Annulus-Sparing Tetralogy of Fallot Repair: Low Risk and Benefits to Right Ventricular Geometry.

BACKGROUND: Annulus-sparing repair of tetralogy of Fallot (TOF) carries a potential increased risk of reoperation for restenosis and unproven benefits on right ventricular (RV) geometry. METHODS: Primary TOF repairs (n = 434) between 2000 and 2012 were studied using risk-adjusted parametric techniques. Progression of cardiac dimensions was analyzed using repeated measures regression using reports of all 2,103 echocardiograms undertaken throughout the study period, to a maximum follow-up of 13.7 years. RESULTS: Repair was at a mean age of 180 days: AS approach in 296 (68%) patients; and transannular patch in 138 (32%). Intraoperative revisions (for residual stenosis) were required in 135 patients (29%). There have been 4 deaths (survival 99%). Surgical reoperation for recurrent right ventricular outflow tract stenosis was occasionally required in both groups at comparable rates (transannular patch, 5 of 136 [4%]; annulus-sparing repair, 14 of 296 [5%]; p = 0.83). Larger increases in RV end-diastolic dimensions were evident in transannular patch patients versus annulus-sparing repair patients (p < 0.0001). Other risks for RV dilation included worse grade of postoperative pulmonary regurgitation, larger right ventricular end-diastolic dimension at the time of diagnosis, and higher operative weight (all p < 0.0001). Factors associated with successful annulus-sparing repair included (1) pulmonary annulus greater than 7 mm, right ventricular end-diastolic dimension greater than 1.2 cm, and tricuspid annulus greater than 1.4 cm (all preoperatively); and (2) right ventricular outflow tract diameter greater than 10 mm and right ventricular systolic pressure less than 50 mm Hg (both intraoperatively after repair). CONCLUSIONS: Pursuit of annulus-sparing repair strategies can lower the use of transannular patch to approximately 30% with low risk of reoperation for the patient. Annulus-sparing repair is associated with significantly reduced long-term RV dilation. Pulmonary valve enlargement to approximately 10 mm and right ventricular systolic pressure less than 50 mm Hg during annulus-sparing repair are associated with low risk of recurrent stenosis.

Young infants with severe tetralogy of Fallot: Early primary surgery versus transcatheter palliation.

BACKGROUND: Infants with severe tetralogy of Fallot may undergo (1) early primary surgical repair (EARLY) or (2) early transcatheter palliation (CATH) before delayed surgical repair. We compared these strategies with (3) elective single-stage tetralogy of Fallot repair (IDEAL). METHODS: From 2000 to 2012, 453 children underwent tetralogy of Fallot repair (excluding systemic-pulmonary shunts), including 383 in the IDEAL (75%), 42 in the EARLY (9%), and 28 in the CATH (6%) groups. IDEAL repair at The Hospital for Sick Children occurs after 3 months. Risk-adjusted hazard analysis compared freedom from surgical or catheter reintervention. Somatic size, branch pulmonary artery size, and right ventricle systolic pressure were modeled using 2780 echocardiogram reports via mixed-model regression. RESULTS: CATH involved right ventricular outflow tract stent in 18 patients, right ventricular outflow tract balloon in 9 patients, and ductal-stent in 1 patient. Three patients died (1 per group). Risk-adjusted freedom from surgical reoperation was 89% ± 4%, 88% ± 5%, and 85% ± 6% for the IDEAL, EARLY, and CATH groups, respectively, at 10 years. Patients in the EARLY and CATH groups had similar reoperation rates, except for neonates (<1 month), for whom EARLY repair conferred an increased risk of reoperation. Risk-adjusted freedom from catheter reintervention was lower in the EARLY group (76%) and especially for the CATH group (53%) at 10 years versus the IDEAL group (83%). Somatic growth and progression of right ventricle systolic pressure were similar among groups at 8 years. Although those undergoing EARLY (P = .02) and CATH (P = .09) tend to have smaller branch pulmonary arteries initially, late pulmonary artery size was not significantly different among groups. CONCLUSIONS: Early primary repair for neonates may increase surgical reoperation, whereas transcatheter palliation comes at a cost of increased catheter reintervention. However, overall outcomes between groups, in terms of survival, growth, and hemodynamic parameters, were comparable, suggesting that both strategies are a reasonable option for children with severe tetralogy of Fallot.

NASA Model of "Threat and Error" in Pediatric Cardiac Surgery: Patterns of Error Chains.

BACKGROUND: We introduced the National Aeronautics and Space Association threat-and-error model to our surgical unit. All admissions are considered flights, which should pass through stepwise deescalations in risk during surgical recovery. We hypothesized that errors significantly influence risk deescalation and contribute to poor outcomes. METHODS: Patient flights (524) were tracked in real time for threats, errors, and unintended states by full-time performance personnel. Expected risk deescalation was wean from mechanical support, sternal closure, extubation, intensive care unit (ICU) discharge, and discharge home. Data were accrued from clinical charts, bedside data, reporting mechanisms, and staff interviews. Infographics of flights were openly discussed weekly for consensus. RESULTS: In 12% (64 of 524) of flights, the child failed to deescalate sequentially through expected risk levels; unintended increments instead occurred. Failed deescalations were highly associated with errors (426; 257 flights; p < 0.0001). Consequential errors (263; 173 flights) were associated with a 29% rate of failed deescalation versus 4% in flights with no consequential error (p < 0.0001). The most dangerous errors were apical errors typically (84%) occurring in the operating room, which caused chains of propagating unintended states (n = 110): these had a 43% (47 of 110) rate of failed deescalation (versus 4%; p < 0.0001). Chains of unintended state were often (46%) amplified by additional (up to 7) errors in the ICU that would worsen clinical deviation. Overall, failed deescalations in risk were extremely closely linked to brain injury (n = 13; p < 0.0001) or death (n = 7; p < 0.0001). CONCLUSIONS: Deaths and brain injury after pediatric cardiac surgery almost always occur from propagating error chains that originate in the operating room and are often amplified by additional ICU errors.

Aggressive Patch Augmentation May Reduce Growth Potential of Hypoplastic Branch Pulmonary Arteries After Tetralogy of Fallot Repair.

BACKGROUND: Potential surgical strategies for hypoplastic branch pulmonary arteries (BPAs) during tetralogy of Fallot repair include (1) extensive patch augmentation to the hilum (PATCH), (2) limited extension arterioplasty to the proximal pulmonary artery (EXTENSION), or (3) leaving the native vessels unaugmented (NATIVE). We explored the effect of these strategies on reintervention and BPA growth. METHODS: From 2000 to 2012, 434 children underwent complete tetralogy of Fallot repair. Risk-adjusted parametric models were used to analyze the risk of BPA reintervention for (1) all children, (2) children with BPAs of 4 mm or smaller, and (3) children with BPAs of 3 mm or smaller. Repeated-measures analysis of more than 2,000 echocardiograms was used to characterize postoperative BPA growth and right ventricular pressure by using nonlinear mixed models. RESULTS: Overall survival (99% [3 deaths]) was excellent. The 10-year freedom from BPA reintervention was 84%. In risk-adjusted models (including baseline BPA z-score), PATCH had a decreased freedom from reintervention (73%; p < 0.01) vs EXTENSION (87%) or NATIVE (91%). For children with BPAs of 4 mm or smaller (28 PATCH, 60 EXTENSION, 75 NATIVE), baseline characteristics were similar. The risk-adjusted 5-year freedom from reintervention was 68% for PATCH, 76% for EXTENSION, and 85% for NATIVE. PATCH trended toward an increased risk of reintervention (p = 0.07). For children with BPAs of 4 mm or smaller left in their NATIVE state, only ∼15% required reintervention. After adjustment for baseline BPA z-score, the time-related BPA growth was decreased (p < 0.014) and right ventricular pressure was increased (p = 0.03) for the PATCH group. CONCLUSIONS: Aggressive PATCH augmentation of hypoplastic BPAs improves the short-term geometry but may lead to late stenosis and higher rates of reintervention. Hypoplastic BPAs in tetralogy of Fallot tend (∼85%) to grow well without instrumentation.

Aortic valve repair for insufficiency in older children offers unpredictable durability that may not be advantageous over a primary Ross operation†.

OBJECTIVES: To evaluate the durability of aortic valve (AoV) repair relative to other strategies for children with significant aortic insufficiency (AI). METHODS: From 2001 to 2012, 90 children with greater than or equal to moderate AI underwent surgery. Resulting procedures were classified according to final operative outcome: AoV repair (repair; n = 46, 51%), Ross procedure (Ross; n = 21, 23%) or replacement with mechanical or tissue prosthesis [aortic valve replacement (AVR); n = 23, 26%]. Repeated measures (n = 1081 echocardiograms) mixed-model analysis and parametric multiphase risk-adjusted hazard analysis were used to evaluate haemodynamic parameters and durability of operations. RESULTS: Mean age at operation was similar for repair and Ross groups, but slightly higher for the AVR group (10.6, 11 and 13.2, respectively; P = 0.04). Baseline annular dimensions were similar among groups. Of 46 repairs, 85% involved pericardial leaflet extensions (commonly with leaflet shaving and/or commisuroplasty). The remaining repairs were commissuroplasties. On multivariable analysis, repair was associated with increased early (∼1-2 years) AI and increased outflow tract peak pressure gradients relative to Ross and AVR procedures. On univariate analysis, repairs tended to have a larger annulus size compared with Ross or AVR; however, this was not significant on multivariable analysis. There were 25 reinterventions (surgical reoperation = 16; transcatheter intervention = 9) for 22 children. Freedom from surgical reoperation was 64, 100 and 51% at 6 years for repairs, Ross and AVR, respectively (P = 0.05); however, three of five reoperations after AVR were for failed bioprosthetic devices. The freedom from reintervention was not significantly influenced by the type of AoV operation (P = 0.43). CONCLUSIONS: Durability of aortic valve repair for children is limited by recurrence of AI and/or stenosis, often within the first few years. After repair, reoperation should be anticipated within ∼7 years.

National Aeronautics and Space Administration "threat and error" model applied to pediatric cardiac surgery: error cycles precede ∼85% of patient deaths.

BACKGROUND: We hypothesized that the National Aeronautics and Space Administration "threat and error" model (which is derived from analyzing >30,000 commercial flights, and explains >90% of crashes) is directly applicable to pediatric cardiac surgery. METHODS: We implemented a unit-wide performance initiative, whereby every surgical admission constitutes a "flight" and is tracked in real time, with the aim of identifying errors. The first 500 consecutive patients (524 flights) were analyzed, with an emphasis on the relationship between error cycles and permanent harmful outcomes. RESULTS: Among 524 patient flights (risk adjustment for congenital heart surgery category: 1-6; median: 2) 68 (13%) involved residual hemodynamic lesions, 13 (2.5%) permanent end-organ injuries, and 7 deaths (1.3%). Preoperatively, 763 threats were identified in 379 (72%) flights. Only 51% of patient flights (267) were error free. In the remaining 257 flights, 430 errors occurred, most commonly related to proficiency (280; 65%) or judgment (69, 16%). In most flights with errors (173 of 257; 67%), an unintended clinical state resulted, ie, the error was consequential. In 60% of consequential errors (n = 110; 21% of total), subsequent cycles of additional error/unintended states occurred. Cycles, particularly those containing multiple errors, were very significantly associated with permanent harmful end-states, including residual hemodynamic lesions (P < .0001), end-organ injury (P < .0001), and death (P < .0001). Deaths were almost always preceded by cycles (6 of 7; P < .0001). CONCLUSIONS: Human error, if not mitigated, often leads to cycles of error and unintended patient states, which are dangerous and precede the majority of harmful outcomes. Efforts to manage threats and error cycles (through crew resource management techniques) are likely to yield large increases in patient safety.