How the model works
Shunt network, Qp/Qs & oxygen mixing
ASD, PAPVR, VSD, PDA and unrepaired TOF use distinct flow-path assumptions.
Overview
Without an anatomic shunt, the ordinary steady native circulation shares one net forward output. Shunts explicitly separate pulmonary flow Qp from systemic flow Qs; device bypass pathways are also accounted for internally when MCS is active.
Shunt flow is signed: positive represents left-to-right recirculation and negative represents right-to-left flow. Pressure/resistance changes after medications or ventilation can alter direction and magnitude. The model is not an operability or closure assessment.
Modeled connections
The ASD example has reference Qp/Qs above 1.5 and an atrial communication that buffers ventricular filling variation. PAPVR routes a selected 40% of pulmonary venous flow to the right heart and cannot reverse on its own. PDA connects the aorta to PA downstream of the RV. Large VSD connects ventricular systolic flow and represents an unrestrictive defect with elevated pulmonary vascular loading.
Eisenmenger adds advanced pulmonary vascular disease to a reversible communication; oxygen does not erase the anatomy. If necessary, the selector adds a suitable communication rather than labeling isolated PAPVR as a reversing defect.
Unrepaired Tetralogy of Fallot
The TOF example combines RV outflow obstruction and an unrestrictive ventricular communication, with an overriding aortic connection represented by right-to-left mixing. Its selected obstructed baseline has reduced pulmonary flow, Qp/Qs below one and systemic desaturation. RV systolic pressure approaches LV/systemic pressure while distal PA pressure can remain low. Numerical endpoints and obstruction response are original teaching assumptions.
This differs from Eisenmenger physiology: the right-to-left flow is driven by RV outflow obstruction rather than an assumed advanced pulmonary vascular resistance. It also differs from the ordinary Large VSD example, which has a separate pulmonary-vascular loading assumption. Oxygen does not remove the anatomic bypass. Pulmonary and systemic flow and content mixing remain explicitly conserved in the bounded model.
Distal PVR is applied in the separate RV-outflow/pulmonary branch balance. The ordinary series-circulation pulmonary preload penalty is bypassed for TOF because right-to-left ventricular flow also supplies the LV; otherwise the same load would be counted twice. Increasing PVR lowers Qp/Qs and saturation in the selected unsupported example, without claiming a measured congenital pressure-flow response.
TOF already includes its ventricular communication. The selector exchanges it with the standalone Large VSD or Eisenmenger example rather than combining duplicate or incompatible ventricular phenotypes. Additional allowed defects and other diseases are bounded extrapolations, not new measured congenital anatomy.
The selected unrepaired ECG illustrates RV hypertrophy/right-axis deviation; post-repair RBBB is not automatically assigned. Its systolic murmur represents RV outflow turbulence, with a reduced pulmonic closure component. The large VSD is not given a separate loud high-gradient murmur. Repaired residual PR/RV dilation, pulmonary atresia, collaterals, surgical shunts and hypercyanotic-spell dynamics require different anatomy or a transient model and are not inferred from this selection.
Conservation and finite reserve
The solver balances the connected branches after rate, vascular loading, therapies and MCS. Qp is used for pulmonary resistance; Qs for systemic resistance. RV outflow is not automatically Qp because PDA joins downstream of the RV; LV aortic ejection is not automatically Qs because recirculation and support can contribute.
Finite pump reserve limits extreme recirculation, reducing useful systemic flow and pressure. Simultaneous PS retains an RV–PA gradient; external constraint remains despite the communication. The branch equations and reserve ceilings are bounded illustrative anatomy.
With no forward circulation, shunt flow is zero and Qp/Qs is withheld. Unsupported arrest uses the existing 10 mmHg illustrative flat pressure reference, not a prediction of post-arrest equilibration. If a support device is selected but produces no flow, its separate right/left filling-pressure reservoirs remain; the shunt solver must not create perfusion or silently equalize those reservoirs.
Saturation run and mixing
Oxygen-content conservation generates the synthetic saturation run and right-to-left admixture. Reversed PDA can lower lower-body saturation while preserving a right-hand sample. PAPVR/ASD/VSD step-ups reflect where oxygenated blood re-enters the right circulation.
Conventional thermodilution is withheld in shunt examples; Fick reports the synthetic systemic Qs where valid. During supported arrest, a steady shunt solution is withheld because transient accumulation, ductal runoff and valve leakage need a different model. P/F and regional MCS/shunt gas mixing are not invented when unresolved.
References & further reading