Background and Aims The treatment of acute ischemic stroke (AIS) still has multiple equipoises and the current time-consuming and inefficient process of clinical translation is a major hindrance. Bayesian approaches offer an alternative method that can accelerate the development of novel treatments. These designs allow multiple treatment options to be tested simultaneously and these features significantly reduce sample size and allow for faster and more efficient clinical translation. Unfortunately, unlike other fields, Stroke researchers are yet to adopt these methodologies.

Methods We propose the “leapfrog design” to evaluate multiple treatments simultaneously. This allows continual treatment improvement and finding one or more treatments that are better than the current ‘gold standard’ which is accomplished by allowing replacement of the comparison arm (the arm against which all other arms are compared) by new treatment arms that achieve a prespecified threshold for demonstrating superiority. For the hypothetical study we simulated a 4-arm study involving – a) Nerinetide b) TRPM7 c) SRC-ND2 and d) uric acid in patients undergoing thrombectomy. We used the BFDA – an R package for Bayes factor (BF) design analysis.

Results We chose Cohen’s d = 0.4, as in early stage of development, with a BFfail of 1/4, and BFsuccess of 5. The Nmin and Nmax were found much smaller than traditional RCT numbers.

Conclusions This will be a much-needed paradigm shift for further progress in stroke research beyond current equipoises.