Abstract
Meta-analyses have been conventionally performed to extract the firmest conclusions from randomized controlled trials while minimizing the risk of bias. However, the field of oncology does not always allow for collecting the best evidence. Radiation oncology is a discipline where intractable or rare diseases are commonly encountered; hence, more practical data suitable for detailed clinical evaluations are needed. This review discusses new viewpoints regarding meta-analyses by pointing out heterogeneities among clinical studies and issues related to analyzing observational studies, thus clarifying the practical utility of meta-analyses in radiation oncology. Limitations of previous systematic reviews or meta-analyses are also assessed to suggest future directions.