The Society of Thoracic Surgeons Congenital Heart Surgery Database: 2022 Update on Outcomes and Research

Volume Alone Does Not Predict Quality Outcomes in Hospitals Performing Pediatric Cardiac Surgery

BACKGROUND

Lower institutional volume has been associated with inferior pediatric cardiac surgery outcomes. This study explored the variation in mortality rates among low-, mid-, and high-volume hospitals performing pediatric cardiac surgery in the United States.

METHODS

The Kids' Inpatient Database was explored for the years 2016 and 2019. Hospitals performing only off-bypass coarctation and ventricular septal defect repair were omitted. The hospitals were divided into 3 groups by their annual case volume. Multivariable logistic regression models were fit to obtain risk-adjusted in-hospital mortality rates.

RESULTS

A total of 25,749 operations performed by 235 hospitals were included in the study. The risk-adjusted mortality rate for the entire sample was 1.9%. There were 140 hospitals in the low-volume group, 64 hospitals in the mid-volume group, and 31 in the high-volume group. All groups had low-mortality (mortality <1.9%) and high-mortality (mortality >1.9%) hospitals. Among low-volume hospitals, 53% were low-mortality (n = 74) and 47% were high-mortality (n = 66) hospitals. Among mid-volume hospitals, 58% were low-mortality (n = 37) and 42% were high-mortality (n = 27) hospitals. Among high-volume hospitals, 68% were low-mortality (n = 21) and 32% were high-mortality (n = 10) hospitals. There was no statistically significant difference in risk-adjusted in-hospital mortality when comparing low-, mid-, and high-volume centers for 7 Society of Thoracic Surgeons benchmark procedures.

CONCLUSIONS

This national, real-world, risk-adjusted volume outcome analysis highlights that volume alone may not be the sole arbiter to predict quality of pediatric cardiac surgery outcomes. Using case volume alone as a surrogate for quality may unfairly asperse high-performing, low-volume programs.

The Society of Thoracic Surgeons Congenital Heart Surgery Database: 2023 Update on Outcomes and Research

The Society of Thoracic Surgeons (STS) Congenital Heart Surgery Database (CHSD) continues to be the most comprehensive database of congenital and pediatric cardiothoracic surgical procedures in the world and contains information on 664,210 operations as of June 30, 2023. The 35th harvest of the STS CHSD data was undertaken in Spring 2023, spanning the 4-year period January 1, 2019, through December 31, 2022, and included 144,919 operations performed at 114 participating sites in North America. The harvest analysis was successfully executed by the STS Research and Analytic Center. The overall unadjusted mortality rate was 2.68% and has remained stable over the 4 years included in the current harvest window. Mortality is highest in neonates (7.4%) and lowest in children (1.1%). As in prior analyses, observed mortality and postoperative length of stay in the database increase with an increase in STS-European Association for Cardio-Thoracic Surgery (STAT) Congenital Heart Surgery Mortality Categories. This quality report summarizes contemporary outcomes, provides the odds ratios for the CHSD risk model variables based on this analysis, and describes on-going efforts to improve data collection and augment analytical approaches. Lastly, 5 research publications completed in the last year using data from the CHSD are also summarized.

Anatomic considerations in the management of complete atrioventricular canal

OBJECTIVE

Patients with complete atrioventricular canal have a variable clinical course prior to repair. Many patients balance their circulations well prior to elective repair. Others manifest clinically significant pulmonary over circulation early in life and require either palliative pulmonary artery banding or complete repair. The objective of this study was to assess anatomic features that impact the clinical course of patients.

METHODS

In total, 222 patients underwent complete atrioventricular canal repair between 2012 and 2022 at a single institution. Twenty-seven (12%) patients underwent either pulmonary artery banding (n = 15) or complete repair (n = 12) at less than 3 months of age (Group 1). The remaining 195 (88%) underwent repair after 3 months of age (Group 2). Patient records and imaging were reviewed.

RESULTS

The median post-operative length of stay following complete repair was 25 [7,46] days for those patients in Group 1 and 7 [5,12] days for those in Group 2 (p < 0.0001). There was relative hypoplasia of left-sided structures in Group 1 versus Group 2. Mean z-score for the ascending aorta was -1.2 (±0.8) versus -0.3 (±0.9) (p < 0.0001), the aortic isthmus was -2.1 (±0.8) versus -1.4 (±0.8) (p = 0.005). The pulmonary valve to aortic valve diameter ratio was median 1.47 [1.38,1.71] versus 1.38 [1.17,1.53] (p 0.008).

CONCLUSIONS

Echocardiographic evaluation of the systemic and pulmonary outflow of patients with complete atrioventricular canal may assist in predicting the clinical course and need for early repair vs pulmonary artery banding.

Predictors of Post-Operative Hospital Length of Stay Following Complete Repair of Tetralogy of Fallot in a Pediatric Cohort in the North of England

We sought to estimate the median post-operative length of stay (PLOS) and predictors of PLOS following tetralogy of Fallot (ToF) repair at a specialist surgical center in the North of England. The local National Congenital Heart Disease Audit dataset was used to identify patients aged < 2 years who underwent surgical repair for ToF between 1 January 1986 and 13 May 2022. Coefficients representing the median change in PLOS (days) according to predictors were estimated using Quantile regression. There were 224 patients (59.4% male, median age = 9 months, interquartile range (IQR) 5-13 months) with a median PLOS of 9 days (IQR 7-13). In the univariable regression, age (months) and weight (kg) at operation (β =  - 0.17, 95% CI: - 0.33, - 0.01) and (β =  - 0.53, 95% CI: - 0.97, - 0.10), previous (cardiac or thoracic) procedure (β = 5, 95% CI:2.38, 7.62), procedure urgency (elective vs urgent) (β = 2.8, 95% CI:0.39, 5.21), bypass time (mins) (β = 0.03, 95% CI:0.01, 0.05), cross-clamp time (mins) (β = 0.03, 95% CI:0.01, 0.06) and duration of post-operative intubation (days) (β = 0.81, 95% CI:0.67, 0.96), were significantly associated with PLOS. Previous procedure and intubation time remained significant in multivariable analyses. Some patient and operative factors can predict PLOS following complete ToF repair. Information on PLOS is important for health professionals to support parents in preparing for their child's discharge and to make any necessary practical arrangements. Health commissioners can draw on evidence-based guidance for resource planning. The small sample size may have reduced the power to detect small effect sizes, but this regional study serves as a foundation for a larger national study.

Intensive Care Unit Analgosedation After Cardiac Surgery in Children with Williams Syndrome : a Matched Case-Control Study

OBJECTIVE

Cardiovascular abnormalities are common in patients with Williams syndrome and frequently require surgical intervention necessitating analgesia and sedation in a population with a unique neuropsychiatric profile, potentially increasing the risk of adverse cardiac events during the perioperative period. Despite this risk, the overall postoperative analgosedative requirements in patients with WS in the cardiac intensive care unit have not yet been investigated. Our primary aim was to examine the analgosedative requirement in patients with WS after cardiac surgery compared to a control group. Our secondary aim was to compare the frequency of major ACE and mortality between the two groups.

DESIGN

Matched case-control study.

SETTING

Pediatric CICU at a Tertiary Children's Hospital.

PATIENTS

Patients with WS and age-matched controls who underwent cardiac surgery and were admitted to the CICU after cardiac surgery between July 2014 and January 2021.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

Postoperative outcomes and total doses of analgosedative medications were collected in the first six days after surgery for the study groups. Median age was 29.8 (12.4-70.8) months for WS and 23.5 (11.2-42.3) months for controls. Across all study intervals (48 h and first 6 postoperative days), there were no differences between groups in total doses of morphine equivalents (5.0 mg/kg vs 5.6 mg/kg, p = 0.7 and 8.2 mg/kg vs 10.0 mg/kg, p = 0.7), midazolam equivalents (1.8 mg/kg vs 1.5 mg/kg, p = 0.4 and 3.4 mg/kg vs 3.8 mg/kg, p = 0.4), or dexmedetomidine (20.5 mcg/kg vs 24.4 mcg/kg, p = 0.5 and 42.3 mcg/kg vs 39.1 mcg/kg, p = 0.3). There was no difference in frequency of major ACE or mortality.

CONCLUSIONS

Patients with WS received similar analgosedative medication doses compared with controls. There was no significant difference in the frequency of major ACE (including cardiac arrest, extracorporeal membrane oxygenation, and surgical re-intervention) or mortality between the two groups, though these findings must be interpreted with caution. Further investigation is necessary to elucidate the adequacy of pain/sedation control, factors that might affect analgosedative needs in this unique population, and the impact on clinical outcomes.

Survival After Single-Stage Repair of Truncus Arteriosus and Associated Defects

BACKGROUND

The goal of this study was to describe in-hospital and long-term mortality after single-stage repair of truncus arteriosus communis (TAC) and explore factors associated with these outcomes.

METHODS

This was a cohort study of consecutive patients undergoing single-stage TAC repair between 1982 and 2011 reported to the Pediatric Cardiac Care Consortium registry. In-hospital mortality was obtained for the entire cohort from registry records. Long-term mortality was obtained for patients with available identifiers by matching with the National Death Index through 2020. Kaplan-Meier survival estimates were created for up to 30 years after discharge. Cox regression models estimated hazard ratios for the associations with potential risk factors.

RESULTS

A total of 647 patients (51% male) underwent single-stage TAC repair at a median age of 18 days; 53% had type I TAC, 13% had interrupted aortic arch, and 10% underwent concomitant truncal valve surgery. Of these, 486 (75%) patients survived to hospital discharge. After discharge, 215 patients had identifiers for tracking long-term outcomes; 30-year survival was 78%. Concomitant truncal valve surgery at the index procedure was associated with increased in-hospital and 30-year mortality. Concomitant interrupted aortic arch repair was not associated with increased in-hospital or 30-year mortality.

CONCLUSIONS

Concomitant truncal valve surgery but not interrupted aortic arch was associated with higher in-hospital and long-term mortality. Careful consideration of the need and timing for truncal valve intervention may improve TAC outcomes.

Congenital Heart Surgery Machine Learning-Derived In-Depth Benchmarking Tool

BACKGROUND

We previously showed that machine learning-based methodologies of optimal classification trees (OCTs) can accurately predict risk after congenital heart surgery and assess case-mix-adjusted performance after benchmark procedures. We extend this methodology to provide interpretable, easily accessible, and actionable hospital performance analysis across all procedures.

METHODS

The European Congenital Heart Surgeons Association Congenital Cardiac Database data subset of 172,888 congenital cardiac surgical procedures performed in European centers between 1989 and 2022 was analyzed. OCT models (decision trees) were built predicting hospital mortality (area under the curve [AUC], 0.866), prolonged postoperative mechanical ventilatory support time (AUC, 0.851), or hospital length of stay (AUC, 0.818), thereby establishing case-adjusted benchmarking standards reflecting the overall performance of all participating hospitals, designated as the "virtual hospital." OCT analysis of virtual hospital aggregate data yielded predicted expected outcomes (both aggregate and for risk-matched patient cohorts) for the individual hospital's own specific case-mix, readily available on-line.

RESULTS

Raw average rates were hospital mortality, 4.9%; mechanical ventilatory support time, 14.5%; and length of stay, 15.0%. Of 146 participating centers, compared with each hospital's overall case-adjusted predicted hospital mortality benchmark, 20.5% statistically (<90% CI) overperformed and 20.5% underperformed. An interactive tool based on the OCT analysis automatically reveals 14 hospital-specific patient cohorts, simultaneously assessing overperformance or underperformance, and enabling further analysis of cohort strata in any chosen time frame.

CONCLUSIONS

Machine learning-based OCT benchmarking analysis provides automatic assessment of hospital-specific case-adjusted performance after congenital heart surgery, not only overall but importantly, also by similar risk patient cohorts. This is a tool for hospital self-assessment, particularly facilitated by the user-accessible online-platform.

Pediatric Anesthesia in the Community

Pediatric anesthesia is a diverse subspecialty practiced at thousands of hospitals and ambulatory surgery centers across the country. Most unusual and high-risk cases are performed in dedicated children's hospitals. However, the majority of cases and practitioners are based in the community. We present a review of demographics in pediatric anesthesia in the United States across 7 years of data from US Anesthesia Partners, a national anesthesia practice, which covers the full range of hospitals and outpatient facilities.

Temporal Cluster Analysis of Deep Sternal Wound Infection in a Regional Quality Collaborative

INTRODUCTION

Deep sternal wound infection (DSWI) is a rare complication associated with high mortality. Seasonal variability in surgical site infections has been demonstrated, however, these patterns have not been applied to DSWI. The purpose of this study was to assess temporal clustering of DSWIs.

METHODS

All cardiac surgery patients who underwent sternotomy were queried from a regional Society of Thoracic Surgeons database from 17 centers from 2001 to 2019. All patients with the diagnosis of DSWI were then identified. Cluster analysis was performed at varying time intervals (monthly, quarterly, and yearly) at the hospital and regional level. DSWI rates were calculated by year and month, and compared using mixed-effects negative binomial regression.

RESULTS

A total of 134,959 patients underwent a sternotomy for cardiac surgery, of whom 469 (0.35%) developed a DSWI. Rates of DSWI per hospital across all years ranged from 0.12% to 0.69%. Collaborative-level rates of DSWIs were the greatest in September (0.44%) and the lowest in January (0.30%). Temporal clustering was not seen across seasonal quarters (high rate in preceeding quarter was not associated with a high rate in the next quarter) (P = 0.39). There were yearly differences across all institutions in the DSWI rates. A downward trend in DSWI rates was seen from 2001 to 2019 (P < 0.001). A difference among hospitals in the cohort was observed (P < 0.001).

CONCLUSIONS

DSWI are a rare event within our region. Unlike other surgical site infection, there does not appear to be a seasonal pattern associated with DSWI.

Perioperative Brain Injury in Children with Aortic Arch Anomalies: A Retrospective Study of Risk Factors and Outcomes

Complex pediatric cardiac disease is associated with brain impairment and neurodevelopmental disorders, particularly in patients requiring cardiac surgery for aortic arch anomalies. This study examines the incidence, risk factors, and outcomes of perioperative brain injury in children undergoing aortic arch repair who had aortic arch anomalies. A total of 145 children with aortic arch anomalies in our center undergoing aortic arch repair between January 2014 and December 2022 were enrolled. There were 129 (89.0%) with coarctation of the aorta (COA) and 16 (9.7%) with interrupted aortic arch (IAA). Risk factor analysis of brain injuries was done using perioperative imaging and included symptoms of hemorrhagic stroke, arterial ischemic stroke, white matter injury, cerebral sinus venous thrombosis, and other pathologies. Preoperatively, 50/145 (34.5%) patients had brain injuries. Multivariate analysis showed that an increased risk of hemorrhagic stroke was associated with newborns (odds ratio [OR], 2.09 [95% CI 0.08-3.50]), isolated COA (OR, 3.69 [95% CI 1.23-7.07]), mechanical ventilation (MV) ([OR, 2.56 [95% CI 1.25-4.03]), and sepsis (OR, 1.73 [95% CI 0.46-3.22]). Newborns ([OR, 1.91 [95% Cl 0.58-3.29]) and weight-for-age z score ([OR, -0.45 [95% CI -0.88 to -0.1]) were associated with an increased risk of white matter injury. New postoperative brain injuries were present in 12.9% of the patients (16/124). Deep hypothermic circulatory arrest (DHCA) was associated with new postoperative brain injuries compared with deep hypothermic low-flow (DHLF) plus antegrade cerebral perfusion (ACP) (([OR, 2.67 [95% CI, 0.58-5.75])). Isolated COA was almost associated with new postoperative brain injuries (OR, 1.13 [95% CI, -0.04 to 2.32]). Children diagnosed with isolated COA appeared to have a higher risk of perioperative brain injury, but the underlying mechanisms are still unclear. We focused on the intrinsic mechanism by which changes in hemodynamics caused by COA result in perioperative brain injury. Further research will be needed to optimize the personalized treatment and cerebral perfusion techniques for complex pediatric cardiac surgery.