ABSTRACT
Chow (2007) classified translational science/research into three areas, namely, translation in language, translation in information, and translation in (medical) technology. Translation in language is referred to the possible lost in translation of inform consent form or case report forms in multinational clinical trials. Lost in translation is commonly encountered because of not only difference in language, but also differences in perception, culture, medical practices, etc. A typical approach for assessment of the possible lost in translation is to first translate the inform consent form or the case report forms by an experienced expert and then translate back by a different experienced but independent expert. The back-translated version is then compared with the original version for consistency. If the back-translated version passes the test for consistency, then the back-translated version is validated through a small-scale pilot study before it is applied to the intended multinational clinical trial. Translation in information is referred to as bench-to-bedside in translational science/research, which is also known as translational medicine (TM). Translation in technology includes biomarker development and translation in diagnostic procedures between traditional Chinese medicine and Western medicine. In this chapter, we will focus on the statistical methods for translation in information and translation in technology. It must be noted that in practice, TM is often divided into two areas, discovery TM and clinical TM. Discovery TM includes biomarker development, bench-to-bedside, and animal model versus human model, while clinical TM includes translation among study endpoints, translation in technology, and the generalization from one target patient population to another.
