Shunt physiology in neonatal hemodynamics

A practical guide to blood flow and targeted neonatal echocardiography

A neonatal shunt changes where blood travels, which chambers carry the extra volume, and what an echocardiographic flow measurement represents. A high ventricular output can coexist with inadequate systemic perfusion. Interpreting the number requires identifying every important shunt and following blood through the circulation. This section explains isolated and combined atrial, ventricular, and ductal shunts; left-to-right, right-to-left, and bidirectional flow; and the contribution of targeted neonatal echocardiography (TNE). The emphasis is on physiology and bedside interpretation, rather than treatment thresholds.

Start with three separate questions

• Where is the blood going? Describe the anatomical route and the direction of flow throughout the cardiac cycle.

• How much blood is circulating? Distinguish total pulmonary flow, total systemic flow, and recirculated shunt flow.

• What is the consequence? Assess ventricular loading and performance, pulmonary congestion, systemic perfusion, and oxygen delivery.

Reading the ratio: Qp/Qs above 1 indicates net pulmonary recirculation (excess pulmonary blood flow, decreased systemic flow; or combination of both depending on capacity to compensate); below 1 indicates net pulmonary bypass. A value near 1 describes a balance of total flows, but does not exclude opposing shunts, hypoxemia, or low absolute flow. The flow identities that follow assume a steady state, normal venous connections and great-artery relationships, and no important valvar regurgitation, other shunts, or collateral supply. They are conservation-of-flow teaching relationships; real Doppler measurements are estimates. [6]

Before and after the tricuspid valve

The pre-tricuspid versus post-tricuspid distinction helps explain the usual loading pattern. It does not replace the specific anatomy: a VSD and a PDA are both post-tricuspid, yet they affect aortic and pulmonary valve flows differently.

Partial anomalous pulmonary venous return also creates pre-tricuspid volume loading, but is not simply flow across an atrial hole. It changes venous routing; the normal-drainage equations in this section must be modified. The existing ASD and VSD sections provide the detailed anatomical context. [17,18]

Ventricular compliance and atrial shunting

Atrial flow follows the instantaneous LA-to-RA pressure difference. The filling properties of the ventricles help generate that difference, particularly when the atrioventricular valves are open. A large ASD can permit substantial exchange despite very small pressure differences. [20]

Compliance C = (ΔV) / (ΔP)

Compliance describes how much volume is accommodated for a given pressure change. Stiffness is the reciprocal relationship, ΔP/ΔV. The ventricular diastolic pressure–volume relationship is nonlinear, so local compliance changes with the operating volume. 

Relative ventricular filling properties matter

A relatively compliant RV can accept more volume at a low filling pressure, favoring LA-to-RA flow. Reduced LV compliance or increased LV filling load can raise LA pressure and augment this tendency. Reduced RV compliance, impaired RV filling, or increased right-sided filling pressure can reduce or reverse it. Valve disease, atrial restriction, and venous streaming also matter. [17,20] During neonatal transition, changes in RV afterload and remodeling alter right-heart filling. However, PVR and ventricular compliance are different quantities: a fall in PVR can influence RV loading and adaptation, but does not directly specify the size of the atrial shunt. A larger left-to-right atrial signal does not necessarily mean a larger anatomical defect. [5,17]

Compliance is not relaxation

Relaxation is the active process by which myocardial tension falls after contraction. Compliance describes the passive pressure response to filling. A ventricle may have impaired relaxation, reduced compliance, or both. Increased LA pressure may accelerate transmitral filling despite abnormal ventricular properties.

Do not confine atrial flow to late diastole

Doppler studies with simultaneous atrial pressures describe atrial left-to-right flow during ventricular systole and diastole, often peaking in late systole or early diastole and increasing again with atrial contraction. Transient reversal can occur. These observations are not a universal neonatal waveform template: use the infant’s actual tracing, ECG timing, respiration, and hemodynamic context. [19] The pressure–volume figure compares hypothetical ventricles. Ventricular filling pressure is not identical to atrial pressure throughout the cycle. 

Follow the flow through each shunt

In a circulation without a shunt, the same net volume crosses the pulmonary vascular bed, pulmonary veins, mitral valve, aortic valve, systemic vascular bed, and pulmonary valve. A shunt introduces a branch in this pathway.

Atrial shunt

With net LA-to-RA flow, some pulmonary venous blood returns to the right heart before crossing the mitral valve. The RV handles the recirculated volume. The mitral and aortic flows remain equal if the ventricular septum is intact and there is no important regurgitation. Thus, an atrial shunt separates pulmonary venous return from transmitral flow; it does not itself separate transmitral from transaortic flow.

Ventricular shunt

With net LV-to-RV flow, blood has already crossed the mitral valve before leaving the LV through the VSD. The LV ejects into two pathways: the aorta and the VSD. LVO measured at the aortic valve excludes the VSD flow. Increased LA/LV filling is therefore compatible with an aortic LVO that represents Qs. The RV outflow carries the pulmonary flow, although the chamber-loading pattern differs from an atrial shunt.

Ductal shunt

With net aorta-to-pulmonary artery flow, recirculating blood has crossed both the mitral and aortic valves before returning to the lungs. Aortic LVO includes systemic delivery plus ductal runoff. With an isolated PDA, LVO represents Qp and RVO represents Qs. These relationships depend on excluding additional important shunts. For isolated net right-to-left flow, the corresponding inequality reverses: an atrial shunt makes QMV exceed QPV; a VSD makes aortic LVO exceed QMV; a PDA makes Qs exceed LVO and RVO exceed Qp. The signed-flow equations apply to all three.

Choose the correct ratio

The correct numerator is the measured flow that represents Qp in that anatomy. “Pulmonary valve flow” and “pulmonary blood flow” are not interchangeable when a PDA is present.

The identities also hold for net right-to-left or bidirectional flow under the stated assumptions. With an isolated PDA and net right-to-left flow, LVO/RVO is below 1. With an isolated intracardiac net right-to-left shunt, RVO/LVO is below 1. In practice, additional communications, unstable physiology, regurgitation, and measurement error often prevent a reliable numerical ratio.

Do not substitute velocity for volume flow

Qp/Qs is a ratio of volume flow rates. It is not a ratio of peak velocities. At each outflow, volume per beat depends on cross-sectional area multiplied by velocity–time integral (VTI); flow per minute also includes heart rate. The squared diameter term matters. A ratio of RVOT VTI to LVOT VTI is not Qp/Qs unless the necessary area relationship has also been accounted for. [6]

Flow (mL/kg/min) = (π × d² × VTI × HR) / (4 × weight)

Use diameter d and VTI in cm, heart rate (HR) in beats/min, and weight in kg. This standard estimate assumes a suitable circular area and matched sampling site; ordinary forward systolic VTI does not correct for important regurgitant flow. [6]

Why a correct equation can still produce a poor estimate

Aortic and pulmonary measurements each have acquisition error. Their ratio compounds those errors, and non-simultaneous acquisitions may describe different physiological states (loading conditions, vascular resistances and heart rate). Use matched measurement sites and a consistent protocol; retain the underlying diameters, VTIs, heart rates, and image-quality assessment. A technically calculated ratio should only be labelled Qp/Qs when the anatomy and acquisition support that interpretation. [1,6] SVC flow is not Qs. It measures venous return from the upper body, including the brain, rather than the whole body. It may add useful information when ventricular outputs contain recirculated blood, but cannot replace Qs in a pulmonary-to-systemic flow ratio. [3] SVC flow estimates are highly variable due to concerns with the technique to estimate this value by echocardiography. Also, some infants may have bilateral SVC.

At the PDA, right-to-left flow indicates pulmonary arterial pressure exceeds local aortic pressure during that interval. It does not, by itself, give an absolute pulmonary vascular resistance (PVR). Low systemic pressure can contribute. A brief early systolic reversal does not necessarily mean the peak pulmonary pressure exceeds the peak systemic pressure. [14] At the atrial level, flow depends on changing RA/LA pressure differences, ventricular filling properties, venous streaming, and the size and geometry of the communication. Atrial direction is not a direct pulmonary artery pressure measurement. At a VSD, direction reflects the ventricular pressure relationship and must be interpreted with both outflow tracts. [4,5] A PDA can be right to left because the aortic filling pressures are very low (ex: interrupted aortic arch). 

Bidirectional does not mean balanced

A short right-to-left phase and a long left-to-right phase can produce net left-to-right flow, but duration alone cannot establish its magnitude. Volume depends on area and velocity integrated over time. Describe “right-to-left in early systole and left-to-right later in systole and diastole” when that is what the tracing shows, rather than reporting only “bidirectional.” Early transitional bidirectionality can occur as the circulation adapts after birth. Persistent or predominant right-to-left flow in an unwell infant warrants assessment of pulmonary vascular disease, ventricular dysfunction, and structural or duct-dependent heart disease. A purely right-to-left atrial pattern also requires careful evaluation of pulmonary venous connections and right-heart anatomy. [4,5]

PDA Doppler informs on instantaneous pressure gradient. The peak of this difference may not occur at the peak of pulmonary pressure; or the peak of aortic pressure; or both. As such, it provides some insight on relationship between Aortic and Pulmonary pressure curves, but does not inform on the absolute peak net difference between those 2 compartments (as the peaks may not occur at the same time, and the gradient at these peaks may not be the largest difference between these 2 curves when they are superimposed on each other). See Understanding PDA Spectral Doppler.

Oxygenation is a different measurement

Right-to-left flow may lower arterial oxygen content even while supporting blood flow. A right-to-left PDA can lower postductal saturation, whereas atrial right-to-left flow can lower both preductal and postductal saturations. Absence of a saturation difference therefore does not exclude important extrapulmonary shunting. Assess oxygen delivery using perfusion, hemoglobin, oxygenation, and ventricular performance together. [4]

Understand combined shunts

A PDA plus an atrial communication

This common neonatal combination creates two possible recirculation loops. With both shunts left to right, ductal blood returns through the pulmonary veins; some then crosses the mitral valve, while some crosses the atrial septum and re-enters the right heart. Both LVO and RVO can contain recirculated blood. [2,3] An atrial communication may decompress the LA. A modest LA size, LA/Ao ratio, or LV volume does not exclude substantial pulmonary flow when part of the pulmonary venous return is diverted into the RA. Look at the atrial shunt as an explanation for apparently discordant left-heart markers. [2,7]

Use pulmonary venous, mitral, and outflow signals to answer different questions. A rise in pulmonary venous diastolic velocity may support increased return, but evaluate the individual veins for focal acceleration, turbulence, or loss of normal phasicity before attributing it to shunt flow. [2,12]

Can transmitral volume flow be quantified

Yes, dedicated methods have been studied. In a study of 30 stable preterm infants, mitral-derived cardiac output agreed best with RVO and LVO measured at the sinotubular junction; the authors proposed it as an alternative output assessment. This does not establish routine subtraction of mitral and aortic Doppler flows as an accurate VSD-volume method in unstable neonates with multiple shunts. Geometry, sample location, time-varying area, and regurgitation must be addressed. [10] Likewise, total pulmonary venous return equals Qp in the model, but summing reliable volume measurements from all pulmonary veins is not a routine bedside neonatal technique. A conservation equation tells us what a flow means; it does not guarantee that TNE can measure it accurately. [6] SVC measurements can complement the examination, especially during transition, but should not be treated as a precise, shunt-independent measure of total systemic flow. Modified acquisition methods continue to be investigated. [3,11]

Separate pressure from flow and resistance

A large communication can transmit pressure while carrying substantial flow. A small restrictive communication can produce a fast jet with relatively little volume. Shunt size, velocity, and physiological effect must be interpreted together.

Velocity estimates a pressure difference

ΔP ≈ 4v²

With velocity v in m/s, the simplified Bernoulli estimate gives ΔP in mmHg. It is an instantaneous pressure difference across the jet, not a flow rate, an absolute pressure, or PVR. Good alignment and a complete envelope matter; proximal velocity and other assumptions may limit the approximation. A well-recorded tricuspid regurgitation jet estimates the RV-to-RA systolic gradient. Adding an appropriate RA pressure estimate yields an approximate RV systolic pressure. Equating that with pulmonary artery systolic pressure requires absence of important RV outflow or pulmonary valve obstruction. An absent or incomplete TR envelope cannot exclude pulmonary hypertension. [1,4] For a left-to-right VSD, subtracting its gradient from LV systolic pressure can approximate RV systolic pressure. However, like the PDA, a VSD doppler provides instantaneous gradient between the RV and LV pressure curves. This means that peak velocity gradient accross the VSD may not necessarily occur at peak PA or peak Aortic pressure (and peak PA and peak Ao pressure may not themselve occur exactly at the same time). Systemic blood pressure approximates LV systolic pressure only if there is no important LV outflow or aortic obstruction. For either VSD or PDA calculations, non-simultaneous peak pressures and peak jet gradients can introduce error. [14]

Why ductal gradients require particular care

The largest aorta-to-pulmonary pressure difference may occur at a different time from either vessel’s peak pressure. Subtracting the peak ductal Doppler gradient from a cuff systolic pressure can therefore misestimate peak pulmonary pressure. Report the direction and timing, compare with contemporaneous blood pressure, and integrate septal shape, TR, RV function, and the clinical setting. [14]

Pulmonary pressure can rise for more than one reason

Mean PAP = mean LA pressure + (PVR × Qp)

This relationship explains why pulmonary artery pressure may be high with elevated PVR, increased pulmonary flow, elevated downstream LA/pulmonary venous pressure, or a combination. Pressure transmission through a large shunt further complicates interpretation. Septal flattening supports an altered RV/LV pressure relationship but does not distinguish these mechanisms on its own. [13] TNE helps to characterize the phenotype; it does not routinely provide a precise invasive PVR value. PA acceleration time, septal configuration, shunt direction, ventricular performance, and venous/inflow findings are complementary clues. Do not label an infant as having high PVR solely because a PDA jet is slow, the septum is flat, or estimated pulmonary pressure is high. [4,12] A left-to-right duct also does not prove that pulmonary pressure is normal: pulmonary pressure may be elevated while remaining below aortic pressure. The actual systemic blood pressure is part of the interpretation.

Left heart volume loading and diastolic filling

In a PDA with net left-to-right flow, the LV must eject both the blood delivered to the systemic circulation and the blood returning through the duct to the lungs. This aortic flow relationship remains true with a concurrent atrial shunt, although LVO then no longer equals total Qp.

LVO = Qs + net left to right ductal flow

Extra preload can increase stroke volume, but that increase may chiefly support recirculation. If filling reserve is limited, LA and LV filling pressures may rise. A hyperdynamic LV or high LVO therefore does not establish adequate systemic perfusion or good diastolic function.

Interpret mitral E and A in context

The E wave reflects early diastolic inflow; the A wave reflects inflow during atrial contraction. Many very preterm infants show A-wave predominance, but E/A below 1 is not a universal neonatal normal value. The pattern evolves with gestational and postnatal age and is influenced by loading, heart rate, and respiratory conditions. [2,16] With a substantial PDA, an increased LA-to-LV early diastolic gradient may increase E velocity, produce E predominance, and shorten the isovolumic relaxation time (IVRT). IVRT is the interval between aortic valve closure and mitral valve opening. A shorter interval may reflect earlier opening driven by higher LA pressure, even when myocardial relaxation has not improved. Serial neonatal data support this loading effect. [8] This can create an apparently more mature or “pseudonormalized” filling pattern. E/A above 1 does not, by itself, prove elevated LA pressure, and it does not prove normal relaxation. If E and A fuse at a high heart rate, do not force a separate-wave ratio. [2,12]

Atrial decompression changes the appearance

When the LA can empty across a PFO or ASD, some of the returning volume bypasses the mitral valve. LA pressure, LA size, and mitral findings may be less conspicuous than expected for the total pulmonary flow. Conversely, a restrictive atrial communication or impaired LV filling can promote pulmonary venous congestion. [2,7]

Build a coherent assessment

Integrate LA and LV dimensions, mitral inflow, IVRT, tissue Doppler where interpretable, pulmonary venous patterns, mitral regurgitation, atrial restriction, and respiratory status. No isolated E/A, E/e′, or IVRT measurement should be converted into a precise neonatal LA pressure or LV end-diastolic pressure. [12] The conceptual Frank–Starling figure illustrates different degrees of preload reserve. Its curves are schematic rather than measured neonatal reference curves. The practical message is that more filling and more total output do not necessarily produce more systemic flow.

Diastolic runoff and systemic perfusion

When aortic pressure remains above pulmonary arterial pressure in diastole, blood continues to leave the aorta through the PDA after aortic valve closure. This runoff can reduce aortic diastolic pressure and forward diastolic flow to systemic territories. A wide pulse pressure and bounding pulses can be clues, but do not quantify the shunt. [7,15]

Look directly for the flow consequence

Assess descending thoracic and/or abdominal aortic Doppler and, when appropriate, celiac, superior mesenteric, renal, and cerebral arterial waveforms. Describe whether diastolic flow is forward, absent, or reversed, including whether reversal persists throughout diastole. Holodiastolic reversal supports significant runoff in the appropriate anatomy. [2,7] Avoid teaching a fixed sequence of “gut and kidneys first, brain later.” Regional effects depend on vascular resistance, autoregulation, blood pressure, shunt magnitude, and illness. Cerebral vessels are generally preductal but can still show altered diastolic flow. An abnormal arterial waveform does not establish tissue ischemia or quantify the absolute flow delivered to that organ.

Why the myocardium may be vulnerable

LV coronary perfusion occurs predominantly during diastole. A useful conceptual approximation for the LV subendocardial driving pressure is:

LV coronary perfusion gradient ≈ aortic DBP − LVEDP

DBP is aortic diastolic blood pressure; LVEDP is LV end-diastolic pressure. A lower DBP and a higher LVEDP narrow this gradient, while a volume-loaded LV may have greater oxygen demand. Coronary flow has been investigated in small neonatal PDA studies, supporting its relevance to the physiological assessment. [9,15] This is an approximation of a pressure gradient, not a measurement of coronary flow or a validated treatment threshold. It should not be generalized to all coronary territories; RV coronary perfusion also has an important systolic component. Routine TNE does not directly measure LVEDP, and cuff DBP is an imperfect substitute for central aortic diastolic pressure.

An illustrative pressure example

If aortic DBP is 18 mmHg and LVEDP is 8 mmHg, the estimated gradient is 10 mmHg. If LVEDP rises to 12 mmHg at the same DBP, the gradient falls to 6 mmHg. These are hypothetical invasive pressures, not echo-derived values, normal ranges, or thresholds for intervention. At the bedside, interpret runoff alongside blood pressure quality and trends, urine output, lactate trajectory, examination, hemoglobin, oxygenation, and regional oxygenation monitoring when available. A high LVO or acceptable mean blood pressure cannot independently establish adequate organ perfusion. [1,12]

Shunt modulation during neonatal care

Ventilation, oxygenation, acid–base status, and cardiovascular support can alter shunt flow without changing the size of the communication. Their effects depend on the starting physiology and must be reassessed rather than inferred from a treatment label.

The lung-volume/PVR relationship is often illustrated as U-shaped: underinflation and overdistension can both increase resistance. Therefore, “higher PEEP raises PVR” is incomplete. These are physiological tendencies, not instructions to adjust ventilator settings or administer vasodilators for a particular infant. [21,23] Pulmonary vasodilation requires particular caution when pulmonary venous hypertension, important LV dysfunction, or duct-dependent systemic flow is present. Increasing pulmonary delivery into a poorly accommodating left heart can worsen pulmonary edema. [22]

Limits of simple shunt resistance equations

For mean pressure and flow measurements in the same steady state, the vascular resistance definitions can be rearranged to give:

Qp/Qs = ((mean PAP − mean LAP) × SVR) / ((mean AoP − mean RAP) × PVR)

PAP is pulmonary artery pressure, LAP is LA pressure, AoP is aortic pressure, and RAP is RA pressure. PVR and SVR must use compatible units. This expression retains both the pulmonary and systemic driving pressure gradients. [13] Qp/Qs approximates SVR/PVR only when those driving gradients are approximately equal, as may occur in selected nonrestrictive common-pressure or parallel-circuit models. It is not a general rule for an atrial shunt, a restrictive VSD or PDA, unequal downstream pressures, or complex obstruction. For example, if the pulmonary driving gradient is half the systemic driving gradient and SVR/PVR is 3, then Qp/Qs is 1.5, not 3. The pressure terms cannot simply be discarded.

What the fourth power of radius does and does not mean

Q = (π × ΔP × r⁴) / (8 × η × L)

Poiseuille’s relationship describes steady laminar flow through an ideal rigid cylindrical tube. Q is flow, ΔP the pressure drop, r radius, η viscosity, and L length. It illustrates why caliber strongly affects resistance under those assumptions. A neonatal duct or septal defect can be short, changing in shape, compliant, pulsatile, and associated with complex or turbulent flow. Do not use the fourth-power relationship as a bedside formula for ductal shunt volume or predict that a small change in diameter causes a fixed clinical change in flow.

Reassess the entire circulation after a change

Record whether the intervention altered lung inflation, oxygenation, systemic pressure, ventricular performance, or more than one of these. Then repeat the relevant shunt and perfusion assessment. A smaller left-to-right Doppler signal can reflect reduced recirculation, increased PVR, reduced output, or increased restriction; those possibilities have different implications.

After PDA closure: anticipate a different circulation

Closing a substantial left-to-right PDA removes a recirculating pathway. Pulmonary venous return to the left heart falls, reducing LV preload, while the LV loses an additional ejection pathway into the pulmonary circulation and faces a greater effective arterial load. These changes can occur after surgical or transcatheter closure; their clinical expression varies. [24,26]

A lower LVO can accompany better systemic delivery

Before closure, LVO = Qs + ductal flow. After complete closure, LVO = Qs under the normal-connection assumptions, even if an atrial shunt remains. Comparing the two LVO values without accounting for the lost ductal component can confuse reduced recirculation with reduced systemic perfusion. For example, a hypothetical infant has LVO 300, Qs 180 and ductal flow 120 mL/kg/min before closure. After closure, an LVO of 220 represents a lower total LV throughput but a higher Qs. This is a conservation example, not a prediction of the response to closure.

Loading changes and myocardial dysfunction are related but distinct

Ejection fraction, shortening fraction and deformation measurements depend on loading conditions. A decline after closure should be interpreted with LV size, stroke volume/output, blood pressure and perfusion. Studies describe different responses: reduced preload and LVO with unchanged load-independent contractility in one cohort, and impaired systolic performance in vulnerable infants in another. [24,25] An acute fall in ejection fraction after transcatheter closure may be transient and need not coincide with clinical circulatory failure. Surgical and catheter procedures should not be assumed to have identical risks or mechanisms of instability. [26,27]

Use Laplace’s relationship as a teaching approximation

Wall stress σ ≈ (P × r) / (2 × h)

For an ideal thin-walled spherical chamber, P is transmural pressure, r is internal radius and h is wall thickness. At fixed geometry, higher pressure raises stress; at fixed pressure, a larger radius or thinner wall raises stress. The real neonatal LV is thick-walled and changes shape during contraction, so this is not a validated bedside calculation of LV afterload. Do not teach an inevitable sequence of “closure → larger LV → reduced contractility.” LV size commonly decreases as preload falls. If ejection becomes impaired, residual volume and dilation can add to wall stress, but the observed trajectory must be established by serial assessment. Increased afterload can reduce shortening without proving that intrinsic contractility has fallen.

TNE reassessment after PDA closure

Obtain a physiological baseline and reassess after closure according to the procedure, clinical course and local program. An infant who deteriorates needs timely evaluation for altered loading, ventricular dysfunction, obstruction, residual shunting and other causes of instability. A single low postoperative output value should not automatically be attributed to afterload mismatch. [1,24,26] Closure of a right-to-left or bidirectional duct requires a different framework: it may remove support for systemic or pulmonary flow, or a route of RV decompression. Establish structural anatomy and the pulmonary vascular/ventricular phenotype before extrapolating the left-to-right model. [1,4]

A practical bedside assessment

Use the same sequence for each study so that a change in one measurement is interpreted within the whole circulation.

1. Establish the anatomy: Confirm systemic and pulmonary venous connections, atrial and ventricular communications, great-artery relationships, valve competence, both outflow tracts, the aortic arch, and ductal anatomy. Unexpected findings or suspected structural disease require comprehensive pediatric cardiology assessment. [1]

2. Characterize every important shunt: For each communication, report location, size, restriction, direction, and timing within the cardiac cycle. At a PDA, distinguish brief systolic reversal from sustained right-to-left flow. Record the blood pressure, heart rate, respiratory support, oxygen exposure, and relevant vasoactive therapy at the examination.

3. Assess the lungs and the left heart: Look for converging evidence of pulmonary overcirculation and congestion: pulmonary arterial diastolic flow, pulmonary venous pattern, LA/LV dimensions, mitral filling, and mitral regurgitation. Interpret a modest LA size in light of atrial decompression. Volume loading and pressure loading are related but are not interchangeable.

4. Assess both ventricles and systemic delivery: Evaluate systolic and diastolic performance, septal configuration, and the RV pressure relationship. Interpret LVO and RVO using the full shunt map. Examine systemic diastolic flow and integrate the clinical perfusion assessment; use SVC flow only as an additional upper-body measure where local expertise supports it. Objective measures can include RV fractional area change (FAC), tricuspid annular plane systolic excursion (TAPSE), LV ejection fraction, systolic eccentricity index, and pulmonary artery acceleration time relative to RV ejection time (PAAT/RVET). Interpret each against appropriate age and loading conditions. Pulmonary venous atrial reversal can occur normally; it is not independently diagnostic of elevated filling pressure. [1]

5. State the mechanism and the uncertainty: Describe whether the dominant problem appears to be pulmonary recirculation, reduced pulmonary flow, ventricular failure, pulmonary venous congestion, systemic runoff, or mixed physiology. Record which measurements were limited and which conclusions remain uncertain. Reassess after clinically meaningful changes; repeat a ratio only if its assumptions still apply.

Example reporting language

“There is a PDA with predominantly left-to-right flow and a concurrent left-to-right atrial communication. Increased pulmonary arterial diastolic flow and pulmonary venous return support pulmonary overcirculation. The atrial shunt may reduce the degree of LA enlargement. LVO and RVO both contain shunt-related flow and cannot independently define Qp or Qs; no numerical Qp/Qs is reported. Abdominal aortic holodiastolic reversal supports systemic runoff. Ventricular performance and clinical perfusion findings should be integrated when determining the overall hemodynamic impact.”

Six domains for a left to right PDA

Use this compact acquisition and interpretation table alongside the full circulation assessment. A PDA study also needs the atrial and ventricular septa, because additional shunts change the interpretation of output and chamber-loading measurements.

The first TNE examination assessing a suspected PDA, including in extremely preterm infants, should be comprehensive enough to identify major structural disease and duct-dependent lesions. This is a standard for the examination’s scope; timing of screening is a separate clinical and local-program decision. [1]

Worked teaching examples

All flow values below are invented and expressed in mL/kg/min. The first four cases use the simplified steady-state model with competent valves and no additional communications.

Bidirectional flow with net left to right transfer

Suppose a hypothetical PDA carries 100 units/min left to right and 40 units/min right to left. Net ductal transfer is 60 units/min left to right. With no other shunt, Qp exceeds Qs by 60 units/min. The right-to-left component still matters for oxygenation and pressure interpretation; the net value does not describe all blood mixing.

Opposing shunts with a ratio of one

Assume Qs is 200, atrial flow is 60 right to left, and ductal flow is 60 left to right. Qp is also 200, but LVO is 260 and RVO is 140. The equal pulmonary and systemic totals do not exclude substantial exchange between the circuits or reduced systemic oxygen saturation. Neither outflow ratio is Qp/Qs.

A physiological change does not have only one meaning

If LVO falls after a left-to-right PDA becomes smaller, the fall could reflect less recirculation, reduced systemic flow, or both. Determine which by reassessing ventricular performance, atrial flow, systemic arterial Doppler, and clinical perfusion rather than interpreting the change in LVO alone.

Conservation equations for combined shunts

This optional section derives the tables. It describes true mean net volume flows, rather than a validated protocol for measuring every shunt with TNE. The model assumes normal biventricular connections, normal venous drainage, no important regurgitation or collateral flow, and no net accumulation of blood in a chamber.

Define A as net LA-to-RA shunt flow, V as net LV-to-RV shunt flow, and D as net aorta-to-pulmonary artery shunt flow. A positive value means left to right; a negative value means right to left. For bidirectional flow, use the signed net transfer over the observation interval.

QPV = Qp

QMV = Qp − A

LVO = QMV − V = Qp − A − V = Qs + D

RVO = Qs + A + V = Qp − D

Qp − Qs = A + V + D

LVO − RVO = D − A − V

The last identity explains why the difference between ventricular outputs does not necessarily equal ductal shunt volume. With a concurrent left-to-right atrial shunt, part of the ductal contribution to LVO is offset by an atrial contribution to RVO.

Why subtracting Doppler flows can be misleading

In the model, QPV − QMV equals A and QMV − LVO equals V. At the bedside, subtracting two large, imperfect estimates can yield a very inaccurate small difference. Atrial or VSD flow should not be assigned a precise volume merely because these equations can be rearranged. Significant mitral regurgitation recirculates blood through the mitral valve; aortic regurgitation separates forward systolic aortic flow from net aortic flow. These additional loops invalidate uncorrected comparisons. The flow identities must be expanded to include the regurgitant components.

Connect to catheter hemodynamics

Qp/Qs = (CaO₂ − CvO₂) / (CpvO₂ − CpaO₂)

In the Fick relationship, CaO₂ is systemic arterial oxygen content, CvO₂ representative mixed venous content, CpvO₂ pulmonary venous content, and CpaO₂ pulmonary arterial content. Oxygen consumption cancels in the ratio. Saturation-only substitutions require additional assumptions, including comparable hemoglobin and negligible dissolved oxygen. Streaming, differential ductal saturations, lung disease, and nonrepresentative samples can distort the calculation. [13] Qp/Qs is a net flow comparison. It does not separately quantify simultaneous left-to-right and right-to-left exchange or define effective oxygen-carrying flow. The catheter section develops these distinctions; the calculations quiz provides further pressure and flow examples. [13,14]