News & Events


Adaptive Trial Design in Clinical Trials: What You Need to Know: Part 2 in a 2-Part Series

Adaptive Trial Design in Clinical Trials: What You Need to Know: Part 2 in a 2-Part Series
21 November 2016

By: Dr. Greg Wei, Statistical Research Fellow

Adaptive designs (AD) when applied to clinical trials allow use of accumulated data to modify a trial after it has commenced. This enables the study to save time and resources, getting to go/no go decisions more quickly. Studies move through phases as the best patient cohorts, dosage and application are determined. Ultimately, the goal for adaptive trial designs is to assist in developing drugs that deliver superior efficacy and reduce serious adverse events (SAE).

Adaptive design for early clinical trials

Early clinical trials are designed to assist in understanding toxicities, find optimal doses, explore biomarkers and surrogate endpoints, and provide proof of concept.

Simon’s One-Arm Two Stage Adaptive Design

One-arm, two stage adaptive design is an approach with which most are familiar. The purpose of this design is to minimize the number of patients exposed to the trial drug and still maintain the power for hypothesis testing.  The Stage 2 is needed only if the hypothesis couldn’t either be rejected or accepted in Stage 1.

The following example demonstrates a situation where, in order to detect a difference of 0.2 in proportion with 80% power using a 5% test, the Stage 1 needs 20 subjects and the Stage 2 needs 15 subjects. If in the Stage 1, more than 12 (R1) subjects show ‘success’, the null hypothesis (the true proportion is not greater than 0.2) can be rejected, whereas if less than 4 (A1) subjects show ‘success’,  the alternative hypothesis can be rejected.  If neither happens, the Stage 2 will be needed. The corresponding cut-off values for the Stage 2 are given as R2 and A2.


Contact Synteract

Tell us how to stay in touch with you: