A Holistic Approach to Management of Digital Radiography Dose and Image Quality across a Large Imaging Enterprise
Abstract
Purpose
Although radiographic exposures are individually low dose, in 2016 there were 275 million in the U.S.; the broad population makes consistent dose and image quality important. A large institution faces significant challenges to achieving this: vast data collection, timely data auditing, and effective intervention in diverse environments. This work describes a novel, decade-long approach addressing these challenges in high-volume digital radiography and the clinical impact.
Methods
The program covered sixty-eight stationary and portable DR units from five major manufacturers at multiple facilities, collectively producing more than 50,000 exposures monthly. Reject rate and average EI per user/protocol were chosen as performance metrics. An in-house, semiautomated system - utilizing remote software, shared network drives, and manual extraction - was developed to streamline exposure record collection. Custom scripts automated dataset aggregation for monthly analysis and review of metrics, incorporating root-cause analysis (alerts) for individual technologist reject rates exceeding 7% or EI +100 relative to EItarget of 225.
Results
This QA program has been running at our institution since 2011, with over 600,000 exposures tracked in 2025. From 2011 the mean retake reject rate has been reduced from 15% to 2.5% and the mean EI from 350 to 220. Mean EI for specific high-impact torso protocols decreased by 50-60%. Interventions reduced the technologists’ alert recurrence; in 2023, 12% of technologist performance metrics exceeded alert levels, but 66% of those were in compliance by the following month. In addition we present specific instances where program findings indicated x-ray machine problems and timely corrections were made.
Conclusion
A semi-automated remote data collection and analysis of reject rates and exposure index has allowed a sustainable review of high-volume DR resulting in significant improvement in consistent patient dose and image quality. Education instead of punitive responses was essential to problem identification and resolution.