Abstract
Background/Aims Creating a virtual tumor registry to function within the Cancer Research Network (CRN) enables us to identify tumors in our members. This is critical for performing cancer research for the CRN. Data standards and staging schema are defined by regulating agencies external to CRN and require all users to adhere to these standards. This establishes a gold standard for modeling our data structure that all participating plan sites should be able to achieve.
Methods Initial datasets were modeled from a commercial registry package. The vendor-provided data dictionary included variable names, formats, and length. The data dictionary also included whether the variable was required or optional, and if the data entry was system-provided or added with keyboard entry. Unfortunately, health plan-specific data characteristics were soon added and the virtual tumor registry began to diverge from the gold standard. It had been assumed that it was a static target (i.e., not requiring adjustment). However, the regulating agencies had changed the standards and a major revision was impending. To remain compliant with the gold standard, the tumor registry would need to be expanded to 117 variables and the data needs to be destroyed and rebuilt each time the data is refreshed.
Results Once expanded to 117 variables, we are having difficulties getting health plans re-synchronized with the data standards. We have developed some QA SAS code to investigate anomalous findings and return results to the health plan.
Conclusions The virtual tumor registry is the key dataset to allow the CRN to function as a cancer research consortium. It allows the identification of tumor diagnoses, tumor staging, initial cancer directed treatment and annual follow-up. Recent additions capture advanced testing including immunostains, nucleic acid polymorphisms or mutations, and other personalized prognostic molecular markers. The virtual tumor registry can be used for cancer research feasibility estimates, treatment modality identification, survival calculations and molecular marker studies.
