Work list prioritization tool cuts reporting and turnaround times by over 70%
An artificial intelligence-powered work list triage tool could significantly reduce report turnaround times in real world hospital settings, according to new research.
Radiology departments commonly use first-in, first-out work lists, meaning examinations are generally interpreted in the order they enter the queue. These workflows may not account for clinical urgency, potentially delaying reports for patients whose examinations require more immediate attention.
But a new analysis, published in the Journal of Medical Internet Research, suggests that use of AI triage for work list prioritization and report generation could help address this issue while also allowing radiologists to reclaim hours of time throughout the course of a shift. The study’s findings suggest that this could help reduce turnaround times by as much as 90%; this would benefit both patients and providers by prioritizing the most urgent cases first and by reducing the amount of time it takes to create individual reports in busy healthcare settings.
“AI systems are increasingly being integrated into radiology workflows, not only for image analysis but also as tools for work list prioritization and report generation,” Srinath Sridharan, PhD, with Data Science & Intelligence at Changi General Hospital in Singapore, and colleagues noted. “However, real-world evidence of their operational impact remains limited.”
To evaluate whether AI could address these challenges, researchers conducted a prospective, single-center crossover study involving eight board-certified radiologists. Each interpreted chest radiographs during two reporting sessions: one with a conventional first-in, first-out work list without AI and another using an AI-triaged work list with integrated report-generation tools. The sessions were separated by a four-week washout period.
The study included 1,054 chest radiographs acquired between November 2023 and January 2024. During that time, radiologists' report generation time—defined as the time from opening an examination to finalizing the report—fell substantially with AI assistance. Median generation times decreased from 2 minutes with the unaided workflow to 0.53 minutes with the AI-assisted workflow, representing a 73.3% reduction. The largest improvement occurred among normal examinations. Median report generation time for normal chest radiographs decreased from 2 minutes to just 0.2 minutes, a 90% reduction.
Average turnaround times also decreased from nearly 15 hours with the standard workflow to 82 minutes with AI assistance, representing a nearly 91% reduction. This finding was consistent across all urgency categories, including critical examinations, with statistically significant differences in each category.
The results highlight a potential role for AI beyond automated image interpretation. Rather than replacing radiologists' diagnostic judgment, these systems can be used to organize the reading queue according to clinical urgency and assist with report generation. This could be particularly beneficial in high-volume radiology departments, where delays can accumulate as readers work through lengthy work lists, the research team suggested.
Read more here.
In addition to her background in journalism, Hannah also has patient-facing experience in clinical settings, having spent more than 12 years working as a registered rad tech. She began covering the medical imaging industry for Innovate Healthcare in 2021.