Room: Exhibit Hall | Forum 3
Purpose: Radiation biology studies require well-described irradiation conditions to be reproducible and interpretable. This work assesses the prevalence in reporting of experimental physics and dosimetry details in radiation biology as a whole over the last twenty years.
Methods: Abstracts were identified using PubMed with search criteria including animal studies/cells and ionizing radiation. Only original studies of biological effects caused by external radiation sources were included, yielding 1758 total articles. The methods section of each article were evaluated following the joint NCI, NIST, and NIAID recommendations on standards of reporting for experimental physics details in radiobiology studies â€“ as published by Desrosiers et al. in 2013. The presence of these experimental details were assessed and scored following a 10-point metric assigned to the following categories: radiation source and beam quality (3 points), dose and dose rate (3 points), dosimetry and calibration (2 points), and irradiation geometry (2 points).
Results: Most authors report the radiation source (92.5%) for the study, as well as the energy (81.4%), dose (99.0%), and dose rate (67.5%). Few authors report the dosimetry protocol used to calibrate the radiation source (3.8%), the field size (12.9%) or SSD (12.2%) used for irradiation. These results are similar to smaller scale reviews completed since the 2013 NIST symposium. The overall quality of physics reporting was quantified through the scoring metric. The majority of authors (75%) scored 6 points or less, with an average overall score of 4.81 Â± 1.6.
Conclusion: While most peer-reviewed articles report basic radiation source details, the majority fail to report further on dosimetry or geometry. These findings show a significant lack of basic physics reporting, which can lead to irreproducible or uninterpretable work, and hinder the translation of research into the clinic. This exposes the need for higher standards of peer-review to ensure research is reproducible.
Not Applicable / None Entered.