Predicting Electronic Health Record (EHR) Usability: A Scoping Review of Models, Metrics, and Opportunities for Qualitative and Quantitative Analysis (Preprint)

Le résumé fourni par la source

BACKGROUND Electronic Health Records (EHRs) have become foundational in modern health care, offering potential benefits in care coordination, data sharing, and patient safety. However, poor EHR usability remains a major barrier, contributing to clinician burnout, inefficiencies, and errors. OBJECTIVE This scoping review examines the current research landscape on predicting EHR usability, with a focus on theoretical models, usability metrics, and analytic approaches. We identify gaps and opportunities for integrating predictive analytics and artificial intelligence (AI) to advance usability research. METHODS Following Joanna Briggs Institute and PRISMA-ScR guidelines, we systematically searched Medline, Web of Science, IEEE Xplore, and Scopus library databases for studies published between 2009 and 2023. Inclusion criteria focused on empirical research using predictive methods or models related to EHR usability. Data were charted and synthesized thematically. RESULTS From 2323 screened articles, 47 studies were selected for detailed review (from across 26 countries). Most of these studies focused on EHR adoption and acceptance. The dominant EHR adoption models discussed, namely the Technology Acceptance Model (TAM), the Unified Theory of Acceptance and Use of Technology (UTAUT), and the DeLone and McLean Information Systems Success (ISS) Model, were critiqued. Usability predictors such as perceived usefulness (PU) and perceived ease-of-use (PEOU) were prevalent. There were 27 specific usability predictors that were identified. Common predictive analytic approaches were many forms of regression analysis and structured equation modelling (SEM). However, no studies were identified that applied predictive modeling (AI) to dynamically forecast EHR usability beyond cross-sectional surveys. Few studies leveraged AI for usability prediction. CONCLUSIONS The explicit focus in this study on prediction of EHR usability is distinctive in the literature. It extends prior usability reviews (mostly focusing on adoption, not prediction). Predictive modeling for EHR usability remains underdeveloped throughout 2009-2025. Existing frameworks primarily assess adoption intent rather than operational usability over time. The critique of reliance on cross-sectional surveys and lack of task-based objective metrics was identified. Potential bias in the self-reported measures dominate in the EHR adoption model literature with one-time period of post-implementation modelled. Current EHR usability research remains largely survey-based, indicating opportunities to apply predictive modeling. Moreover, in the models there were no feedback loops from output back to inputs to dynamically change the construct determinants of overtime. However, the large volume of predictive techniques shows clear unique effort to utilize many variants of regression and structured equation modelling to establish predictor variables for post-implementation EHR usability. Therefore, to advance the field, future research is needed into the effectiveness of models that quantitatively analyze usability in the form of predictive analytics or categorization.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.