Analysis of statistical methods for estimating standard radiation doses for patients undergoing computed tomography at the A.I. Burnazyan Federal Medical Biophysical Center

This study compares standard radiation dose estimates for patients undergoing computed tomography (CT) of major anatomical regions at the A.I. Burnazyan Federal Medical Biophysical Center using three different statistical approaches. A total of 4,854 CT examinations performed in 2024 were analyzed. For most anatomical regions, the standard dose calculated using Method 1 (entire patient cohort with an average body weight of 76–85 kg) was significantly higher than that obtained using Method 2 (patients selected with a body weight of 70 ± 5 kg), with statistically significant differences (p < 0.001).Comparison of doses calculated by Method 1 and Method 3 (linear regression modeling based on the entire cohort to estimate the dose for a "standard" 70 kg patient) revealed similar values, reflecting the influence of the overall weight distribution on the final dose estimates. When comparing Method 2 and Method 3—both applied to estimate the dose for a “standard” patient—it was found that Method 2 yielded lower values, with differences ranging from 1.1 to 1.5 folds. For example, for the most extensive combined anatomical regions, the effective dose per a non-contrast study was 6.3 mSv (Method 2) and 9.34 mSv (Method 3), and for the contrast-enhanced CT of the abdominal and retroperitoneal organs, 18.8 mSv (Method 2) and 20.11 mSv (Method 3). Conclusion: The application of three statistical methods for assessing standard effective radiation doses in CT was analyzed. The accuracy and comparability of standard dose estimates for the purpose of exposure control depend on the consistency of methodological approaches used to determine the need for dose optimization.