Room: Exhibit Hall | Forum 1
Purpose: The radiation exposure from medical imaging has increased significantly in recent decades. CT is by far the largest source of medical radiation exposure. We aimed to investigate our facility dose levels using dose monitoring software and compared them with external benchmarks.
Methods: CT dose data acquired at different facilities across our academic campus from December 2017 to February 2018 were retrospectively analyzed using dose monitoring software (Nexodose, Bracco, NJ, USA). The software enables capturing, tracking, and reporting radiation doses delivered to patients and allows us to compare ourselves to national and regional data (Diagnostic Reference Levels, DRLs). All data were exported as Excel spreadsheets, and further analysis was conducted using an in-house built Matlab script. Those examinations which included CT Biopsy, dual energy CT, or were missing study description, study name or body part were excluded. All examinations were categorized into different protocol groups based on RadLex playbook ID (RPID). The 50% percentile of CTDIvol, DLP and SSDE were calculated and compared to DRLs.
Results: CT examinations were performed on 17140 individual adult patients (mean age: 56years). The head (34.4%) and the abdomen and pelvis (30%) were the more frequent CT exams performed, followed by CTA (19%). The 50% percentile of dose metrics (CTDIvol, DLP and SSDE) of our facility for each protocol group was calculated and compared with DRLs. Median dose metrics of chest, abdomen and pelvis, chest, abdomen and pelvis groups were lower than published national DRLs. CTDIvol and DLP from both head and neck or cervical spine exams were higher than DRLs, which indicated that these protocols may need further optimization to lower the dose.
Conclusion: The present study demonstrated that dose monitoring software provides a valuable tool for CT dose protocol optimization, as well as allowing for external or internal comparison of dose data.