Registry-based trials: a potential model for cost savings?