Neither bold nor insolent. The safety of routine primary sternal closure after the Norwood procedure is the central question addressed in the article under discussion. Routine primary sternal closure after the Norwood procedure is a relevant issue in complex neonatal cardiac surgery, particularly because many centers still perform delayed closure. Traditionally, the arguments supporting delayed rather than primary closure have included its hemodynamic and respiratory advantages, despite a higher risk of infectious complications and longer hospital stay. The uncertainty between these two approaches and the lack of consensus in the literature provide the rationale for this study.
Against this background, a tertiary center in the United States, where primary sternal closure is systematically attempted, presents its experience, analyzes its outcomes, and describes the indications used to proceed with, or avoid, this strategy according to different clinical parameters. The primary outcomes were operative survival and infectious complications. Secondary outcomes included the need for postoperative mechanical circulatory support, duration of mechanical ventilation, and hospital length of stay.
To analyze primary sternal closure after Norwood surgery, it is useful to understand its alternative: delayed sternal closure. This strategy is widely used to reduce the adverse effects of myocardial dysfunction and edema after CPB and XC. In the United States, 74% of patients undergoing the Norwood procedure leave the operating room with an open chest. Sternal closure is associated with reduced cardiac output and pulmonary compliance, particularly in neonates, in whom the heart is proportionally larger relative to the thoracic cage than in adults. Neonates with single-ventricle physiology have limited cardiac reserve and are more vulnerable to adverse cardiopulmonary interactions. Therefore, the hemodynamic and respiratory benefits of delayed closure may offset the increased infectious risk and the need for additional procedures to achieve definitive sternal closure. Current literature offers no consensus regarding the most appropriate chest-closure strategy after stage 1 single-ventricle palliation, although many centers routinely favor delayed closure, in contrast to the institution presenting this study, where routine closure is primarily performed.
A retrospective, single-center study was designed and conducted between 2017 and 2022, including 116 patients who underwent the Norwood procedure. Two study groups were considered: primary sternal closure, comprising 80 patients (68.9%), and delayed sternal closure, comprising 36 patients. To allocate patients to each group, high-risk clinical criteria were defined, with their presence favoring delayed closure. These high-risk criteria included baseline features such as prematurity, low birth weight, tricuspid regurgitation, or right ventricular dysfunction; preoperative criteria such as the need for inotropes or mechanical circulatory support, renal dysfunction, necrotizing enterocolitis, stroke, or preoperative atrial septal intervention; and intraoperative criteria such as procedural complexity, CPB time, XC time, circulatory arrest time, multiple CPB runs, excessive bleeding, space considerations, hemodynamic instability, poor respiratory status, need for catheterization or mechanical circulatory support, severe arrhythmia, tricuspid regurgitation, or right ventricular dysfunction. Intraoperative considerations, assessed jointly by the surgeon and anesthesiologist, were the most decisive factors. Statistical analysis was then performed using chi-square tests for categorical variables, Kruskal-Wallis tests for continuous variables, and multivariable logistic regression to assess the primary and secondary outcomes.
At the preoperative level, baseline clinical characteristics and echocardiographic findings were analyzed, with no differences between groups (p = .03), except for a lower preoperative need for inotropes in the primary closure group.
At the perioperative level, among patients undergoing primary closure, fewer required multiple CPB runs (p < .001), and there was also a lower need for postoperative circulatory support (p < .001) and a shorter duration of mechanical ventilation (3 vs 8 days; p < .001). Hospital survival (91.3% vs 66.7%; p = .04), hospital length of stay (29 vs 78 days; p < .01), surgical site infection (p = .33), and sepsis (p = .38) were also lower in the primary closure group. Within this group, 6.2% required sternal reopening (5 patients).
Third, multivariable analysis yielded relevant findings for the assessment of the primary outcomes: delayed closure was not an independent risk factor for increased hospital mortality (p = .63), nor was it independently associated with infectious complications (p = .15). Only the need for postoperative circulatory support (p < .01) and chromosomal abnormalities (p < .01) increased the risk of hospital mortality. Chromosomal abnormalities were also a risk factor for infectious complications (p = .01).
Finally, the authors describe the indications they followed for delayed closure and the reasons that led to reopening in patients who had undergone primary closure. The main indications for delayed closure were hemodynamic instability (12 patients), intolerance of closure (7 patients), prolonged CPB time (5 patients), elective decision (4 patients), ECMO (3 patients), bleeding (2 patients), arrhythmia (1 patient), tissue edema (1 patient), and high-risk status (1 patient). The indications for sternal reopening were hemodynamic instability requiring ECMO, a concomitant intra-abdominal process, shunt thrombosis, residual aortic arch obstruction, and Sano revision.
In their conclusions, the authors state that primary sternal closure can be performed in most neonates after the Norwood procedure with favorable outcomes. Selected patients with high-risk preoperative or intraoperative features may benefit from delayed sternal closure. Delayed sternal closure does not independently increase the risk of mortality or infectious complications.
COMMENTARY:
The authors report that outcomes among patients undergoing delayed closure were worse than previously published, but they justify this by emphasizing the patient-selection strategy used to allocate patients to the delayed closure group. Delayed closure was reserved for patients with a worse and more complex clinical status, which in itself made poorer outcomes more likely. They consider this a strength, as it reinforces the effectiveness of their patient-selection process based on the high-risk criteria they used. This point is supported by the multivariable analysis, which did not identify delayed closure as an independent factor for either hospital mortality or increased infectious complications.
The authors also emphasize the 90% success rate achieved with routine primary closure and the low number of reopenings recorded, which may translate into patient benefit and improved healthcare-system efficiency. This is a positive aspect, although it implies a considerable degree of team specialization and procedural volume.
Therefore, two important conclusions can be drawn from this study. First, the mortality observed in the delayed closure group cannot be attributed to the sternal-management strategy itself, but rather reflects the greater severity of these patients or nonmodifiable baseline factors, such as chromosomal abnormalities; therefore, delayed closure should still be used when the clinical situation requires it. Second, primary sternal closure appears to be a safe strategy for most neonates undergoing the Norwood procedure. Both points may be regarded as favorable and challenge routine delayed sternal closure after the Norwood procedure as the most widely established practice. However, several critical issues should also be acknowledged: these results come from a single tertiary center and reflect the management of a specific medical team, rather than a reproducible protocol that can be implemented universally. Team experience and pathophysiological interpretation played a major role in decision-making, but a precise and reproducible algorithm would be needed to standardize the strategy and shift the broader paradigm. Finally, the structural limitation of its retrospective design should be mentioned, as it is more prone to selection bias and potential confounding factors.
REFERENCE:
Mills MF, John MM, McKiernan M, Chanani N, Horan V, Rosenblum JM, et al. Routine Primary Sternal Closure After the Norwood Procedure. World J Pediatr Congenit Heart Surg. 2026 Mar;17(2):177-184. doi: 10.1177/21501351251363167. Epub 2025 Sep 9. PMID: 40924785.
