The D2Refine Platform for the Standardization of Clinical Research Study Data Dictionaries: Usability Study (Preprint)

BACKGROUND D2Refine provides a Web-based environment to create clinical research study data dictionaries and enables standardization and harmonization of its variable definitions with controlled terminology resources. OBJECTIVE To assess the usability of the functions D2Refine offers, a usability study was designed and executed. METHODS We employed the TURF (task, user, representation, and function) Usability Framework of electronic health record usability to design, configure, and execute the usability study and performed quantitative analyses. D2Refine was compared for its usability metrics against two other comparable solutions, OntoMaton and RightField, which have very similar functionalities for creating, managing, and standardizing data dictionaries. We first conducted the function analysis by conducting one-on-one interviews armed with questionnaires to catalog expected functionality. The enrolled participants carried out the steps for selected tasks to accomplish specific goals and their feedback was captured to conduct the task analysis. RESULTS We enrolled a group (n=27) of study developers, managers, and software professionals to execute steps of analysis as specified by the TURF framework. For the within-model domain function saturation, D2Refine had 96% saturation, which was 4 percentage points better than OntoMaton and 28 percentage points better than RightField. The manual examination and statistical analysis of the data were conducted for task analysis, and the results demonstrated a significant difference for favorability toward D2Refine (P<.001) with a 95% CI. Overall, 17 out of 27 (63%) participants indicated that D2Refine was their favorite of the three options. CONCLUSIONS D2Refine is a useful and promising platform that can help address the emerging needs related to clinical research study data dictionary standardization and harmonization.