5 Ways to Optimize Trial Designs

1 - Better Design

• Every design choice will have some effect on the required sample size
• Good designs are efficient and thus reduce sample size
• Design decisions include - randomization level (Subject level Vs Cluster level etc)
  • Crossover vs parallel (Crossover design can be more efficient if chronic disease)
  • Choice of endpoint - closer endpoint is to actual measure the better 
  • Some unneeded simplifications increase costs - dictomisation, treating TTE as binary, recurrent events as TTE
  • Covariate Selection should be part of the study design and protocol

2 - Real World Evidence

• Using RWE can help researchers determine a viable study length and frequency of study events when designing their trial
• Enable researchers to assess site feasibility from previous data from site-specific clinical trials that were successful
• Create Synthetic Control Arms and therefore allow all recruited patients to be in the active arm
• Trial planners can assess and model protocols to reduce potential patient issues which lead to dropout, prior to commencing the trial

3 - Bayesian Assurance

• Frequentists can gain greater insight and ideally make better decisions to complement their frequentist trial with Bayesian assurance
• Take advantage of prior information and expert opinion
• Bayesian (Power) Assurance - “True Probability of Success” very useful at financial & scientific Review boards.
• Can help you formalize your sample size sensitivity analysis

4 - Adaptive Trials 

• Adaptive Designs can lead to more efficient clinical trials
• Group Sequential Design & Sample Size Re-estimation facilitates interim analyses where you can either stop for benefit or futility or increase your sample size if the effect size is promising
• Other Adaptive areas include enrichment, seamless trials and dose finding
• Adaptive Trials can reduce costs and risks and result in earlier decisions
• Patients may also favour enrolling in adaptive trials because of the increased probability they will receive more effective treatment

5 - Master Protocols

• The FDA guidelines describe two types of master protocol designs. 
• The basket trial design tests a single oncology drug in various populations 
• The umbrella trial design is used to evaluate multiple investigational oncology drugs in a single disease population
• Ability to “bundle” treatments and garner economies of scale (reduce paperwork and duplication of activities)
• From the perspective of the sites there can be a great reduction in overhead resulting in significant cost savings
• Evaluate safety and efficacy of treatments in parallel instead of sequentially
• Reduce patient exposure to control arms

Suggested Reading

A guest blog post from .

(Updated July 28th 2020)