Personalised randomised controlled trial designs, or PRACTical designs, allow researchers to compare multiple treatments for diseases that don’t have a single “standard-of-care” treatment that every patient could take.

What is the PRACTical design and why is it relevant?

In certain clinical areas such as multi-drug resistant infections, multiple treatments are potentially useful, but there is no single “standard-of-care” that every patient could take. For example, one of the available antibiotic treatments might be unsuitable for patients with additional medical conditions or for patients with infections that are already resistant to this type of antibiotic.

To tackle this problem, we have proposed a personalised randomised controlled trial (PRACTical) design. In a PRACTical design, patients are randomised only among treatments that are suitable for them (see Figure below). As a result, the trial produces a ranking list that can inform clinical decisions about which treatments are likely to perform best for each patient.

Using the PRACTical design, researchers can carry out clinical trials and evaluate which treatments work best in diseases where standard trial designs cannot easily be used.

In a PRACTical design, patients are randomised only among treatments that are suitable for them. The figure below illustrates randomization for a patient who could take any of treatments A, D, H or I.

What have we done?

We have implemented the PRACTical design in our NeoSep1 trial. This trial is comparing multiple antibiotic treatments for treating sepsis in newborn babies, including some treatments that are routinely used and some new combinations of existing antibiotics.

PRACTical trial designs in the areas of snakebite, carbapenem-resistant infections and multi-drug-resistant typhoid are currently under development.

How will this make a difference?

We are developing the methodology for designing, analysing and implementing PRACTical trials.

Our work provides guidance on how PRACTical designs can be used in practice, including the statistical methods for evaluating the design, and the methods for determining the sample size.




Recordings from the "Innovative Phase III trial designs Symposium" (June 2023) - YouTube