Interoperability and the New SNOMED CT - Encoded CAP Cancer Checklists (SECCC)
Richard Moldwin, MD, PhD MD; College of American Pathologists;
Content:
The College of American Pathologists (CAP) publishes cancer case synoptic forms ("checklists"). The American College of Surgeons requires the use of the checklist data elements for accreditation. Previously, CAP provided the checklists in PDF and Word formats, encoded with SNOMED CT to help enable semantic interoperability. However the checklists were difficult to computerize, and interoperability between institutions was not achieved. We propose a solution to these problems.
Technology:
Microsoft (Redmond, WA): Word, Excel, Access, SQL Server, VBA6, C#; Oracle (Redwood Shores, CA); Innovasys (Westowe, UK): Document!X and Help Studio; Altova (Beverly, MA): XML Spy; XML, XSD.
Design:
Each Word-formatted SNOMED CT-Encoded CAP Cancer Checklist (SECCC) was copied into an Excel spreadsheet. VBA6 code was written to import each spreadsheet into a custom Access database, as checklist templates. A template editor was programmed in Access/VBA6 for editing, copying, versioning and creating the new computerized SECCC templates. A module was programmed in C# to automatically transform each template into a custom XML format. XML schemas were constructed using XML SPY to assist in validation of the various types of XML files. A screen generator web component was written in C# to JIT generate web data-entry forms from the XML files on-the fly, and to provide a user customizable method of data storage.
Results:
This system allows us to easily map each checklist data item to specific metadata, such as SNOMED CT Concept IDs, NAACCR data item numbers and Collaborative Staging data elements, as part of a collaborative project with NAACCR, CDC and others. The first version of the SECCC database file was distributed to users in Jan 2007, complete with SNOMED CT encodings. Ongoing work is directed to the production of new and revised templates with mappings to the above systems, refinement of the tools, and integration with caBIG.
Conclusion:
We have created a mechanism for the interoperable and collaborative creation, modification, deployment and use of the SECCCs. Once fully operational, this system will allow pathologists, oncologists, surgeons, research groups, cancer registrars (NAACCR), CDC, NCI and SEER to have an end-to-end interoperable flow of aggregated SECCC data from source to destination.
