Table 2.

Adaptive sequences of trials

What is adaptive	Based on	Example	Benefit	Special trial requirement	Caution	Reference
Decision to conduct a full-size trial	Pilot trial	Pilot trial result determines whether intervention is worth testing with a fully powered trial	Avoid funding unworthy programs	Cannot use for long-term outcomes	Sample size and power can be incorrect	Kraemer et al. 2006 (78)
Merging multiple stages of research	Combining research questions	Roll-out trial Dynamic wait list	Ability to evaluate as a program is implemented	Community wants full implementation and is willing to use wait listing	Challenges in implementing program with sufficient strength	Brown et al. 2006 (25)
Reusing a trial sample in a new trial	Developmental appropriateness of further intervention	Rerandomize subjects to a second trial	Ability to examine early or late intervention	Maintain low attrition	Differential attrition	Campbell et al. 1994 (27)
Screening of intervention components	Fractional factorial designs and analysis	Multiphase optimization strategy	Increased efficiency	Large samples are required	Can fail if there are high-level interactions between components	Collins et al. 2007 (35)
Modification of intervention as it is implemented	Sequence of small experiments that can be combined	Cumulative adaptive trial	Streamlined improvement of the intervention	Sufficiently strong mediators or proximal targets to demonstrate effectiveness	Potential effect on Type I error	Brown et al. 2007 (24)