PFO and Scuba Diving: Who Should Get Tested and What Comes Next

What is a PFO and how common is it?

A patent (or persistent) foramen ovale is a small flap-like opening between the two upper chambers of the heart that failed to seal after birth. Before birth, the foramen ovale lets blood bypass the lungs. When the baby starts breathing, pressure in the left atrium rises and pushes the flap shut, and it normally fuses soon after. DAN puts the share of people in whom it never fully fuses at about 27%, citing a Mayo Clinic study that found no difference between men and women. StatPearls gives a range of 20-34% of adults.

Most PFOs are small. In a post-mortem series of 263 people used as a comparison group in a 2015 study, the median PFO diameter was 5 mm, and only 1.3% measured 10 mm or more. That size difference turns out to matter more for divers than the simple presence of a hole.

This article summarizes what diving medicine bodies publish. It is general information, not medical advice. Decisions about testing or closure belong with a physician trained in diving medicine, together with a cardiologist. More articles on health and diving are in our safety guides.

Why a PFO matters for decompression sickness

Bubbles in the veins after a dive are common, and the lungs normally filter them out. A right-to-left shunt lets some of them skip that filter. If pressure in the right atrium rises above the left, blood and any bubbles in it can cross the PFO straight into the arterial circulation. Cardiologist Peter Wilmshurst's 2015 review describes what happens next: the bubbles reach tissues that are still supersaturated with nitrogen and grow there, a process he calls peripheral amplification.

DAN lists four conditions that have to line up: a large PFO, venous bubbles after the dive, something that opens the shunt (straining, lifting or a forceful Valsalva, for example), and bubbles arriving at a tissue while it is still supersaturated. That explains why shunt-related DCS tends to hit the brain, spinal cord, inner ear and skin, and why it often follows an ordinary dive. StatPearls describes the typical case as sudden, severe symptoms after a dive that followed an established decompression protocol with no other risk factors. More than one such episode should raise suspicion of a shunt.

Joint pain on its own is not among the DCS types linked to shunts. If you are unsure what DCS looks like, our guide to decompression sickness symptoms and first aid covers the signs and the emergency steps.

How much does a PFO raise the risk of DCS?

The relative risk goes up several times, but the absolute risk stays low, and most of it sits with the largest shunts. Estimates range from 2.5 to about 5 times the odds of serious DCS. In absolute terms, the 230-diver cohort in the table found about 5 major DCS events per 10,000 dives in divers with a PFO.

Those low absolute numbers are the argument against testing every diver. The 1998 analysis in the table concluded that the risk of DCS with a PFO remains small and does not warrant routine echocardiography of sport, military or commercial divers.

Should divers be tested for PFO?

Not routinely. The current reference is the 2025 update of the joint position statement by the South Pacific Underwater Medicine Society (SPUMS) and the United Kingdom Diving Medical Committee (UKDMC), which came out of a SPUMS workshop in May 2024 and replaces the 2015 version. Its first statement: routine screening for a right-to-left shunt at an initial or periodic dive medical is not indicated. The UHMS guidance for physicians agrees that routine screening of all diving candidates is not recommended.

Compared with 2015, the update adds cardiovascular DCS and cryptogenic stroke to the reasons for testing and covers atrial septal defects and lung shunts as well as PFO. DAN's guideline page, built on the 2015 statement and a 2015 DAN workshop, is narrower: it suggests testing after more than one DCS episode of the relevant type, and says mild DCS without skin signs, or a headache on its own after diving, is not a reason to test. The 2025 statement lists five situations in which testing should be considered:

• A history of cerebral, spinal, vestibulocochlear (inner ear), cardiovascular or cutaneous (skin) decompression sickness.
• Migraine with aura.
• A cryptogenic stroke, meaning a stroke with no identified cause.
• A PFO or atrial septal defect in a first-degree relative.
• Congenital heart disease.

How the PFO test works and how results are graded

The recommended test is a bubble contrast echocardiogram, ideally transthoracic (a scan through the chest wall, no sedation). A bubble contrast agent is injected into a vein while the diver performs manoeuvres that promote right-to-left flow, a Valsalva release and a sniff, and the scan watches for bubbles reaching the left side of the heart. The 2025 statement says a standard echo with colour Doppler but no bubble contrast is not adequate, and that testing should be done by centres well practised in the technique. Before the test, the consequences of a positive result should be discussed as part of consent.

A spontaneous shunt, or a large shunt with provocation, is an unequivocal risk factor for the types of DCS listed above. Medium shunts carry a lower, poorly defined risk. A small shunt has to be read against the reason for testing, and after a relevant DCS episode it may need further investigation to confirm it really is small. Shunts through the lungs also cause a larger share of shunt-related DCS than expected, and the statement wants advice on those to come from a cardiologist with diving expertise.

One caution from DAN: finding a PFO after an episode of DCS does not prove that the PFO caused it.

Options after a PFO is found

The 2025 statement gives a diver with a risk-relevant shunt three options, to be weighed with a diving physician against the reason for testing and the diver's future plans: stop diving, which it calls the safest; close the shunt if it is in the atrial wall; or dive more conservatively. It adds a consensus view that dives deeper than 15 m raise the risk of bubbles and therefore of shunt-related DCS. StatPearls adds that closure should not be routinely considered in a diver with an incidental PFO and no DCS history, who should instead be counselled to dive conservatively.

The conservative strategies named in the 2025 statement and in DAN's guideline:

• Keep dive times well inside the no-decompression limit.
• Stay shallower than 15 m (49 ft).
• Make only one dive a day.
• Breathe nitrox but plan with air limits.
• Lengthen the safety stop or shallow decompression stops on purpose.
• Avoid heavy exercise, unnecessary lifting or straining, and forceful Valsalva manoeuvres for at least three hours after diving. The statement rates this one as having a less established basis.

Worked example: nitrox with air limits at 14 m

Here is what the nitrox advice looks like in numbers. A diver told to stay shallower than 15 m plans one dive to 14 m. On the DECO 2000 table, which rounds depth up to the next row, the air no-decompression limit is read from the 15 m row: 72 minutes. The advice is to stay well inside that, not to use all of it.

Now the diver breathes EAN32 but keeps the air plan. Our nitrox calculator gives an equivalent air depth of 10.7 m for EAN32 at 14 m, rounded up. The nitrogen the body takes up matches a dive about 3 m shallower than the one the air limit assumes, and that difference becomes margin instead of extra bottom time. The oxygen partial pressure at 14 m on EAN32 is 0.77 bar, far below the 1.4 bar working limit. These tools help with planning and do not replace training, a dive computer or the diving physician's advice for your case.

What the evidence on closure and conservative diving shows

Both approaches are backed by observational data only, and the numbers favour both. There are no randomised controlled trial data on PFO closure to prevent DCS, and a 2021 European position paper on PFO, written with the European Underwater and Baromedical Society, described the evidence for these conditions as of low or very low certainty.

For closure: one study followed 104 divers with a history of major DCS for about 5 years. Divers chose their own group, so it was not randomised. Major neurological DCS occurred at 0.5 per 10,000 dives after closure against 35.8 per 10,000 in divers who kept an open PFO. A 2022 meta-analysis of four observational studies (309 divers) found DCS in 2.84% after closure against 11.3% without it, but adverse events in 7.63% of closures, nearly one in thirteen, including heart rhythm problems and bleeding. In a UK audit of 105 divers closed by one cardiologist, 81 of the 95 with follow-up echo returned to unrestricted diving and 2 needed treatment for atrial fibrillation.

For conservative diving: another study interviewed 27 divers who had been advised to reduce their nitrogen load after DCS. They had 34 DCS events in 17,851 dives before the advice and 4 in 9,236 dives after it, a drop seen with and without a shunt. It was a retrospective telephone survey, so it cannot rule out recall bias, and the 15 m depth figure in the 2025 statement is expert consensus, the lowest evidence grade it uses. DAN adds that closing a PFO after DCS gives no assurance that DCS will not happen again.