Systematic review on adult diagnostic reference levels in computed tomography, and analysis of regional variability

Background Computed tomography (CT) is vital in diagnostic imaging but involves ionizing radiation, requiring careful dose management. Diagnostic Reference Levels (DRLs), recommended by the International Commission on Radiological Protection (ICRP), serve as benchmarks for optimizing radiation doses and improving patient safety. While many DRL studies exist, most focus on individual countries, anatomical regions, or clinical indications, limiting broader applicability. This systematic review addresses this gap by analyzing adult CT DRLs across continents and major anatomical areas. It compares international benchmarks, highlights methodological inconsistencies, and explores drivers of regional variability to inform global dose optimization. Methods Following PRISMA-S guidelines, a systematic search of PubMed, Web of Science, Scopus, and Embase identified studies published in English from 2018 to 2025. Keywords related to DRLs and CT imaging were used. Studies reporting DRLs for adult CT using standardized dose metrics were included. Pediatric studies, narrative reviews, Phantom studies, case reports, pilot studies, and gray literature were excluded. Results Of 2719 identified articles, 63 met the inclusion criteria. Local DRLs were reported in 48 % of studies, and national DRLs in 52 %. Substantial variability was observed across examination types and geographical regions. For head CT, CTDI vol values ranged from 38.2 to 117.6 mGy, while DLP values ranged from 512 to 1943 mGy·cm across the included studies. Similar variability was observed for other examination categories, which may reflect differences in patient characteristics, equipment, and imaging protocols. Conclusion This systematic review identified substantial international variability in DRLs for adult CT examinations, associated with differences in imaging practices, patient characteristics, available technology, and methodological approaches to DRL establishment. The 75th percentile was the most commonly used method for defining DRLs across the included studies. These findings support regular DRL review and greater methodological standardization to improve comparability across regions and strengthen CT dose optimization practices.