Allogeneic Transfusion-related Adverse Events in Pediatric Idiopathic Scoliosis Surgery Authors: Vivien Chan, Geoffrey Shumilak, Armaan K Malhotra, Andy M Liu, Andrew Chan-Tai-Kong, Jalen Dansby, David E Lebel, David L Skaggs Published: 2026 ABSTRACT AbstractBackgroundPediatric spinal deformity surgery can involve significant perioperative blood loss, and allogeneic blood transfusion has been linked to an increased risk of adverse events. This study aimed to quantify adverse events associated with allogeneic blood transfusion in pediatric idiopathic scoliosis surgery.Study design/settingRetrospective observational cohort study.MethodsThe National Surgical Quality Improvement Program (NSQIP) Pediatric database (2016-2022) was queried. Patients <18 years undergoing posterior arthrodesis with 7 or more surgical levels for idiopathic scoliosis were included. The primary outcome was a composite 30-day adverse event outcome. Chi-square tests were used to compare the rates of adverse events associated with allogeneic blood transfusion. A multivariable logistic regression analysis was performed to assess the association between allogeneic blood transfusion and the occurrence of adverse events.ResultsA total of 23,106 patients were included in this study, of which 12.3% received allogeneic transfusion. The rate of the composite 30-day adverse event outcome was 1.1%. Patients that received allogeneic transfusion had a higher rate of 30-day adverse event (2.5% vs 0.9%; X2 ​= ​62.5, P ​< ​.001). Allogeneic transfusion was associated with a higher rate of renal failure, cardiac arrest, venous thromboembolism, deep wound infection/dehiscence, sepsis/septic shock, pneumonia, and unplanned reintubation. In a multivariable logistic regression analysis, allogeneic transfusion was associated with higher odds of the composite 30-day adverse event outcome (odds ratio [OR]: 2.09, P ​< ​.001).ConclusionsIn this large national pediatric registry cohort, allogeneic transfusion was associated with higher odds of 30-day adverse events after posterior spinal fusion for idiopathic scoliosis. Further prospective and center-level studies are needed to clarify transfusion thresholds, blood management strategies, and potential contributors to adverse outcomes.Key Concepts(1)In 23,106 pediatric idiopathic scoliosis patients from the American College of Surgeons National Surgical Quality Improvement Program, 12.3% received allogeneic transfusion and the overall 30-day composite adverse event rate was 1.1%.(2)Patients who received allogeneic transfusion had a significantly higher rate of composite 30-day adverse events compared to those who did not (2.5% vs 0.9%, P ​< ​.001).(3)Allogeneic transfusion was associated with higher rates of specific complications, including renal failure, cardiac arrest, venous thromboembolism, unplanned reintubation, deep wound infection/dehiscence, sepsis/septic shock, pneumonia, and 30-day mortality.(4)On multivariable logistic regression, allogeneic transfusion independently increased the odds of composite adverse events (OR: 2.09, 95% CI: 1.37-3.19, P ​< ​.001).Level of EvidenceLevel IIIKeywords: Pediatric, Scoliosis, Transfusion, Adverse events, Morbidity, Surgical outcomes Introduction Posterior arthrodesis for pediatric spinal deformity correction can be associated with significant blood loss and perioperative transfusion ​[1,2]. Historically, concerns with allogeneic blood transfusion were related to risks of communicable disease transmission and ABO mismatch ​[3]. Improved screening for transfusion-transmissible infections and hospital risk reduction protocols have significantly reduced the risk of these complications. However, allogeneic blood transfusion is being increasingly associated with negative clinical outcomes, including higher rates of infectious complications, poor postoperative functional recovery, and immunological reactions ​[2,4]. Compared to adults, transfusion in children has been associated with substantial risks due to a higher incidence of noninfectious adverse events and allergic reactions ​[[5], [6], [7], [8]]. Adverse events related to transfusion in children remains understudied ​[9]. The aim of this study was to investigate the association between allogeneic red blood cell transfusion and perioperative adverse events in pediatric patients receiving surgery for idiopathic scoliosis. Methodology Source of data The American College of Surgeons National Surgical Quality Improvement Program (NSQIP) Pediatric database was harmonized with the procedure-targeted spinal fusion dataset for years 2016–2022. NSQIP is a nationally validated, risk-adjusted, outcomes-based program that measures the quality of surgical care. In each participating hospital, a trained surgical clinical reviewer collects preoperative through 30-day postoperative data on randomly selected patients. Data are entered online in a secure, web-based platform available for open access by request. Participants Patients were included in this study if they were: 1) under the age of 18 years, 2) received posterior spinal arthrodesis with 7 or more surgical levels (CPT 22802 and 22804), and 3) had a diagnosis of idiopathic scoliosis. Patients were excluded if they had anterior approaches for deformity correction. Patients with kyphosis, neuromuscular scoliosis, congenital scoliosis, and syndromic scoliosis were excluded. Outcomes of interest and statistical analysis The primary outcome of this study was a composite outcome of 30-day adverse events. The composite 30-day adverse event outcome was defined as the occurrence of any of the following: acute kidney injury, unplanned reintubation, cardiac arrest requiring cardiopulmonary resuscitation, cerebrovascular accident, venous thromboembolism, catheter-related bloodstream infection, pneumonia, deep or organ space surgical site infection, wound dehiscence, sepsis, and septic shock. Renal failure is defined as a rise in serum creatinine to greater than or equal to 2 ​mg/dL from the patient's baseline and/or the acute need for renal replacement therapy. Deep wound infection is defined as an infection that occurs within 30 days after the operation and the infection appears to be related to the operation and infection involves deep soft tissues (eg ​fascial and muscle layers) of the incision and at least one of the following: 1) Purulent drainage from the deep incision but not from the organ/space component of the surgical site. 2) A deep incision spontaneously dehisces or is deliberately opened by a surgeon when the patient has at least one of the following signs or symptoms: fever (>38 ​C), localized pain, or tenderness, unless the site is culture-negative, 2) An abscess or other evidence of infection involving the deep incision is found on direct examination, during reoperation, or by histopathologic or radiologic examination, or 4) Diagnosis of a deep incision surgical site infection (SSI) by a surgeon or attending physician. A composite 30-day adverse event outcome was defined to capture the overall morbidity following posterior spinal fusion. While these events differ in mechanism, combining them allowed for sufficient statistical power to detect associations in the context of a low overall event rate. Secondary outcomes included the individual adverse events and mortality. Baseline patient demographics and surgical characteristics were reported using descriptive statistics with continuous variables reported as means with standard deviation and categorical variables reported as counts with proportions. The rates of allogeneic red blood cell transfusion and adverse events were characterized as proportions. In the adjusted analysis, Chi-square tests were used to compare the rates of adverse events between patients that received allogeneic transfusion and those that did not. A multivariable logistic regression analysis was performed to investigate the association between adverse events and allogeneic blood transfusion, controlling for age in years, sex, preoperative hematocrit<35, number of operative levels (7-12 levels vs ​≥13 levels), operative time in hours, use of antifibrinolytics, use of cell salvage, and total volume of blood transfused (allogeneic and cell-saver). A variable representing total transfusion volume (mL) was included in the multivariable model as a surrogate measure of intraoperative blood loss, which is not available within the NSQIP-Pediatric database. Total transfusion volume represented the combined volume of allogeneic transfusion and cell-salvage transfusion. The purpose of including total transfusion volume was to partially account for differences in overall transfusion requirements and surgical blood loss while evaluating the independent association between allogeneic transfusion exposure and postoperative adverse events. Variables for regression analyses were chosen a priori based on the literature and expert opinion. Listwise deletion was used for missing data for complete case analysis. Statistical analysis was performed using IBM SPSS Statistics version 29.0.2.0. Results Participants and surgical metrics A total of 23,106 patients were included in this study (Table 1). The mean age was 14.4 ​± ​1.95 years and 76.4% (n ​= ​17,646) were female. The mean operative time was 4.6 ​± ​1.6 ​h. There were 71.6% (n ​= ​16,539) that had seven to 12 surgical levels and 28.4% (n ​= ​6,567) that had 13 or more surgical levels. Table 1Patient characteristics and surgical factors.Table 1n ​= ​23,106Age (mean)14.4 y ​(SD: 2.0)Female76.4% (n ​= ​17,646)Preoperative hematocrit<357.1% (n ​= ​1,633)Surgical factors Operative time (mean)4.6 ​h (SD: 1.6) Number of surgical levels 7 to 12 levels71.6% (n ​= ​16,539) ≥13 levels28.4% (n ​= ​6,567)Intraoperative antifibrinolytics90.2% (n ​= ​20,841)Transfusion Allogeneic transfusion12.3% (n ​= ​2,843) Allogeneic transfusion volume (mean)56.70 ​mL (SD: 186.53) Cell-salvage used70.1% (n ​= ​16,190) Cell-salvage transfusion volume (mean)163.96 ​mL (SD: 192.37) Total transfusion volume (mean)220.67 ​mL (SD: 283.93)30-d ​adverse event1.1% (n ​= ​252)30-d ​mortality0.01% (n ​= ​3)SD, standard deviation. Patient characteristics and surgical factors. SD, standard deviation. Allogeneic blood transfusion was used in 12.3% (n ​= ​2,843). Cell-salvage was used in 70.1% (n ​= ​16,190). The mean volume of cell-salvage transfusion was 164.0 ​± ​192.4 ​mL. Mean total transfusion volume, including both cell-salvage and allogeneic blood transfusion, was 220.7 ​± ​283.9 ​mL. Intraoperative antifibrinolytics were used in 90.2% (n ​= ​20,841). The rate of the composite 30-day adverse events outcome was 1.1% (n ​= ​252). Of the 23,106 patients, 199 (0.9%) had one adverse event, 43 (0.2%) had two adverse events, 9 (0.04%) had three adverse events, and 1 (0.004%) had 4 adverse events. The 30-day mortality rate was 0.01% (n ​= ​3). Of the 23,106 patients, 30-day mortality was missing in 0.1% (n ​= ​23). Adverse events and 30-day mortality Patients that received allogeneic blood transfusion compared to patients that did not receive allogeneic blood transfusion had a significantly higher unadjusted rate of the composite 30-day adverse events outcome (Allogeneic transfusion: 2.5% vs ​no allogeneic transfusion: 0.9%; X2 ​= ​62.5, P ​< ​.001) (Table 2). Allogeneic blood transfusion was associated with a significantly higher rate of renal failure (allogeneic transfusion: 0.14% vs ​no allogeneic transfusion: 0.02%; X2 ​= ​8.6, P ​= ​.003), cardiac arrest requiring cardiopulmonary resuscitation (allogeneic transfusion: 0.14% vs ​no allogeneic transfusion: 0.01%; X2 ​= ​13.0 P ​< ​.001), and venous thromboembolism (allogeneic transfusion: 0.2% vs ​no allogeneic transfusion: 0.05%; X2 ​= ​9.4, P ​= ​.002) (Table 2). Patients who received allogeneic blood transfusion also had a higher rate of infectious complications, including deep wound infection or dehiscence (allogeneic transfusion: 1.1% vs ​no allogeneic transfusion: 0.6%; X2 ​= ​12.0, P ​< ​.001) and sepsis or septic shock (allogeneic transfusion: 0.5% vs ​no allogeneic transfusion: 0.1%; X2 ​= ​15.1, P ​< ​.001). Allogeneic blood transfusion was associated with a higher rate of respiratory adverse events such as pneumonia (allogeneic transfusion: 0.5% vs ​no allogeneic transfusion: 0.2%; X2 ​= ​24.0, P ​< ​.001) and unplanned reintubation (allogeneic transfusion: 0.6% vs ​no allogeneic transfusion: 0.06%; X2 ​= ​41.4, P ​< ​.001) (Table 2). There was no significant difference in rate of cerebrovascular accident (allogeneic transfusion: 0.0% vs ​no allogeneic transfusion: 0.005%; P ​= ​.708) (Table 2). There were no catheter-related bloodstream infections. There was a higher rate of 30-day mortality with allogeneic transfusion (allogeneic transfusion: 0.07% vs ​no allogeneic transfusion: 0.005%; X2 ​= ​8.2, P ​= ​.004). Table 2Rates of adverse events and mortality with allogeneic transfusion.Table 2Allogeneic transfusion (n ​= ​2,843)No allogeneic transfusion (n ​= ​20,263)X [2]P ​valueComposite 30-d ​adverse event outcome2.5% (n ​= ​72)0.9% (n ​= ​180)62.5<0.001∗Noninfectious adverse events Renal failure0.14% (n ​= ​4)0.02% (n ​= ​5)8.60.003∗ Unplanned reintubation0.6% (n ​= ​18)0.06% (n ​= ​34)41.4<0.001∗ Cardiac arrest0.14% (n ​= ​4)0.01% (n ​= ​3)13.0<0.001∗ Venous thromboembolism0.2% (n ​= ​6)0.05% (n ​= ​10)9.40.002∗ Cerebrovascular accident0.0% (n ​= ​0)0.005% (n ​= ​1)––Infectious adverse events Deep wound infection or dehiscence1.1% (n ​= ​32)0.6% (n ​= ​116)12.0<0.001∗ Sepsis or septic shock0.5% (n ​= ​13)0.1% (n ​= ​27)15.1<0.001∗ Pneumonia0.5% (n ​= ​14)0.2% (n ​= ​13)24.0<0.001∗ Catheter-related bloodstream infection0.0% (n ​= ​0)0.0% (n ​= ​0)––Mortality0.07% (n ​= ​2)0.005% (n ​= ​1)8.20.004∗∗ Denotes statistical significance. Rates of adverse events and mortality with allogeneic transfusion. ∗ Denotes statistical significance. Multivariable regression analysis on composite 30-day adverse event outcome In a multivariable logistic regression analysis, allogeneic blood transfusion was associated with a higher odds of the composite 30-day adverse event outcome (odds ratio [OR] ​= ​2.09 [95% confidence interval [CI] = 1.37-3.19], P ​< ​.001) (Table 3). Longer operative time (OR ​= ​1.13 [95% CI = 1.04-1.23], P ​= ​.005), ≥13 surgical levels (OR ​= ​2.25 [95% CI = 1.69-3.00], P ​< ​.001) were associated with higher odds of adverse events (Table 3). Female sex (OR ​= ​0.68 [95% CI = 0.50-0.94], P ​= ​.018) and use of intraoperative antifibrinolytics (OR ​= ​0.65 [95% CI = 0.42-0.98], P ​= ​.041) were associated with lower odds of adverse events (Table 3). Use of cell-salvage (P ​= ​.556), age in years (P ​= ​.658), preoperative hematocrit<35 (P ​= ​.284), and total blood volume transfused (P ​= ​.971) were not significantly associated with the odds of adverse events (Table 3). Table 3Multivariable logistic regression analysis for composite 30-d ​adverse event outcome.Table 3Odds ratio95% confidence intervalP-valueAge (y)1.017[0.945-1.095]0.658Female0.683[0.498-0.938]0.018∗Preoperative hematocrit<351.271[0.820-1.972]0.284Operative time (hours)1.128[1.037-1.226]0.005∗≥13 surgical levels (ref. [[7], [8], [9], [10], [11], [12]] levels)2.251[1.687-3.004]<0.001∗Intraoperative antifibrinolytics0.645[0.423-0.983]0.041∗Use of cell-salvage0.904[0.647-1.264]0.556Allogeneic transfusion2.090[1.367-3.194]<0.001∗Total blood volume transfused (mL)1.000[0.999-1.001]0.971∗ Denotes statistical significance. Multivariable logistic regression analysis for composite 30-d ​adverse event outcome. ∗ Denotes statistical significance. Discussion Posterior spinal fusion for pediatric scoliosis can be associated with major blood loss requiring allogeneic blood transfusion ​[1,4]. Historically, the concern with allogenic blood transfusion has been focused on transmission of viruses, bacterial contamination, and ABO mismatch ​[3]. The risk of these complications has been significantly reduced with improved blood collection, storage, testing, and implementation of hospital risk reduction protocols ​[10,11]. More recently, studies on adult patients have reported negative clinical outcomes associated with allogeneic blood transfusion, including higher infection rates and poor postoperative functional recovery ​[[12], [13], [14]]. Transfusion-related adverse events and the impact of allogeneic transfusion on clinical outcomes in pediatric patients remains understudied ​[9,15]. We found a higher rate of the 30-day composite adverse event outcome in patients that received allogeneic transfusion. Specifically, there was a higher rate of renal failure, cardiac arrest requiring cardiopulmonary resuscitation, unplanned reintubation, venous thromboembolism, deep wound infection/dehiscence, sepsis/septic shock, pneumonia, and mortality. Pediatric patients are physiologically and biochemically different from adults ​[16]. Compared to adults, pediatric patients have a higher reported incidence of transfusion reactions. A retrospective study reported a reaction rate of 528 per 100,000 red blood cell (RBC) transfusions in children, compared to 252 per 100,000 RBC transfusions in adults ​[16]. The authors noted higher rates of allergic reactions and febrile nonhemolytic reactions in pediatric patients [16]. Similarly, Wang et al. [15] reported a higher rate of rare transfusion reactions in children, including acute hemolytic transfusion reactions and transfusion-related acute lung injury. In our study, we examined the association between allogeneic transfusion and adverse events. Patients that received allogeneic transfusion had a significantly higher rate of adverse events (2.5% vs 0.9%). In our multivariable regression analysis, controlling for various confounding factors including preoperative hematocrit and total volume of blood transfused, which included cell-salvage transfusion volume and allogeneic transfusion volume, allogeneic transfusion was associated with higher odds of adverse events. However, it is likely that there are residual confounding variables not collected by the database, such as estimated blood loss, postoperative hematocrit, deformity severity, curve rigidity, transfusion thresholds, and center-level practice patterns. Additionally, patients requiring allogeneic transfusion may represent a higher-risk subgroup characterized by increased blood loss, greater physiologic stress, or operative complexity, which are factors that are not fully captured by the NSQIP database. Cell-salvage is an important component of perioperative blood management in pediatric spinal deformity surgery and may reduce exposure to allogeneic blood products. In the present cohort, cell-salvage was used in 70.1% of patients and was not independently associated with the composite 30-day adverse event outcome after adjustment for patient and surgical factors. These findings support the role of cell-salvage as part of a multimodal blood management strategy in idiopathic scoliosis surgery, alongside antifibrinolytic use, preoperative optimization, meticulous surgical technique, and standardized transfusion thresholds. However, the absence of an observed association between cell-salvage and adverse events should not be interpreted as proof of a protective effect, as patients selected for cell-salvage may differ from those who do not receive it, and NSQIP-Pediatric does not capture the granular intraoperative details needed to evaluate this relationship fully. Our findings are consistent with prior work by Minhas et al., [17] who used NSQIP-Pediatric data from 2012 to 2013 to evaluate complications after posterior spinal fusion for adolescent idiopathic scoliosis. Similar to the present study, they identified longer operative time and fusion of 13 or more levels as factors associated with postoperative complications. They also reported that transfusion volume of 20 ​mL/kg or greater was associated with an increased risk of adverse events. In the present analysis, total transfusion volume was included as an adjustment variable; however, we did not perform a dedicated threshold analysis to determine whether a specific transfusion-volume cutoff was associated with increased risk. Future studies should evaluate whether a dose-response relationship exists between transfusion volume and adverse events, and whether clinically meaningful thresholds can be identified after accounting for estimated blood loss, postoperative hematocrit, transfusion indications, and institutional blood management protocols. The primary outcome was a composite 30-day adverse event outcome to capture the overall postoperative morbidity after posterior spinal fusion for idiopathic scoliosis. However, the use of a composite outcome has important limitations ​as it combines heterogeneous complications that may differ substantially in mechanism, severity, and relationship to transfusion. The individual components of the composite outcome also vary considerably in their clinical significance, ranging from wound complications to life-threatening events such as cardiac arrest and mortality. Consequently, the observed association between allogeneic transfusion and the composite outcome should not be interpreted as evidence that transfusion has a uniform relationship with each individual complication. Rather, the composite outcome was selected to provide a global measure of postoperative morbidity and to improve statistical power in the setting of a low overall event rate. Therefore, we also examined individual adverse events to better understand their specific relationships with allogeneic transfusion. When we looked at specific adverse events, we found a higher rate of renal failure, cardiac arrest requiring cardiopulmonary resuscitation, unplanned reintubation, venous thromboembolism, and mortality in patients that received allogeneic transfusion. Several biologically plausible mechanisms have been proposed in the literature to explains associations between allogeneic transfusion and postoperative adverse outcomes. Potential mechanisms include transfusion-related immunomodulation, immune-mediated injury, and proinflammatory responses. Allogeneic transfusion has been hypothesized to contribute to renal injury from donor leukocytes initiating an immune response against host kidneys or from donor leukocytes secreting inflammatory cytokines and cytotoxic proteins ​[18]. Transfusion-related immunomodulation from release of inflammatory cytokines and mediators, such as interleukin-6 and interleukin-10, has been associated with systemic inflammation, endothelial dysfunction, and activation of coagulation pathways ​[[19], [20], [21]]. This may contribute to a hypercoagulable state and has been proposed as potential mechanisms underlying the increased risk of myocardial infarction and venous thromboembolism ​[[19], [20], [21], [22], [23], [24], [25], [26]]. Other hypothesized mechanisms include transfusion-associated circulatory overload, transfusion-related acute lung injury, and hemolytic transfusion reactions, which may contribute to cardiopulmonary complications ​[27,28]. A meta-analysis of adult spine surgery patients reported similar findings of increased morbid complications in patients that received perioperative allogeneic blood transfusion ​[29]. Although our study found higher rates of adverse events associated with transfusion, the relationship is an association and cannot be interpreted as causative. Given the observational nature of our study, the mechanisms underlying this association remain incompletely understood. Transfusion-associated postoperative infections have been described in various surgical subspecialties, including cardiac surgery, colorectal surgery, and spine surgery ​[[30], [31], [32]]. A systematic review of 13 studies by Zhou et al. [27] concluded a higher rate of postoperative infections in adult spine surgery patients that received perioperative blood transfusion. This association between allogeneic transfusion and postoperative infection in pediatric spine patients remains understudied. In our study, patients that received allogeneic transfusion had a higher rate of infectious adverse events, including deep wound infection or dehiscence, sepsis/septic shock, and pneumonia. This is in keeping with results from a study by Eisler et al. [33] on postoperative infections in pediatric patients after spinal fusion that reported a higher rate of any infection associated with allogeneic transfusion. Various mechanisms for increased rate of bacterial infections after allogeneic transfusion have been proposed. Donor white blood cells that are immunologically active can suppress the recipient's immune function ​[34]. Additionally, soluble biologic response modifiers are released, in a time-dependent manner, from white blood cell granules and their cell membranes into the supernatant of red blood cells during storage ​[35]. These mechanisms cause suppressed immune function, potentially leading to increased vulnerability to bacterial infections. Patients who require allogeneic transfusion may have experienced greater blood loss, longer operative time, more extensive surgery, or other perioperative factors that also increase infection risk. As such, the observed association supports continued attention to blood conservation and perioperative optimization, while also highlighting the need for prospective studies with more granular clinical data. The present findings add to the literature by identifying an association between allogeneic transfusion and rare 30-day adverse events in a large cohort of pediatric idiopathic scoliosis patients undergoing posterior spinal fusion. To our knowledge, this is the first study investigating transfusion-related adverse events in pediatric spine surgery patients using a large registry database. We found a significantly higher rate of noninfectious and infectious adverse events in patients that received allogeneic transfusion. Rather than establishing allogeneic transfusion as a direct cause of these events, these results suggest that transfusion may identify a subgroup of patients with increased perioperative risk. This distinction is important when interpreting registry-based findings. Clinically, these data support continued efforts to reduce unnecessary allogeneic transfusion through evidence-based blood management strategies, including preoperative optimization, antifibrinolytic use, cell-salvage when appropriate, and standardized transfusion thresholds. Further work is needed to determine whether specific transfusion practices, institutional protocols, or center-level factors contribute to variation in transfusion rates and adverse outcomes. This study has several limitations. First, this was a retrospective observational study using registry data, and therefore the findings demonstrate association rather than causation. Although allogeneic transfusion was associated with higher odds of adverse events, transfusion may be a marker of greater blood loss, physiologic stress, operative complexity, or other unmeasured factors. Second, NSQIP Pediatric does not include several clinically important variables, including estimated blood loss, postoperative hematocrit, transfusion thresholds, hemodynamic instability, radiographic deformity severity, curve rigidity, surgical technique, surgeon experience, or detailed perioperative decision-making. Although we adjusted for available markers of complexity, including operative time and number of fusion levels, residual confounding remains likely. Third, NSQIP Pediatric does not release center identifiers, surgeon identifiers, or center-level case volume. Therefore, we were unable to assess whether adverse events or transfusion practices clustered within specific centers, whether high- or low-performing centers contributed disproportionately to the observed findings, or whether institutional volume or expertize modified the relationship between transfusion and adverse events. Fourth, indications for transfusion were not standardized across institutions, and we cannot determine whether transfusions were administered according to restrictive or liberal thresholds ​[36,37]. Finally, individual adverse events were rare, and the composite adverse event outcome combined complications with heterogeneous mechanisms. For these reasons, the individual complication findings should be interpreted cautiously and primarily as hypothesis-generating. Author contributions Vivien Chan: Writing – review & editing, Writing – original draft, Visualization, Validation, Software, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Geoffrey Shumilak: Writing – review & editing, Writing – original draft, Supervision, Methodology, Formal analysis, Conceptualization. Armaan K. Malhotra: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology, Investigation, Formal analysis. Andy M. Liu: Writing – review & editing, Writing – original draft, Validation, Methodology, Investigation, Formal analysis. Andrew Chan-Tai-Kong: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology, Investigation, Data curation. Jalen Dansby: Writing – review & editing, Writing – original draft, Validation, Methodology, Formal analysis, Conceptualization. David E. Lebel: Writing – review & editing, Validation, Supervision, Methodology, Investigation, Formal analysis, Data curation. David L. Skaggs: Writing – review & editing, Validation, Supervision, Resources, Methodology, Investigation, Formal analysis, Conceptualization. Ethics approval and consent The author(s) declare that no patient consent was necessary as no images or identifying information are included in the article. Data availability The NSQIP Pediatric database is available for use by all participating NSQIP centers. Funding This project was not funded. Declaration of competing interests G.S., A.K.M., J.D., A.M.L, A.C. and D.E.L. declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. V.C. is a consultant for OrthoPpediatrics and Depuy Synthes. DLS is a consultant for Highridge, Globus Medical, and Orthobullets and receives royalties from Medtronic, Highridge, Globus Medical, and Wolters Kluwer Health. DLS has received grant funding from Pediatric Spine Study Group and Safety in Spine Surgery. DLS has patents with Medtronic and Highridge. DLS has stock options in Orthobullets and Green Sun Medical. Acknowledgments None. References Associated Data Data Availability Statement The NSQIP Pediatric database is available for use by all participating NSQIP centers.