Atrial Septal Defect

Definition

Atrial septal defect (ASD) is a direct communication between the left and right atria, resulting in a left-to-right shunt whose magnitude depends on defect size and relative ventricular compliance. (sources/asd-ehj-2022 — high; sources/asd-lancet-2014 — high)

Key Concepts

Epidemiology

• Incidence ~56–100/100,000 livebirths; third most common type of CHD overall (sources/asd-ehj-2022 — high; sources/asd-lancet-2014 — high)
• Most common CHD diagnosed in adulthood (25–30% of new adult CHD diagnoses), owing to slow clinical progression and frequent asymptomatic presentation in early life (sources/asd-ehj-2022 — high)
• Sex distribution by subtype: ~65–70% of secundum, ~50% of primum, ~40–50% of sinus venosus defects are female (sources/asd-lancet-2014 — high)

Anatomy and Classification

• Secundum ASD (80%): Fossa ovalis; most amenable to percutaneous device closure (sources/asd-ehj-2022 — high)
• Primum ASD / Partial AVSD (15%): Adjacent to AV valves; near-always abnormal AV valve morphology; surgical repair required; more prone to AV block and AV valve regurgitation (sources/asd-ehj-2022 — high)
• Sinus venosus ASD: Superior type (~5%) at SVC–right atrium junction with frequent partial anomalous pulmonary venous drainage (PAPVD); inferior type (<1%) at IVC junction; predominantly surgical but covered stent catheter technique emerging for superior type (sources/asd-ehj-2022 — high)
• Coronary sinus defect (rarest): Unroofing between coronary sinus and left atrium; strongly associated with persistent left SVC (Raghib syndrome) (sources/asd-ehj-2022 — high)
• Common atrium (very rare): Absent septum primum + septum secundum + AV canal septum; often associated with heterotaxy syndrome (sources/asd-lancet-2014 — high)
• Distinguished from PFO: PFO is not a true deficiency of atrial septal tissue (tunnel communication); complete anatomical closure occurs in 70–75% of adults; does not require closure for ASD indications (sources/asd-ehj-2022 — high; sources/asd-lancet-2014 — high)

Genetics and Aetiology

• Most sporadic; genetic syndromes: Down (~80%), Holt-Oram (66% have ASD), Ellis van Creveld (~60%), Noonan (~20%) (sources/asd-ehj-2022 — high; sources/asd-lancet-2014 — high)
• Implicated genes: NKX2-5 (also linked to AV block), GATA4, TBX5, MYH6 (chromosome 14q12) (sources/asd-ehj-2022 — high; sources/asd-lancet-2014 — high)
• Maternal exposures: alcohol, smoking, SSRIs/antidepressants, diabetes, high dietary glycaemic index, advanced maternal age ≥35 years (sources/asd-ehj-2022 — high; sources/asd-lancet-2014 — high)

Pathophysiology

• Left-to-right shunt → right atrial and RV volume overload → RV dilatation, myocardial stretch, increased pulmonary blood flow
• Shear stress on pulmonary endothelium → activation of growth factors, vasoconstrictors, smooth muscle hypertrophy → PAH (pulmonary vascular disease relatively rare, but risk increases with age and large defects)
• LV volume unloading and adverse ventricular–ventricular interaction also occur (not purely a right-heart lesion) (sources/asd-ehj-2022 — high)
• As LV stiffens with age (hypertension/CAD), degree of L→R shunting increases further (sources/asd-ehj-2022 — high)

Natural History

• Life expectancy without closure (Campbell, untreated large secundum): annualised mortality 0.6–0.7%/yr in first two decades; rising to 4.5%/yr in 4th decade; 7.5%/yr in 6th decade (sources/asd-lancet-2014 — high)
• Sinus venosus and primum defects do not decrease in size; almost always require surgical closure (sources/asd-lancet-2014 — high)
• Secundum spontaneous closure rates (Hanslik 2006, n=200, median age 5 months):
  • 4–5 mm: 56% spontaneous closure
  • 6–7 mm: 30% spontaneous closure
  • 8–10 mm: 12% spontaneous closure

  • 10 mm: 0% spontaneous closure

• Change in defect size over time: initially ≤4 mm → 70% decrease; initially >8–12 mm → 76% increase (sources/asd-lancet-2014 — high)
• Cardiac remodelling after closure: Begins immediately but continues ≥1 year; extent of RV reverse remodelling inversely related to age at closure (sources/asd-ehj-2022 — high)
• RVESVI >75 mL/m² at closure → persistent RV dilatation, residual TR, elevated BNP, RV dysfunction (sources/asd-ehj-2022 — high)
• Long-term survival after surgery vs general population (Murphy, 27–32 year follow-up):
  • Operated <12 yr: 98% vs 99% (near-normal)
  • Operated 12–24 yr: 93% vs 97%
  • Operated 25–41 yr: 84% vs 91%
  • Operated >41 yr: 40% vs 59% (sources/asd-lancet-2014 — high)

Diagnosis

• First-line: TTE (2D + colour Doppler); RV enlargement with normal/hyperdynamic function = haemodynamically significant ASD
• TOE: Near-universally performed after TTE; essential for sinus venosus defects, suboptimal TTE windows, intraprocedural guidance
• 3D echocardiography: Defect characterisation, rim assessment, device sizing
• CMR: Gold standard for ventricular volumes and Qp/Qs measurement; indispensable for inferior sinus venosus defects and when echocardiography leaves unanswered questions; non-invasive gold standard for absolute Qp and Qs measurement — see concepts/Intracardiac-Shunts
• Transpulmonary ultrasound dilution (COstatus): In critically ill or surgical paediatric patients who cannot undergo MRI, this minimally invasive bedside method estimates Qp/Qs via dilution curve asymmetry using existing arterial + central venous catheters; sensitivity 95.7%, specificity 97.6% for L→R shunt detection, but underestimates Qp/Qs magnitude in moderate/small shunts (sources/shunt-nature-sr-2020 — medium); see concepts/Intracardiac-Shunts
• Cardiac CT: High-resolution anatomy at <1 mSv; when echocardiography or CMR insufficient
• Cardiac catheterisation: NOT routine; mandatory when PH suspected (sPAP estimate >40 mmHg), LV dysfunction, or CAD evaluation needed (sources/asd-ehj-2022 — high)
• Qp:Qs — limitations of a relative metric:
  • Qp:Qs ≥1.5:1 is the standard criterion for haemodynamically significant ASD; influenced by defect size and ventricular compliance (sources/asd-ehj-2022 — high)
  • Critical limitation: Qp:Qs is a relative index — does not capture absolute pulmonary blood volume, which may be the more pathophysiologically relevant metric (sources/asd-japc-2025 — medium)
  • Illustrative example: Two patients both with Qp:Qs 1.8:1 but resting cardiac outputs of 4 L/min and 6 L/min have pulmonary flows of 7.2 L/min vs 10.8 L/min — a 50% difference in volume overload burden (sources/asd-japc-2025 — medium)
  • A tall, athletic individual with hyperdynamic circulation may carry a substantially higher right heart burden at the same Qp:Qs as a sedentary patient
  • Qp:Qs assumes stable vascular resistances; does not account for physiological fluctuations during exercise or neurohumoral modulation (sources/asd-japc-2025 — medium)

Management

Indications for Closure

• General principle: Timely closure in presence of RV volume overload irrespective of age (excluding advanced pulmonary vascular disease or significant LV impairment); multidisciplinary CHD team decision (sources/asd-ehj-2022 — high)
• Optimal age: Before 25 yr → normal survival; benefits observed at all ages
• PVR thresholds (invasive haemodynamic assessment mandatory when PH present):
  • PVR <5 WU: closure safe and beneficial
  • PVR ≥5 WU: closure likely harmful → initiate PAH pharmacotherapy first; reassess
  • PVR ≥5 WU + Qp/Qs >1.5 on PAH therapy: fenestrated closure may be considered (long-term evidence lacking)
  • PVR ≥5 WU despite PAH treatment: closure contraindicated
• Vasoreactivity testing with inhaled NO: NOT recommended in ASD-PAH decision-making (sources/asd-ehj-2022 — high)
• LV impairment: balloon test occlusion pre-device deployment; pre-conditioning with HF medical therapy (sources/asd-ehj-2022 — high)

Percutaneous (Catheter) Closure

• Treatment of choice for eligible secundum ASDs (>80% of all ASDs); low complication rate vs surgery; no sternotomy; shorter stay (sources/asd-ehj-2022 — high)
• Anatomic contraindications: insufficient/absent rims (except isolated absent aortic rim which can often be accommodated), very large defect (>36–40 mm), excessively floppy aneurysm; interference with AV valves or venous drainage (sources/asd-lancet-2014 — high)
• Devices: self-centring Amplatzer-type shape memory alloy occluders; ASO (St. Jude Medical) can close defects up to 38 mm; biodegradable devices in development (sources/asd-closure-jc-2015 — high)
• Device complication rates (meta-analysis, 28,142 patients, 203 studies):
  • Major periprocedural: 1.6% (95% CI 1.4–1.8); device embolisation 0.7%; pericardial tamponade 0.1%
  • Late complications: arrhythmias 1.5%, stroke 0.4%, device thrombosis 0.2%, cardiac erosion 0.1%, embolisation 0.1%, death 0.1% (sources/asd-lancet-2014 — high)
• Nickel allergy: new-onset/worsening migraine reported; clopidogrel may help (sources/asd-lancet-2014 — high)
• Antiplatelet drugs routinely prescribed post-device but supporting data lacking (sources/asd-lancet-2014 — high)
• Emerging: covered stent for superior sinus venosus ASD (catheter); still requires surgical backup (sources/asd-ehj-2022 — high)

Cardiac Erosion — Mechanism and Risk Factors

• Rare (<0.1%) but serious complication; complete mechanism not established (sources/asd-lancet-2014 — high; sources/asd-closure-jc-2015 — high)
• Established risk factors: aortic (superoanterior) rim deficiency + oversized device selection (sources/asd-closure-jc-2015 — high)
• Atrial septal malalignment (novel risk factor): Surfaces of septum primum and secundum are vertically offset, causing tight impingement of the right atrial disk toward the aortic root; changes the device axis angle after cable release → increased erosion risk. Many patients with aortic rim deficiency do NOT erode — malalignment likely the additional required factor (sources/asd-closure-jc-2015 — high)
• Updated ASO Instructions for Use (IFU): IVC rim <5 mm added as contraindication (in addition to existing aortic rim deficiency criterion), because this anatomy tends to cause oversized device selection (sources/asd-closure-jc-2015 — high)
• Pre-deployment assessment of malalignment is difficult; both disk surfaces must be observed pre- and post-cable release
• Longitudinal follow-up is essential — late erosion has been documented

Technical Techniques for Difficult ASD

• Large ASD (>30 mm) and wide rim deficiency require modified techniques; operators should master 1–2 alternatives (sources/asd-closure-jc-2015 — high)
• LUPV/RUPV (pulmonary vein) technique: Left atrial disk deployed inside left or right upper pulmonary vein, then retracted to optimal septal position; enables large devices up to 38 mm; pulmonary vein injury is a concern
• Hausdorf sheath (Cook, Bloomington, IN): Designed to align left atrial disk parallel to atrial septum; improves approach in large defects
• Balloon assist technique: Partially inflated Amplatzer sizing balloon holds left atrial disk inside left atrium, preventing prolapse into right atrium; relatively simple, requires additional venous access; high success rates even for large ASD
• Sheath rotation/remolding: Rotation or remolding outside the body to improve alignment at the interatrial septum
• Straight side-hole delivery sheath (Kutty): Enhanced delivery in difficult anatomy

Hemodynamic Management in Elderly Patients

• ~10% of transcatheter ASD cases may be aged >70 years; majority have ≥1 major comorbidity; most are NYHA class >II (sources/asd-closure-jc-2015 — high)
• LV diastolic dysfunction is masked by the L→R shunt (unloads LV); closure causes abrupt LV preload rise → risk of acute CHF, especially in elderly with impaired diastolic function (sources/asd-closure-jc-2015 — high)
• Pre- and peri-procedural anti-congestive therapy (diuretics, optimised fluid balance): essential to prevent post-closure CHF in elderly; series shows no acute CHF despite impaired diastolic function when pre-treated (sources/asd-closure-jc-2015 — high)
• PCWP monitoring protocol (author's institution): (sources/asd-closure-jc-2015 — high)
  • PCWP continuously monitored during device deployment
  • Abandon procedure if mean PCWP rises >10 mmHg from baseline during test balloon occlusion
  • Abandon procedure if mean PCWP >20 mmHg absolute (pulmonary oedema risk)
• Fenestrated device: May moderate abrupt hemodynamic change post-closure; optimal fenestration size not yet evaluated; limited evidence (sources/asd-closure-jc-2015 — high)
• Balloon test occlusion pre-device deployment is recommended when LV dysfunction is suspected (sources/asd-ehj-2022 — high; sources/asd-closure-jc-2015 — high)

Surgical Closure

• Indicated for sinus venosus, primum, coronary sinus defects, and secundum not amenable to device closure
• Mortality <1%; major complications <7%; most common complication: postoperative arrhythmia
• Minimally invasive approaches (mini-sternotomy, thoracotomy, robotics) comparable outcomes to sternotomy (sources/asd-ehj-2022 — high)

Arrhythmia Management

• Atrial flutter and AF prevalence increases steeply with age: >20% atrial flutter, >50% AF by age 60 (sources/asd-ehj-2022 — high)
• AF prevalence by decade: ~1/3 of patients aged >60 years; ~1/2 of patients aged >70 years have AF (sources/asd-closure-jc-2015 — high)
• Substrate: right atrial fibrosis, reduced voltage, prolonged refractory periods, Bachmann's bundle conduction disturbances (sources/asd-ehj-2022 — high)
• ASD closure does NOT reliably reduce AF burden (meta-analysis of 25 studies: percutaneous closure not associated with arrhythmia reduction) (sources/asd-ehj-2022 — high)
• Closure before age 40 may offer some AF protection; persistent AF unlikely to benefit; paroxysmal AF may improve at lower ages (sources/asd-ehj-2022 — high)
• Key management principle — PVI before ASD closure when symptomatic AF present (freedom from post-closure AF 79% vs 37%); closing ASD first increases complexity of trans-septal access (sources/asd-ehj-2022 — high)
• PVI-first criteria (Okayama University protocol): age <75 years + symptomatic paroxysmal or persistent AF + LA dimension <50 mm; ASD closure scheduled when sinus rhythm persists >3 months after ablation (sources/asd-closure-jc-2015 — high)
• Permanent AF: NOT a contraindication to ASD closure; closure resolves L→R shunt and improves symptoms; strict anticoagulation required indefinitely post-closure (sources/asd-closure-jc-2015 — high)
• Sinus node dysfunction: more prevalent in superior sinus venosus defects (sources/asd-ehj-2022 — high)
• AV block: more frequent in primum ASDs (inferiorly displaced AV conduction system) (sources/asd-ehj-2022 — high)

PAH Management

• Oral ERA ± PDE5 inhibitor when PVR ≥5 WU (sources/asd-ehj-2022 — high)
• Eisenmenger syndrome: proactive PAH therapy with prostacyclins; subcutaneous/inhaled preferred over parenteral (sources/asd-ehj-2022 — high)
• Closure contraindicated in established Eisenmenger syndrome (sources/asd-ehj-2022 — high)

Exercise Physiology and Shunt Dynamics

• Left atrial pressure rises more than right atrial pressure during exercise; the left atrium is described as stiffer in ASD patients (sources/asd-japc-2025 — medium)
• Exercise-induced changes in right and left atrial compliance, PVR-compliance, and systemic arterial pressure could dynamically alter trans-atrial pressure gradients and shunt volume
• The assumption that Qp:Qs remains stable at rest and during exercise is largely unvalidated (sources/asd-japc-2025 — medium)
• Three published studies with contradictory findings on exercise shunt response:
  • Bay et al. 1971 (stress testing): increased absolute shunt volume
  • Nielsen & Fabricius 1968 (exercise): decreased shunt volume
  • Stephensen et al. 2017 (dobutamine + atropine stress, CMR): no change in shunt volume
  • Caveats: very small sample sizes; some used pharmacological rather than true exercise stressors; individual responses within each study contradicted average trends (sources/asd-japc-2025 — medium)
• Key conceptual point: A declining Qp:Qs during exercise does NOT necessarily indicate reduced absolute pulmonary flow — if systemic cardiac output rises faster than the ratio falls, net pulmonary flow may still increase
• Endurance exercise in unrepaired ASD may exacerbate right heart dilation and trigger maladaptive pulmonary vascular remodeling in susceptible individuals; current guidelines provide no clear directive on exercise restriction (sources/asd-japc-2025 — medium)
• Advanced characterisation tools: real-time 4D flow MRI; invasive exercise hemodynamics — not in routine clinical use

Post-Closure Outcomes

Stroke

• Open ASD: ~4% stroke prevalence (paradoxical embolism primary mechanism) (sources/asd-ehj-2022 — high)
• Closed ASD: 1.4% stroke prevalence (residual risk from AF and pulmonary venous remodelling) (sources/asd-ehj-2022 — high)
• Consider anticoagulation 6–12 months post-closure in older patients irrespective of pre-procedural AF (sources/asd-ehj-2022 — high)

Predictors of Late-Onset PAH After Closure

• The belief that ASD closure reduces pulmonary pressures in all patients is overly simplistic (sources/asd-japc-2025 — medium)
• A subset of patients develops PAH before or even after closure; late-onset PAH can occur years to decades post-intervention — an increasingly recognised phenomenon (sources/asd-japc-2025 — medium)
• Risk factors for post-closure PAH:
  • Delayed closure (typically >40 yr) — irreversible pulmonary vascular changes may have occurred before repair
  • Pre-closure elevated PVR even if below Eisenmenger threshold; pre-closure hemodynamic assessment including vasoreactivity testing and PVR quantification is essential in older adults and symptomatic patients
  • BMPR2 genetic variants — limited data in ASD-specific cohorts; data extrapolated from IPAH/HPAH literature (sources/asd-japc-2025 — medium)
  • Unrecognised pulmonary microvascular disease, especially in women or patients with autoimmune conditions
• Some patients develop PAH despite early closure — suggesting nonhemodynamic mechanisms: endothelial dysfunction, inflammation, in situ thrombosis (sources/asd-japc-2025 — medium)
• An abnormal rise in pulmonary arterial pressure in response to augmented blood flow (exercise or medication) — before or after closure — may help identify patients at higher PAH risk (sources/asd-japc-2025 — medium)

Pregnancy

• Well tolerated in uncomplicated ASD; advise closure before conception (sources/asd-ehj-2022 — high)
• ASD discovered during pregnancy: continue, close electively post-partum (sources/asd-ehj-2022 — high)
• Repaired ASD: pregnancy outcomes equivalent to general population (sources/asd-ehj-2022 — high)
• ASD + PAH: pregnancy risks prohibitive; effective contraception essential (sources/asd-ehj-2022 — high)

Contradictions / Open Questions

• PVR threshold for ASD closure — 5 WU vs 8 WU: Brida et al. EHJ 2022 cites PVR ≥5 WU as contraindication to closure (per ESC 2020 guidelines); Geva et al. Lancet 2014 cites PVR >8 WU as generally precluding closure (per ACC/AHA 2008 and ESC 2010). The 5 WU threshold represents the more recent and current standard. Clinically significant: a patient with PVR 6–7 WU would be refused closure under ESC 2020 but potentially acceptable under older guidelines. (sources/asd-ehj-2022 — high; sources/asd-lancet-2014 — high)
• ASD closure and AF — short-term benefit but long-term attenuation: Geva 2014 Lancet meta-analysis (26 studies, 1841 surgical + 945 transcatheter) shows OR 0.66 reduction in arrhythmias short-term, but beneficial effect lost at ≥5 years follow-up. Brida 2022 EHJ meta-analysis (25 studies) shows percutaneous closure NOT associated with arrhythmia reduction overall. Both support that ASD closure is not an indication for closure based on AF alone. (sources/asd-ehj-2022 — high; sources/asd-lancet-2014 — high)
• Treat-and-repair strategy in ASD-PAH: Long-term benefit of PAH pharmacotherapy followed by closure (treat-and-repair) is unproven. Evidence for benefit is limited and quality is low. (sources/asd-ehj-2022 — high)
• Prophylactic catheter ablation before ASD closure in patients without documented AF: Whether to ablate prophylactically, and whether targeting Bachmann's bundle improves arrhythmia-free survival, remains unresolved. (sources/asd-ehj-2022 — high)
• Fenestrated closure in borderline PVR (≥5 WU + Qp/Qs >1.5): Long-term impact of fenestrated vs complete closure in this subgroup is unknown. (sources/asd-ehj-2022 — high)
• Sinus venosus defect catheter closure: Covered stent approach is promising but data are very limited (small series); remains investigational with surgical closure as standard of care. (sources/asd-ehj-2022 — high)
• No RCT comparing transcatheter vs surgical closure: As of Lancet 2014 and still relevant — comparative effectiveness relies on observational and registry data only. (sources/asd-lancet-2014 — high)
• PVR threshold units — indexed vs absolute: Akagi 2015 JC cites hemodynamic criterion as PVR <5 WU/m² (indexed to BSA); EHJ 2022 and ESC 2020 guidelines use absolute PVR <5 WU. For an average adult (BSA ~1.7–1.9 m²), indexed PVR <5 WU/m² corresponds to absolute PVR ~3.0–3.5 WU — a meaningfully stricter criterion. A patient with absolute PVR 4 WU would be eligible under ESC guidelines but potentially borderline under the indexed criterion. Clinical practice should clarify which metric is being used. (sources/asd-closure-jc-2015 — high; sources/asd-ehj-2022 — high)
• Cardiac erosion mechanism — incomplete: The definite mechanism is not established. Aortic rim deficiency + oversized device are accepted risk factors, but most cases with these features do NOT erode. Atrial septal malalignment has been proposed as an additional required factor but evidence is mechanistic/case-series level, not prospective. (sources/asd-closure-jc-2015 — high)
• Qp:Qs ratio — relative metric masking absolute flow burden: Qp:Qs does not capture absolute pulmonary blood volume. Two patients with identical Qp:Qs 1.8:1 can differ by >3 L/min in absolute pulmonary flow. Clinical decision thresholds (e.g., Qp:Qs ≥1.5) may underappreciate hemodynamic burden in individuals with high cardiac output (athletes, hyperdynamic states). No validated absolute flow threshold has been proposed. (sources/asd-japc-2025 — medium)
• Exercise-induced shunt dynamics — unresolved and contradictory: Three published studies show opposing directions of change in absolute shunt volume during exercise (Bay 1971: increased; Nielsen 1968: decreased; Stephensen 2017: no change). None used invasive exercise hemodynamics or 4D flow MRI with adequate sample sizes. True impact of exercise on ASD physiology across the age spectrum remains unclear. (sources/asd-japc-2025 — medium)
• Exercise restriction in unrepaired ASD — no guideline guidance: Whether patients with unrepaired ASD — particularly those engaged in endurance sports — should limit physical activity to prevent right heart volume overload and pulmonary vascular remodeling is unanswered. Current guidelines contain no specific directive. (sources/asd-japc-2025 — medium)
• Post-closure PAH risk stratification — no validated model: Late-onset PAH after ASD closure is recognised but not reliably predictable. Candidate risk factors (delayed closure >40 yr, pre-closure elevated PVR, BMPR2 variants, pulmonary microvascular disease) are hypothesis-generating only; no prospective validated risk stratification tool exists. (sources/asd-japc-2025 — medium)
• BMPR2 variants in ASD-associated PAH — limited data: BMPR2 is established in IPAH/HPAH, but evidence for its role in ASD-associated PAH specifically is described as "limited" with no quantification. Genetic testing is not part of current ASD-PAH evaluation protocols. (sources/asd-japc-2025 — medium)

Connections

• Related to concepts/ACHD-AP-Classification — secundum ASD classified as Anatomic Class I (Simple)
• Related to concepts/Pulmonary-Hypertension — ASD-PAH is Group 1 pre-capillary PAH; PVR thresholds guide closure decisions
• Related to concepts/PAH-Risk-Stratification — risk stratification determines pharmacotherapy intensity
• Related to entities/Atrial-Fibrillation — AF most common arrhythmia; pre-closure ablation strategy critical
• Related to concepts/SVT-Management — AF management principles apply
• Related to concepts/Intracardiac-Shunts — Qp:Qs measurement methods and limitations

Sources

• sources/asd-ehj-2022 — primary source (2022 EHJ state-of-the-art review)
• sources/asd-lancet-2014 — foundational Lancet Seminar (natural history, spontaneous closure, surgical survival, device complication data)
• sources/asd-closure-jc-2015 — technical review (cardiac erosion mechanism, difficult ASD closure techniques, elderly PCWP protocol, AF management strategy)
• sources/shunt-nature-sr-2020 — paediatric Qp/Qs estimation; COstatus ultrasound dilution method-comparison study
• sources/asd-japc-2025 — brief communication; Qp:Qs limitations (absolute vs relative flow); exercise-induced shunt dynamics; post-closure PAH risk factors