Last, we concluded that patent foramen ovale closure on clinical grounds was not justified. While we maintain this stance, we do accept that a trial using a different device, and powered on appropriate pilot data with different entry criteria, could yield a positive result. However, our data give no reason to believe so, and the variation that we observed leads us to suspect that the sample size required would be large. For this reason, we are not planning to do such a study ourselves. Of course, we agree with their final comment that additional data are needed.