Design aspects for prognostic factor studies

Prognostic research is clinically relevant and ultimately facilitates stratified medicine. However, its quality and output are limited. More guidance is needed to improve understanding and thus quality. On behalf of the topic group ‘TG5: study design’ of the STRATOS initiative and for the general readership, this article describes key concepts and issues for prognostic factor studies, a sub-area of prognosis research. After providing a general overview on prognosis research, the article covers aspects such as aims, estimands and designs of prognostic factor studies, highlighting standards and current practice. Focusing on prognostic factor studies that assess a single factor at a time and a binary outcome, this article is complemented by a glossary of terms and a list of general aspects to consider in prognostic factor studies.