USABILITY OF AI-ENABLED MEDICAL INFORMATICS REPOSITORY FOR ACCELERATED CLINICAL RESEARCH ACCESS
Abstract
This study evaluated the usability of an AI-ready, web-based Digital Research Repository developed for the Southwestern Medical Center Research Department to mitigate research access delays, physical document degradation, and administrative communication gaps. Utilizing an explanatory sequential mixed-methods design, the study established baseline metrics from 32 resident doctors, engineered the system using user-centered design principles, conducted task-based usability evaluations, and administered the System Usability Scale (SUS) paired with qualitative interviews. Pre-implementation diagnostics revealed prominent baseline barriers, specifically in document organization and storage security (M = 4.20, SD = 0.57) and document location latency (M = 3.80, SD = 0.57), culminating in an overall baseline mean friction score of 3.80 (SD = 0.51). Post-implementation usability testing yielded a composite SUS score of 72.5 (SD = 7.07), outperforming the standard industry benchmark of 68.0 and positioning the platform firmly within the acceptable, "Good" usability classification. Qualitatively, clinical practitioners reported drastically accelerated file retrieval speeds and structurally clearer archive navigation, while noting minor recommendations for loading optimization and real-time request-status indicators. These findings demonstrate that deploying structured, AI-ready data layers effectively resolves localized knowledge-management friction, establishing a scalable infrastructure that prepares traditional medical clinical archives for future intelligent systems integration, natural language processing, and advanced semantic discovery.
