SUPPLEMENTALFILES
Early Shunt Failure: Prevalence, Risk Factors and Outcomes
An Analysis of the Society of Thoracic Surgeons Congenital Heart Surgery Database
1Nhue Do M.D., 2Kevin D. Hill M.D. M.S., 2Amelia S. Wallace M.S., 3Luca Vricella M.D., 3Duke Cameron M.D., 4Jim Quintessenza M.D., 5Neil Goldenberg M.D., PhD., 6Constantine Mavroudis M.D., 5Tom Karl M.D., 7Sara K. Pasquali M.D., M.S., 5Jeffrey P. Jacobs M.D., 3Marshall L. Jacobs M.D.
1Children’s Hospital of Philadelphia, Philadelphia, PA
2Duke Clinical Research Institute, Durham, NC
3Johns Hopkins Medical Institutions, Baltimore, MD
4CincinnatiChildren’s Hospital, Cincinnati, OH
5Johns Hopkins All Children’s Hospital, St. Petersburg, FL
6Florida Hospital, Orlando, FL
7CS Mott Children's Hospital, Ann Arbor, MI
Correspondence:
Marshall L. Jacobs, Division of Cardiac Surgery, The Johns Hopkins Hospital, Zayed 7107, 1800 Orleans Street, Baltimore, MD 21287, USA. Email:
Supplemental Table 1.Index Operation included one or more of the following procedure codes (from Versions 3.0 and/or 3.22) of the STS CHS database:
Eligible
Shunt, Systemic to pulmonary, Modified Blalock-Taussig Shunt (mBTS)
Shunt, Systemic to pulmonary, Central (shunt from aorta)
Shunt, Systemic to pulmonary, Central (shunt from aorta) Central shunt with an end-to-side
connection between the transected main pulmonary artery and the side of the ascending
aorta (i.e. Mee shunt)
Shunt, Systemic to pulmonary, Other
Norwood Stage 1 Procedure with source of pulmonary blood flow specified as any one of the
above, OR as “Sano modification” OR as ventricle-to-pulmonary artery (Ventricle-to-PA)
conduit.
Damus-Kaye-Stansel Procedure (DKS) with source of pulmonary blood flow specified as any
one of the above, OR as “Sano modification” OR as ventricle-to-pulmonary artery
(Ventricle-to-PA) conduit.
Total anomalous pulmonary venous connection (TAPVC) repair + Shunt, systemic-to-
pulmonary
Ineligible
Hybrid approach Stage 1
Hybrid approach Stage 2: Aortopulmonary amalgamation + Superior cavopulmonary
anastomosis + PA Debanding + Aortic arch repair (Norwood [Stage I] + Superior
Cavopulmonary anastomosis + PA Debanding)
Combined Norwood/Rastelli Procedure: Conduit insertion right ventricle to pulmonary artery
+ Intraventricular tunnel left ventricle to neoaorta + Arch reconstruction (Rastelli and
Norwood type arch reconstruction) (Yasui)
Supplemental Table 2.
Patients were considered to have experienced Early Shunt Failure if any one or more of the following codes were entered during the same hospitalization as the Index Operation
- DIAGNOSIS: “Shunt Failure” Coded in association with a subsequent event (after the Index Operation) during the same hospital admission
- PROCEDURE: “Shunt, Reoperation” Coded in association with a subsequent event (after the Index Operation) during the same hospital admission
- Additional PROCEDURE criterion: Any one or more of the following procedures performed during the same hospitalization as the eligible Index Operation, but as all or part of a subsequent event (not part of Index Operation).
- Shunt, Systemic to pulmonary, Modified Blalock-Taussig Shunt (MBTS)
- Shunt, Systemic to pulmonary, Central (shunt from aorta)
- Shunt, Systemic to pulmonary, Central (shunt from aorta) Central shunt with an end-to-side connection between the transected main pulmonary artery and the side of the ascending aorta (i.e. Mee shunt)
- Shunt, Systemic to pulmonary, Other
- Conduit placement, RV to PA
- Conduit placement, LV to PA
- Cardiovascular catheterization procedure, Therapeutic, Balloon dilation
- Cardiovascular catheterization procedure, Therapeutic, Stent insertion
Supplemental Table 3. Multivariable Analysis of Early In-Hospital Shunt Failure in Neonates and Infants
INCLUDES Ventricle to Pulmonary Shunts (Sano) / EXCLUDES Ventricle to Pulmonary Shunts (Sano)
Risk Factor / OR (95% CI) / p-value / OR (95% CI) / p-value
Patient Factors
Prematurity / 0.80 (0.62,1.04) / 0.0909 / 0.84 (0.64,1.10) / 0.2146
Neonate (vs. infant) / 1.04 (0.46,2.34) / 0.9288 / 0.91 (0.38,2.17) / 0.8369
Weight, neonates (per kg decrease) / 1.35 (1.13,1.61) / 0.0007 / 1.30 (1.07,1.58) / 0.0076
Weight, infants (per kg decrease) / 1.34 (1.15,1.56) / 0.0002 / 1.35 (1.15,1.59) / 0.0002
Previous cardiac operations / 1.37 (0.90,2.08) / 0.1398 / 1.34 (0.85,2.09) / 0.2064
Any non-cardiac abnormalities / 1.29 (0.91,1.82) / 0.1569 / 1.37 (0.95,1.97) / 0.0940
Any genetic abnormalities / 1.09 (0.89,1.33) / 0.3945 / 1.01 (0.81,1.25) / 0.9240
Fundamental Dx: HLHS vs. other / 0.97 (0.72,1.33) / 0.8709 / 1.34 (0.92,1.95) / 0.1274
Fundamental Dx: non-HLHS single ventricle vs other / 1.02 (0.82,1.27) / 0.8490 / 1.08 (0.86,1.36) / 0.5147
Preoperative Factors
Mechanical circulatory support / 1.45 (0.62,3.36) / 0.3878 / 1.59 (0.67,3.74) / 0.2903
Shock, persistent at time of operation / 0.85 (0.46,1.57) / 0.6067 / 0.86 (0.44,1.66) / 0.6433
Preoperative Renal failure requiring dialysis / 1.03 (0.58,1.82) / 0.9207 / 0.86 (0.43,1.73) / 0.6777
Preoperative mechanical ventilatory support / 1.12 (0.93,1.35) / 0.2451 / 1.16 (0.94,1.43) / 0.1688
Preoperative Neurological deficit / 1.11 (0.51,2.39) / 0.7986 / 0.99 (0.41,2.42) / 0.9909
Coagulopathy, hypercoagulable state / 2.47 (1.09,5.60) / 0.0306 / 2.79 (1.20,6.48) / 0.0167
Coagulopathy, intrinsic hypocoagulable state / 0.78 (0.29,2.07) / 0.6187 / 0.64 (0.21,1.91) / 0.4218
Any other preoperative risk factor / 1.24 (1.01,1.53) / 0.0378 / 1.22 (0.97,1.53) / 0.0859
Operative Factors
Operation type (CPB vs. non-CPB) / 1.05 (0.83,1.33) / 0.6857 / 1.07 (0.84,1.35) / 0.5997
Shunt type: Central vs. mBTS / 1.30 (0.99,1.69) / 0.0569 / 1.30 (0.99,1.71) / 0.0560
Shunt type: Norwood - mBTS vs. mBTS / 1.27 (0.91,1.76) / 0.1578 / 1.05 (0.73,1.52) / 0.7789
Shunt type: Norwood – Ventricle to pulmonary vs. mBTS / 0.65 (0.45,0.93) / 0.0200
Shunt type Other vs. mBTS / 1.07 (0.68,1.69) / 0.7654 / 1.07 (0.68,1.70) / 0.7565
Shunt type: TAPVC vs. mBTS / 1.20 (0.63,2.30) / 0.5754 / 1.18 (0.61,2.27) / 0.6159
HLHS = hypoplastic left heart syndrome; CPB = cardiopulmonary bypass; mBTS = modified Blalock-Taussig shunt; TAPCV = total anomalous pulmonary venous connection
Supplemental Results
Association of Early Shunt Failure with Gender
The entire study cohort was composed of 5,377 males and 3,785 females. The event rate for the endpoint Early Shunt Failure was 6.6% among males and 8.4% among females. For both neonatal and infant subgroups, weight was lower among females than males. And, the multivariable analysis identified lower weight at operation as a risk factor for in-hospital shunt failure in both neonates and infants (OR=1.35, p=0.0007 and OR=1.34, p=0.0002, respectively). In addition, patients receiving a right ventricle-to-pulmonary artery shunt, which was the operative procedural category associated with the lowest rate of occurrence of in-hospital shunt failure, accounted for 23% (870/3785) of the female neonates and infants while accounting for 25% (1361/5377) of the male neonates and infants. This may be in part a reflection of the fact that the HLHS diagnostic group accounted for a smaller fraction of female patients than male patients (29.4% vs 34.7%). Thus, the observation that shunt failure was more prevalent among females than males may be reflective of lower weight at operation among females (a covariate that was a risk factor for in-hospital shunt failure by multivariable analysis), and in addition may be related to the smaller fraction of all female shunt patients receiving a systemic-ventricle-to-pulmonary artery shunt in the context of a Norwood procedure. Thus females were “under-represented” in the procedural group that was associated with the lowest risk of early shunt failure.
Shunt Failure and Diagnostic Category
From the standpoint of diagnostic groups, we intentionally classified as “other” those patients who did not have HLHS or one of the additional specific single ventricle diagnoses for which there are unique codes in the STS CHSD (e.g. “Single Ventricle Tricuspid Atresia,” “Single Ventricle Mitral Atresia,” “Single Ventricle DILV,” among others). Thus, it was not the intent that “other” would be synonymous with, or equivalent to “biventricular.” The “other” diagnostic category includes many cases for which eventual achievement of biventricular physiology after subsequent reparative surgery is highly likely (e.g. tetralogy of Fallot), but it also includes cases with diagnoses such as Ebstein anomaly or pulmonary atresia with intact ventricular septum, for whom either univentricular or biventricular physiology may be the eventual outcome. As such, the data, as presented, are not intended to support inferences regarding the relative risk of shunt failure in relation to the potential for biventricular vs univentricular circulation.
A more granular examination of the “other” diagnostic category led to the following observations:
Among the most prevalent fundamental diagnoses in this category are: Pulmonary Atresia w/ Intact Ventricular Septum, Pulmonary Atresia w/ VSD +/- MAPCAs, tetralogy of Fallot w/ Pulmonary Stenosis, Ebstein Anomaly, Double Outlet Right Ventricle (various sub-types), and TGA/VSD +/- LVOTO. Our interpretation of the available data is that the event rate (early shunt failure) is lower in the HLHS group than the rest, and roughly comparable between the non-HLHS Single Ventricle and “Other” groups (7.8% vs 7.7%). We feel that this is consistent with the observation that a substantial fraction of HLHS patients received right ventricle-to-pulmonary artery shunts, and that this procedural strategy was observed to be associated with a lower rate of occurrence of the Shunt Failure endpoint (5.2%; Table 1).
Supplemental Figure 1. Shunt volume and Shunt Failure Rate by Center
Supplemental figure 1: Scatter plot displaying the number of eligible shunt patients by center (n; horizontal axis)* and the early shunt failure rate (%; vertical axis).
*Sites with <5 shunt records are excluded due to the instability of their rate calculations.