate Only (DCO) cases originally coded under CSv020302. The spreadsheet is organized by schema and is for cases which have a behavior code of 3. Instructions are included. The previous version was for v0202 and not v0203. CSv2_Valid_Invalid_&_Obsolete_Codes_Tables_2010_12_06.zip (155K ZIP) Valid, Invalid, and Obsolete Codes Tables: Excel Spreadsheets are provided which show schema-specific valid and invalid values for each field for CS version 02.03.02. An Excel spreadsheet is also included which contains the obsolete codes for version 02.03.02. Please note that the obsolete spreadsheet has a table for each obsolete code type and one of all obsolete codes combined. These three spreadsheets are not mutually exclusive since the obsolete codes are also contained in the valid code list. NAACCRv12_Valid_Invalid_v020302_2010_12_06.zip (1870K ZIP) Test Cases for the Algorithm: Vendors have expressed an interest in obtaining test records which have a known behavior when the CS algorithm is run on them. Two sets of random test cases, one for valid CS codes (valid.NAACCRv12.in.v0203020.txt) and one set with one and only one CS invalid code (invalid.NAACCRv12.in.v02030200.txt) in each record were generated from the XML files. These input files (.in.) are in NAACCR v12 format (LRECL = 3339) and only contain the CS input fields and other fields necessary for the DLL to successfully run. On the input file, the record is un-staged, i.e., all derived CS fields are blank. These files can be used as input to the corresponding version of the CS algorithm (v020302). The output files (.out.) contain the staging results generated by the CSv2 algorithm. When staging the Valid “in” file, no errors should be generated by the algorithm. When staging the Invalid “in” file, one specific error message should be generated per data record. Please note that each record of the invalid “out” file contains the first 200 characters of the error message generated by the CSv2 algorithm. The error message has been written to column 1236 (Reserved 05). For Questions: As the CSv2 algorithm is implemented and employed with real case scenarios, questions regarding the CS site-specific schemas should be sent to csv2@facs.org. If you would like to receive vendor/software developer e-mails, you can sign up by e-mailing csv2@facs.org For questions and comments on CSv2 production version software, please e-mail Joseph Rogers at JRogers@cdc.gov Collaborative Stage Version 02.02.00 Release Announcement with Implementation InstructionsContents of Release The new version includes corrections to the software and corresponding API documentation. Two new components have been added: VB source code and HTML documentation for schema selection. The function cstage_get_version was added to the EDITS interface. A blank in the field Lymph-Vascular Invasion will now be considered valid when the year of diagnosis is 2009 or earlier and CS Version Input Original is less than 020000. See the “Release Notes” and the file “Changes for CSv2.2.doc” for more details. The release also contains changes to the following schemas.
Implementing Version 02.02.00 If you have not yet implemented CS version 02.00.01 (released in 01/10), or 02.01.00 (released in 2/10), you can implement this new release instead. Note that all conversion specifications and implementation instructions published for the initial version 2 release must still be followed. If you have implemented CS version 2 and converted your old data, but have not yet collected any new cases under CSv2, you can just replace the DLL with the new one and re-derive all stage data with the new version, setting the CS Version Derived to ‘020200’. If you have implemented CSv2, converted old data, and collected any new data under CSv2, you may also need to review some cases for which the schema selection algorithm may have returned an incorrect result. This new release includes a change in the handling of blanks in schema discriminator, Site-Specific Factor 25 (SSF25). The schema selection algorithm will no longer resolve to a schema when a discriminator is needed but is blank. Any cases matching the following sites and histologies and having a blank string in SSF25 should be reviewed. The blank string should be replaced with an appropriate non-blank code and all of the CS fields should be reviewed and recoded if necessary.
CSv2_Production_Release_2010_04_29.zip The contents of the zip file include the following in separate folders:
Testing the CS Algorithm - tools to ensure that your software program calculates stage correctly.
Conversion Specifications 07/30/10 - Revised Specifications for Converting CSv1 to CSv2 Released Specifications for converting from version 1 to version 2 of Collaborative Stage have been revised to accommodate Site-Specific Factors 1 through 6 for which blank was an acceptable value but which are now required to derive staging information. After the first release of the conversion specifications, a problem was discovered that resulted in cases failing to derive either TNM, 6th ed., or Summary Stage 77 and Summary Stage 2000, although the cases had derived all stage information in version 1. The problem occurred when site-specific factors which were either not defined or not required in the past became required for staging in version 2, and only when the newly required site-specific factor had been left blank. The specifications have been revised to rectify this problem, which only affected five of over 150 schemas. There are changes in both components of the specifications, the Word document and the Excel spreadsheet. The revisions to the conversion of Collaborative Stage will be incorporated into the next release of CDC’s general program for converting cancer registry records from the NAACCR 11 layout to the NAACCR 12 layout. The new version of the CDC program, released as NCONV12.DLL, will be version 1.0.2. Users who have not already converted from NAACCR 11 to NAACCR 12 should use the new specifications and the new version of NCONV12.DLL, version 1.0.2, which will also include a change to the conversion of Class of Case described in a revised implementation document published by NAACCR. Users who have already converted can use the revised specifications to program corrections to any CS data they may have with the problem. The revised files are available for download on the CS Software page at: http://cancerstaging.org/cstage/software/index.html. Modification of the CS Testis Schema - updated July 9, 2010 Since the publication of the AJCC 7th edition, important information related to the capture of serum tumor markers has been brought forward to the AJCC chapter authors. READ MORE
Resources
Program to Convert from Collaborative Stage Version 1 to Version 2 CDC has released Northcon12, a freestanding Windows-based utility program that converts a file of cancer registry records in NAACCR11 record layout format to the NAACCR 12 record layout. Conversion of Collaborative Stage (CS) data from version 1 to version 2 is embedded in this program. The CS conversion is based on the published specifications and has been extensively tested. When the conversion is run, the user has the option to derive the Collaborative Stage version 2 stage outputs on appropriate records. For more information or to download the program, see the CDC-NPCR Web page at: http://www.cdc.gov/cancer/npcr/tools/registryplus/up_download.htm Instructions for accessing the program’s underlying dynamic link libraries (DLLs) are also available on this Web page. These can be used by anyone who wishes to perform the conversions using another program. Support Files and Test Files CSv2 Defaults version 020200 Final.xls (115K XLS) An Excel spreadsheet which provides the default values for the CSv2 data fields for DCO cases originally coded under CSv2. The spreadsheet is organized by schema and is for cases which have a behavior code of 3. CSv2_Valid_Invalid_&_Obsolete_Codes_Tables_2010_04_29.zip (130K ZIP) Valid, Invalid, and Obsolete Codes Tables: Excel Spreadsheets are provided which show schema-specific valid and invalid values for each field for CS version 02.02.00. An Excel spreadsheet is also included which contains the obsolete codes for version 02.02.00. Please note that the obsolete spreadsheet has a table for each obsolete code type and one of all obsolete codes combined. These three spreadsheets are not mutually exclusive since the obsolete codes are also contained in the valid code list. NAACCRv12_Valid_Invalid_v020100_2010_04_29.zip (2095K ZIP) Test Cases for the Algorithm: Vendors have expressed an interest in obtaining test records which have a known behavior when the CS algorithm is run on them. Two sets of random test cases, one for valid CS codes (valid.NAACCRv12.in.v020200.txt) and one set with one and only one CS invalid code (invalid.NAACCRv12.in.v020200.txt) in each record were generated from the XML files. These input files (.in.) are in NAACCR v12 format (LRECL = 3339) and only contain the CS input fields and other fields necessary for the DLL to successfully run. On the input file, the record is un-staged, i.e., all derived CS fields are blank. These files can be used as input to the corresponding version of the CS algorithm (v020200). The output files (.out.) contain the staging results generated by the CSv2 algorithm. When staging the Valid “in” file, no errors should be generated by the algorithm. When staging the Invalid “in” file, one specific error message should be generated per data record. Please note that each record of the invalid “out” file contains the first 200 characters of the error message generated by the CSv2 algorithm. The error message has been written to column 1236 (Reserved 05). Software Products for Collaborative Staging Version 1 (CSv1)Collaborative Staging Library - computer routines to derive stage and read the Collaborative Staging tables. Calculates TNM and AJCC Stage Group according to AJCC 6th edition, SEER Summary Stage 1977, and SEER Summary Stage 2000. Supplied as source code with build files for Microsoft Visual C++ version 6 and Makefile for gcc on Linux. Compiled DLL provided for Windows. Also a second DLL shell is provided to adapt library for Microsoft Visual Basic and some other programming environments.
Collaborative Staging Tables - available in different formats
Testing the CS Algorithm - tools to ensure that your software program calculates stage correctly.
Valid value tables - gives the valid storage values and corresponding labels for the derived CS fields for T, N, M and Stage and SEER Summary Stage. Unknown Values for Death Certificate Only Cases - provides the default/unknown values for the CS schemas. It can be used by vendors to populate a DCO record, filling in unknown values. Default and unknown values for CS items (80K XLS)Editing CS data fields It is highly recommended that all CS fields be checked for validity before running the CS Algorithm Valid and Invalid Codes Tables Excel Spreadsheets are provided which show schema-specific valid and invalid values for each field for CS version 01.04.01. An Excel spreadsheet is also included which contains the obsolete codes for version 01.04.01. These three spreadsheets are not mutually exclusive since the obsolete codes are also contained in the valid code list:
These Excel spreadsheets are also provided by site/histology within schema name. Please note that there are two rows each for lymphomas and hemeretic because of two histology codes that can be either lymphomas or leukemias depending on the site code. In addition, the site codes and histology codes are given in ranges for ease of use and that this is not an indication that all codes in the range are valid sites or histologies. The histology codes in these files have been corrected.
Test Cases for the Algorithm Random input test files of valid and invalid codes were generated from the XML files and used as input to the corresponding version of the CS algorithm. The input and corresponding output files for each are attached. Please note that the name of the schema is not included in output for v010400. Verified file: For version 010100 a set of input values were run through the algorithm and the derived fields were verified by hand that the results were correct (verify.out.v010100.txt). The output values were then run through the subsequent versions of the algorithm and the derived fields updated (see verify.out file for specific CS version). Any differences in derived values between the last two versions have been verified.
Suggested Method for Editing CS Before Running the Algorithm Below is a front end program that will allow the editing of each CS field before the algorithm is run so that error messages can be displayed on a specific field.
There are readme files with each. Please note that all of the executable files need to be in the same directory as the current version of the DLL. The Zip files are password protected and the password is "cstage." Questions or comments about these files should be sent to Lynn Ries at lynn_ries@nih.gov. For Questions: The CS algorithm is a very complex algorithm that has been extensively tested by the developers. As the CS algorithm is implemented and employed with real case scenarios, questions regarding the CS site-specific schema should be sent to Donna Gress, AJCC Technical Specialist at dgress@facs.org.
Revised December 5, 2011
Collaborative Staging Home Page
|