CT Dose Optimization in Eastern Libya: A Prospective Multicenter Audit of Examination-Specific Variation in CTDIvol, DLP, and Derived Scan Length

Abstract

CT dose depends partly on scanner design, but day-to-day practice is also shaped by how each service applies its protocols, including scan start and end points, additional phases, and local review against diagnostic reference levels (DRLs). This prospective multicenter assessment assessed examination-specific variation in volume CT dose index (CTDIvol), dose-length product (DLP), and derived scan length across CT services in Eastern Libya. The final cleaned dataset included 229 adult CT examinations from four hospitals. Albyda Medical Center was analyzed as two separate service contexts because the Casualty and In-Patient CT units followed different workflows. Examination names were standardized before analysis. As direct z-axis scan range was unavailable, derived scan length was calculated as DLP/CTDIvol and used only as a practical indicator of longitudinal coverage, rather than as a direct anatomical measurement. Dose indicators were summarized using medians and interquartile ranges, and site comparisons were performed within matched examination or protocol groups. Chest CT and abdomen/abdomen-pelvis examinations showed significant inter-site differences in DLP and derived scan length, while brain CT differed mainly in CTDIvol. Paranasal sinus CT and the local CTU/KUB protocol group showed significant variation across CTDIvol, DLP, and derived scan length. Local 75th percentile DLP values were above selected external comparators for brain CT, chest CT, HR chest, and paranasal sinus CT. These findings support targeted protocol review, tighter control of scan range, and repeated dose audit linked to DRL-based optimization.