Beyond non-inferiority: a practical trial design for optimising antibiotic treatment duration

There is substantial interest in investigating the use of shorter durations of antibiotic treatment for bacterial infections to counter the global threat of AMR. Classic designs compare two arbitrarily chosen durations in a non-inferiority trial. Given their important limitations (including arbitrary non-inferiority margins), our aim was therefore to develop new alternative designs to optimise duration of treatment. Our main idea was to recast the problem of determining an optimal duration as estimating the shape of the `Duration-Response Curve’ (DRC), by means of flexible regression modelling strategies.

Project Duration: 2017 – 2018
LSHTM staff: Matteo Quartagno, James Carpenter
Funding: MRC Hub for Clinical Trial Methodology
Collaborators: CTU