The Common Data Elements fo rC ancer Research :R emarks on Function sa nd Structur e

Summary Objectives: Th eN ational Cance rI nstitute (NCI) ha s develope dt he Commo nD ata Elements (CDE) to serv e as ac ontrolle dv ocabulary of dat ad escriptors fo rc ance r research, to facilitate data interchange an di nter-operability between cance rr esearch centers. We evaluate d CDE’s structure to se ew hether it could represent the elements necessary to support its intended purpose , an dw hether it could prevent errors an di nconsistencie s from bein ga ccidentally introduced. We also performed automated checks for certai nt ypes of conten te rrors that provide dar ough measure of curation quality . Methods: Evaluation wa sp erforme do nC DE content downloade dv ia the NCI’s CD EB rowser, an dt ransforme di nto relational database form. Evaluation wa s performe du nder three categories: 1) compatibility with the ISO/IE C1 1179 metadata model, on which CD Es tructure is based, 2) feature sn ecessary fo rc on trolle dv ocabulary support ,a nd 3) support fo ras tated NCI goal, se tu po fd ata collection forms fo rc ance r research. Results: Various limitation sw er ei dentified both with respec tt oc onten t( inconsistency, insufficien td efinition of elements ,r edundancy )a sw el la ss tructure –p articularl yt he need fo rt er ma nd relationship support ,a s wel la st he need fo rm etadata supportin gt he explici t representation of electronic forms that utilize sets of common dat ae lements . Conclusions: While ther ea re numerous positive aspect st ot he CD Ee ffort ,t her ei sc onsiderable opportunity for improvement. Our recommendations includ e revie wo fe xisting content by diverse experts in th e cance rc ommunity ;i ntegration with the NCI thesaurus to tak ea dvantage of th el atter’s links to nationally use dc ontrolle dv ocabularies, an dv arious schem a enhancements require df or electronic form support .