Design of Experiments > Adaptive design clinical trial
What is an Adaptive Design Clinical Trial?
An adaptive design clinical trial (also called adaptive randomization) is any design that allows adaptations to a clinical trial as it progresses. This type of trial is strongly recommended by the FDA for several reasons, including that it increases the odds a patient will receive a beneficial drug. Trials tend to complete faster, and are more efficient—making the best use of available resources (Mahajan & Gupta, 2010), shortening the trial length, and using fewer patients (FDA, 2010).
The trial modifications, based on accumulating data, are made without any threat to the trial’s validity (i.e. a test accurately measures what it’s supposed to) or integrity (the data is well managed and reflects the trial results). At each stage of the study, data is analyzed and a decision is made to continue or stop the trial. If the trial continues, modifications might be made to a number of different areas, such as diagnostic procedures, eligibility criteria study dose.
Adaptive trials are sometimes called flexible designs. While the two terms can be used synonymously, the term “flexible design” is informal, and can refer to a wider variety of non-rigid designs. On the other hand, an adaptive design clinical trial has a very precise definition.
According to the FDA, an adaptive design clinical trial is:
“…a study that includes a prospectively planned opportunity for modification of one or more specified aspects of the study design and hypotheses based on analysis of data (usually interim data) from subjects in the study” (FDA, 2010).
Prospectively planned means that any changes must be made before the trial data is examined. In addition, changes must be made based on data itself, and not from an outside source. For example, results from a separate trial might indicate a particular drug is not effective. Making changes based on this fact alone does not make a trial “adaptive”, although adaptive designs do sometimes use a combination of information gleaned from inside the trial and outside sources.
Adaptive Design Clinical Trial Rules
An adaptive design clinical trial has one or more “rules”, which guide how the trial is conducted (Krams et. al, 2006). Different rules apply to the different subsets of adaptive design. For example, adaptive allocation, sample size re-assessment, or group sequential have different rules:
Allocation rule: how participants are allocated to different treatments. Adaptive allocation designs only use an allocation rule. At the end of each stage, the trial is tweaked so that the maximum number of patients receive the beneficial treatment.
Sampling rule: how many participants are sampled at each stage. Sample size re-assessment designs only use the sampling rule. These designs modify the sample size as the trial progresses to achieve the desired statistical power. The statistical power of a study (sometimes called sensitivity) is how likely the study is to distinguish an actual effect from one of chance. See: Adaptive sampling.
Stopping rule: what will cause the trial to be stopped (for example, an ineffective treatment). Group sequential designs only use the stopping rule.
Decision rule: any rules included in the study that don’t fall into one of the previous three categories.
Advantages and Disadvantages
According to Chow and Chang (2008), adaptive designs have several advantages over traditional designs, which are “fixed” at the outset.
- The design is ethical, as it allows for a trial for ineffective or toxic treatments to be stopped part-way. By the same token, patients can also be funneled into more beneficial treatments.
- The design more closely matches the “real world” of medicine, where patients continue to enter more promising trials and cease to partake in trials with ineffective treatments.
- Trials are more efficient. An efficient trial maximizes use of materials, time and energy.
However, adaptive design do have some notable disadvantages:
- P-values and confidence intervals related to the treatment effect may be unreliable after the modifications are made.
- The modified trial might be completely different from the one originally intended; the original study question might not be answered with any accuracy, if at all.
Chow, S. and Chang, M. (2008). Adaptive design methods in clinical trials – a review. Orphanet J Rare Dis. 2008; 3: 11. Retrieved December 4, 2017 from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2422839/
Cook, T. (2005). Introduction to Statistical Methods for Clinical Trials (Chapman & Hall/CRC Texts in Statistical Science) 1st Edition. Chapman and Hall/CRC
Friedman (2015). Fundamentals of Clinical Trials 5th ed. Springer.
Krams, M. et. al (2006). Adaptive Clinical Trials. Posted on the American Statistical Association website. Retrieved December 4, 2017 from: https://ww2.amstat.org/meetings/fdaworkshop/presentations/2006/phrma_adapt_fda_sept27_09122006.ppt
FDA (2006). Innovation or Stagnation: Critical Path Opportunities List. Washington DC, USA. Food and Drug Administration. Retrieved January 15, 2018 from: http://www.fda.gov/downloads/ScienceResearch/SpecialTopics/CriticalPathInitiative/CriticalPathOpportunitiesReports/UCM077258.pdf
FDA (2010). Guidance for industry: Adaptive design clinical trials for drugs and biologics. Washington DC, USA: Food and Drug Administration. Retrieved January 15, 2018 from: http://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM201790.pdf
Mahajan, R. & Gupta, K. (2010) Adaptive design clinical trials: Methodology, challenges and prospect. Indian J Pharmacol. 2010 Aug; 42(4): 201–207. Retrieved January 15, 2018 from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2941608/