Differences in radiation dose for computed tomography of the brain among pediatric patients at the emergency departments: an observational study

Background Computed tomography (CT) is associated with a risk of cancer development. Strategies to reduce radiation doses vary between centers. We compared radiation doses of CT brain studies between pediatric and general emergency departments (EDs), and determine the proportion studies performed within the reference levels recommended by the International Commission on Radiological Protection (ICRP). Methods A retrospective review was carried out in a healthcare network consisting of one pediatric ED and three general hospital EDs. Pediatric patients less than 16 years old with CT brain studies performed between 1 January 2015 and 31 December 2018 were included. Information on demographic, diagnosis, volume-averaged computed-tomography dose index and dose length product (DLP) were collected. Effective dose was then calculated from DLP using conversion factors, termed k-coefficients which were derived using a 16 cm head CT dose phantom. Results Four hundred and seventy-nine CT brain studies were performed - 379 (79.1%) at the pediatric ED. Seizure (149, 31.1%), head injury (147, 30.7%) and altered mental status (44, 9.2%) were the top three ED diagnoses. The median effective dose estimates were higher in general than pediatric EDs, particularly for those aged > 3 to <= 6 years old [1.57 mSv (IQR 1.42-1.79) versus 1.93 mSv (IQR 1.51-2.28), p = 0.047], > 6 to <= 10 years old [1.43 mSv (IQR 1.27-1.67) versus 1.94 mSv (IQR 1.61-2.59), p = 0.002) and > 10 years old (1.68 mSv (IQR 1.32-1.72) versus 2.03 mSv (IQR 1.58-2.88), p < 0.001). Overall, 233 (48.6%) and 13 (2.7%) studies were within the reference levels recommended by ICRP 60 and 103 respectively. Conclusions Radiation doses for CT brain studies were significantly higher at general EDs and less than half of the studies were within the reference levels recommended by ICRP. The development of diagnostic reference levels (DRLs) as a benchmark and clinical justification for performing CT studies can help reduce the radiation risks in the pediatric population.