Local diagnostic reference levels and optimisation priorities for paediatric oncology computed tomography at Gaza’s sole specialist paediatric hospital

BACKGROUND Published evidence on paediatric computed tomography (CT) dose indicators and diagnostic reference levels in Palestine is limited, particularly for children with cancer who may undergo repeated imaging.Local diagnostic reference levels provide practical optimisation benchmarks where national paediatric reference levels are unavailable.We therefore examined whether examination-specific and size-specific dose distributions could identify actionable local optimisation priorities.AIM To establish paediatric oncology CT local diagnostic reference levels and identify actionable dose-optimisation priorities at Gaza's sole specialist paediatric referral hospital.METHODS This retrospective cross-sectional examination-level dose survey analysed 193 paediatric oncology CT examinations performed between March 2019 and December 2019.Age-specific and protocol-specific P75 values of scanner-recorded CTDIvol and dose-length product (DLP) were proposed as local dose audit benchmarks; P50 values and non-parametric bootstrap 95%CIs were also calculated.P75 estimates from groups of 4-9 examinations were treated as provisional.Additional descriptive analyses assessed patient-size grouping, scan range and protocol variability.Modelderived size-specific dose estimates, effective-dose and organ-dose estimates were not included because their computational inputs and conversion factors could not be recovered. RESULTSThe principal CTDIvol P75 values were 23.46-23.73mGy for brain CT, 2.44-3.15mGy for chest CT, 4.68-6.78mGy for abdomen CT and 4.66-7.50mGy for 4 / 43 chest/abdomen/pelvis CT in children aged 0-10 years.Provisional values for older small subgroups are reported separately.Chest/abdomen/pelvis CT demonstrated substantial DLP variability and extensive anatomical coverage, although brain CT had higher DLP in the two younger age groups.Confidence intervals were wider in smaller groups; an examination-level audit identified some discordance between recorded scan length and DLP/CTDIvol-derived length.Secondary associations were descriptive rather than evidence of avoidable dose. CONCLUSIONThese scanner-specific CT dose benchmarks provide a basis for local dose audit.Review of patient-size adaptation and scan range should be informed by clinical indication and image quality.Small-subgroup estimates remain provisional and patient-level dependence could not be quantified.