Desaturation Ten Days After a Norwood With a 3.5 mm Shunt

The number: 4 x (3.6)^2 = 52 mmHg peak gradient across the shunt.

It does not prove stenosis. The gradient across a Blalock-Taussig shunt equals neo-aortic pressure minus pulmonary artery pressure; a high gradient is equally consistent with a low PA pressure (a good thing) as with a narrowed anastomosis. What matters is where the gradient sits and what the flow looks like:

• Look at the 2D/colour anastomosis, both ends: a discrete colour alias and a visible narrowing at the shunt-to-PA or shunt-to-innominate junction localises a stenosis; a uniformly narrow shunt with a high gradient throughout indicates diffuse hypoplasia.
• Flow timing: normal shunts flow continuously, with systolic predominance. Loss of diastolic flow, or flow that becomes systolic-only, suggests raised downstream resistance or a kink.
• Distal pulmonary arteries: small, poorly filled PAs with a high-velocity jet at the anastomosis support stenosis; generous PAs with a high gradient suggest the shunt is simply appropriately sized for a low PA pressure.
• The neo-aortic arch must be interrogated for recoarctation or arch hypoplasia, which raises the upstream pressure and mimics a shunt problem while actually being the cause of poor systemic output.
• Ventricular function and AV valve regurgitation determine whether the desaturation is a shunt problem or a failing circulation.

A 3.6 m/s jet with continuous flow and mildly reduced function is ambiguous; the anatomy and the diastolic flow pattern decide.

Patient safety takes precedence over completeness of the study, and the modification is planned before the probe touches the chest.

• Prioritise. Decide in advance the four questions that change management: is the shunt patent, is there a stenosis, is the arch obstructed, is ventricular function adequate. Everything else (full segmental sweep, pulmonary veins, detailed valve interrogation) is deferred.
• Minimise handling. Keep the infant settled in a neutral thermal environment; use warmed gel; avoid repositioning more than necessary by using subcostal and apical windows sequentially from one side. Do not press on the shunt or the sternotomy.
• No sedation, no feeding challenge. Sedation and oral feeds both risk decompensation in a shunt-dependent circulation; if sedation is genuinely needed this becomes a planned, monitored procedure.
• Watch the lines. The umbilical arterial line and central line are the infant's lifeline; keep the probe and cable away from them and never lift the infant to get a window. Monitor the arterial trace and saturation on the screen during the study.
• Time-box the study and stop for bradycardia or a desaturation that does not recover with a brief pause; communicate with the bedside nurse throughout so that handling is coordinated with suctioning and blood gas sampling rather than stacked on top of it.
• Document the limitation in the report so the team knows which questions remain unanswered rather than assuming a negative study.

Why Qp/Qs fails: in a shunt-dependent single-ventricle circulation, pulmonary blood flow comes from the shunt, not from a ventricular outflow tract, so the standard RVOT and LVOT stroke-volume calculation has no anatomical basis. The systemic and pulmonary flows are both supplied by the same ventricle, and the 'shunt flow' cannot be derived from a diameter and a VTI with any reliability. Quoting a ratio here would be false precision.

What the report should contain instead:

• Shunt patency and calibre, with the site of any narrowing described anatomically and the peak and mean gradient stated alongside the diastolic flow pattern.
• Neo-aortic arch: dimensions, presence of recoarctation, and descending aortic flow.
• Ventricular systolic function (qualitative, with fractional area change if measurable) and tricuspid regurgitation severity.
• Atrial septum: patency of the septostomy or fenestration, since a restrictive atrial communication in this circulation causes desaturation and low output.
• Interpretation tying the findings to the clinical picture - saturations of 68% with a lactate of 5.2 and falling urine output is a low-output state, not simply a shunt problem, and the team needs to know whether the imaging supports shunt stenosis, arch obstruction, or ventricular dysfunction.
• A clear recommendation: if a discrete shunt stenosis or arch obstruction is demonstrated, catheterisation or surgical revision is indicated; if the shunt is patent and unobstructed with reduced function, escalating inotropic and ventilatory support rather than re-imaging is the priority.