Electronic Health (eHealth) and Artificial Intelligence-based Tools to Optimize In-hospital Patient Flow: A Scoping Review

Objectives: Congested hospitals are increasingly common. Electronic health (eHealth) and artificial intelligence (AI)-based tools may improve in-hospital patient flow, however their implementation into practice varies. This study aims to identify and synthesize evidence on implementing eHealth and AI-based tools to manage in-hospital patient flow. Methods: Structured language and keywords related to patient flow and eHealth or AI-based tools were searched in five databases. Studies were eligible if they reported barriers or facilitators (determinants) to implementing eHealth and/or AI-based tools, and/or key metrics for patient flow. Study characteristics, tool characteristics, study population, setting, and outcome measures were abstracted. Information related to determinants of implementation were categorized using the Theoretical Domains Framework and interventions were mapped to the Expert Recommendations for Implementing Change Taxonomy. Results: Twenty-five studies were included; 40% were quasiexperimental studies and most (n=19) were conducted in the United States. Four categories of tools were identified with imbedding eHealth or AI-based tools into an existing electronic medical or health record being the most common. Barriers to tool implementation were commonly linked to the environmental context and resources (n=5), while facilitators were linked to social influence (n=4). Conclusions: This scoping review classified the reported barriers and facilitators to implementing eHealth and AI-based tools to improve in-hospital patient flow. Future research on in-hospital patient flow should adopt the identified measures when reporting tool effectiveness. To improve implementation efforts, more consistent reporting of determinants of tool implementation is needed.