In the medical device industry it is common to run a clinical trial to compare a new device to an approved device. For this purpose, a Bayesian hierarchical model is useful for incorporating historical data on the approved device as a control, potentially decreasing the sample size needed in the control arm in the new clinical trial. In this case study, I present the application of such a model to a randomized, controlled, non-inferiority trial in a regulatory setting. A more conservative Bayesian ‘conditional borrowing’ strategy was employed. This approach restricted the scenarios under which historical data could be borrowed to those that would not bias in favor of the new device. This design resulted in a reduced sample size of 25% of patients compared to the frequentist alternative. The statistical challenges in this trial design and the unblinded results will be presented.