Colloquium: A Dose of Reality: How 20 years of incomplete physics and dosimetry reporting in radiobiology studies may have contributed to the reproducibility crisis
Speaker: Yannick Poirier, PhD, MCCPM, DABR, University of Maryland School of Medicine
Time: 2:00 p.m., Friday, Nov 8 (refreshments served at 1:45 p.m.)
Place: P-148 (refreshments across from SPS)
Abstract:
The majority of pre-clinical research fails to transition successfully to clinical practice, due in part to poor reproducibility. This has also been the case in radiation biology studies, which are often used to develop new cancer therapies in Radiation Oncology. The precise details of an experimental irradiation protocol of a radiation biology experiment can have a dramatic impact on the biological effects induced by the radiation. In the first large-scale study of its kind, our team at the University of Maryland analyzed the methods of 3542 manuscripts from 469 journals to see if the experimental irradiation protocol was sufficiently well-described for other independent researchers to understand and replicate the experiment. However, we found that the vast majority of papers described few if any of the essential experimental details. More significantly, highly-cited articles in high impact factor publications were statistically more likely to have poorly described protocols than publications in low-visibility technical journals. These findings show a crucial deficiency in the reporting of basic experimental details, and severely impact the reproducibility and translatability of a large proportion of radiation biology studies. While the scientific findings of these studies are not under question, the ability of other teams to understand and contextualize these results is diminished if the experiment cannot be understood or replicated.