A physiological left-to-right shunt through a patent foramen ovale (PFO) can be demonstrated in 32–55% of young adults with unexplained stroke compared with 10–35% of the general population.1 Transoesophageal or transthoracic echocardiographic evidence of a PFO consists of demonstration of a defect in the wall between the atria and ideally the passage of bubble contrast from right to left at this level. Provocative measures may be used to raise right atrial pressure, such as coughing, Valsalva manoeuvre or pressure on the liver, and increase the detection rate of PFO by at least 25%, but are not always performed rigorously or even routinely in children. The characteristics of the PFO, such as the size of the defect and of the shunt, are also important considerations; associated atrial septal aneurysm, larger defects and those with a greater right-to-left shunt may carry a higher risk of recurrence, particularly in patients with cryptogenic stroke. However, prospective population-based studies have failed to confirm the association (even with atrial septal aneurysm), a recent study failed to find any link with recurrent stroke2 and there are very few data in children. Despite this, studies of the effect of closure of PFO detected echocardiographically with devices placed at catheterisation are in progress,3 and such procedures have been advocated in children who have had a stroke.4

Transcranial Doppler (TCD) with bubble contrast and Valsalva appears to be a more sensitive technique than echocardiography for detecting PFO in adults.5 Benidik et al's careful bubble contrast TCD study published in this issue6 (in …