Adaptive Trial Design Playbook
Build the trial that adjusts as data accumulates.
A practical playbook for adaptive trial design — sample-size re-estimation, arm dropping, dose finding, seamless phase trials and Bayesian designs.
"Should we run an adaptive design — and what kind?"
Adaptive designs trade upfront design complexity for in-trial flexibility. They work well in dose finding, biomarker enrichment and seamless P2/3 trials.
Adaptive design is now mainstream. Best teams pre-specify rules rigorously, engage regulators early and use SSR, arm-dropping and seamless P2/3 designs to compress timelines.
What we’re seeing in the data.
Bayesian designs gain regulatory acceptance
Especially for early-phase and rare disease.
Sample-size re-estimation is low-risk
Most teams under-use it.
Arm-dropping speeds dose finding
Can cut trial size 30–50%.
How to think about it.
-
01
Define adaptation rules
Pre-specified, rigorous.
-
02
Choose adaptive type
SSR, arm drop, seamless, Bayesian.
-
03
Plan SAP
Adaptive-specific statistical plan.
-
04
Engage regulators early
Type B / scientific advice.
What separates a good answer from a defensible one.
IRT, randomization, blinding.
Must control rigorously.
Fixed-design teams resist.
Where the signal comes from.
Common questions.
Will FDA accept Bayesian P3?
Increasingly yes, especially with pre-specified rules.
When should we avoid adaptive?
When the science doesn’t support flexibility — e.g., simple safety endpoints.
Want this answered on your data?
We build decision systems on top of analyses like this — so the next question takes minutes, not weeks.
Talk to a strategist