Although randomisation ensures the elimination of selection bias the ‘standard’ two armed trial is not without its difficulties. Many participants have a treatment preference, which can introduce post-randomisation bias. Attrition is a problem in most trials with 20% or more attrition not being uncommon. Similarly trials can be slow and expensive.
A TWICs design can potentially address some of these problems. Essentially a TWIC is a RCT that is embedded within a larger cohort. Elements of the design can mean that attrition is lower, participant preferences are dealt with, costs are lowered due to the use of unequal allocation and existing cohorts. In this talk I will discuss the basic elements of the design and the experience of the York Trials Unit in its use.