Impact of a Digital Intelligence Platform on Radiology Workflow Efficiency and Patient Satisfaction: A Retrospective Cohort Study.
Authors
Affiliations (2)
Affiliations (2)
- Department of Radiology, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China (J.F., X.C., J.F., L.C., Q.W., J.Z.).
- Department of Radiology, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China (J.F., X.C., J.F., L.C., Q.W., J.Z.). Electronic address: [email protected].
Abstract
Traditional radiology management models may face challenges in meeting the growing demand for medical imaging services, potentially related to resource scheduling and workflow coordination. This supply-demand imbalance necessitates comprehensive digital transformation and workflow optimization to achieve systematic enhancement of operational efficiency. To evaluate the efficacy of a digital intelligence platform in restructuring radiology workflows, specifically assessing its impact on appointment scheduling, examination throughput, and report turnaround times across routine and emergency tracks. This paper conducted a retrospective cohort study comparing a pre-implementation phase (January-October, 2022) with a post-implementation phase (January-October, 2024) at a tertiary hospital in China. The platform integrated an intelligent scheduling algorithm and a dynamic report dispatching system. Key metrics included Order-to-Appointment Time (OAT), Order-to-Examination Time (OET), Exam-to-Report Time (ERT) and patient satisfaction scores. Multivariable linear regression was performed to adjust for confounding factors (patient age, type), and subgroup analyzes were conducted for emergency examinations. A total of 1149,154 examinations were analyzed. Post-implementation, the total examination volume surged by 66.4% against stable staffing levels. Despite this load, the platform significantly reduced the median OAT (CT: -72.4%; MRI: -33.8%; P<.001) and OET (CT: -50.6%; MRI: -34.6%; P<.001). Regression analysis confirmed the platform as an independent predictor of reduced pre-exam wait times (B = -0.359; P<.001). Conversely, the overall median ERT increased significantly (CT: +130%; MRI: +222%; P<.001) due to the disparity between the 66% volume surge and stable staffing. However, the emergency subgroup analysis revealed a significant reduction in ERT for emergency CT (-22.1%) and MRI (-26.6%), validating the effectiveness of the priority dispatching algorithm. Patient satisfaction scores regarding waiting time improved from 78.2 to 85.7 (P<.001). The platform successfully resolved pre-examination bottlenecks and improved patient satisfaction through algorithmic resource orchestration. The increased routine report turnaround time highlights that human interpretation capacity remains the ultimate rate-limiting step amidst surging demands, while the platform's intelligent prioritization effectively ensured the timeliness of emergency diagnoses. Future efforts should focus on integrating generative artificial intelligence (AI) and diagnostic algorithms to close the efficiency loop.