Adjusting for Congenital Heart Surgery Risk Using Administrative Data.

BACKGROUND: Congenital heart surgery (CHS) encompasses a heterogeneous population of patients and surgeries. Risk standardization models that adjust for patient and procedural characteristics can allow for collective study of these disparate patients and procedures. OBJECTIVES: We sought to develop a risk-adjustment model for CHS using the newly developed Risk Stratification for Congenital Heart Surgery for ICD-10 Administrative Data (RACHS-2) methodology. METHODS: Within the Kids' Inpatient Database 2019, we identified all CHSs that could be assigned a RACHS-2 score. Hierarchical logistic regression (clustered on hospital) was used to identify patient and procedural characteristics associated with in-hospital mortality. Model validation was performed using data from 24 State Inpatient Databases during 2017. RESULTS: Of 5,902,538 total weighted hospital discharges in the Kids' Inpatient Database 2019, 22,310 pediatric cardiac surgeries were identified and assigned a RACHS-2 score. In-hospital mortality occurred in 543 (2.4%) of cases. Using only RACHS-2, the mortality mode had a C-statistic of 0.81 that improved to 0.83 with the addition of age. A final multivariable model inclusive of RACHS-2, age, payer, and presence of a complex chronic condition outside of congenital heart disease further improved model discrimination to 0.87 (P < 0.001). Discrimination in the validation cohort was also very good with a C-statistic of 0.83. CONCLUSIONS: We created and validated a risk-adjustment model for CHS that accounts for patient and procedural characteristics associated with in-hospital mortality available in administrative data, including the newly developed RACHS-2. Our risk model will be critical for use in health services research and quality improvement initiatives.

Machine Learning to Predict Interstage Mortality Following Single Ventricle Palliation: A NPC-QIC Database Analysis.

There is high risk of mortality between stage I and stage II palliation of single ventricle heart disease. This study aimed to leverage advanced machine learning algorithms to optimize risk-prediction models and identify features most predictive of interstage mortality. This study utilized retrospective data from the National Pediatric Cardiology Quality Improvement Collaborative and included all patients who underwent stage I palliation and survived to hospital discharge (2008-2019). Multiple machine learning models were evaluated, including logistic regression, random forest, gradient boosting trees, extreme gradient boost trees, and light gradient boosting machines. A total of 3267 patients were included with 208 (6.4%) interstage deaths. Machine learning models were trained on 180 clinical features. Digoxin use at discharge was the most influential factor resulting in a lower risk of interstage mortality (p < 0.0001). Stage I surgery with Blalock-Taussig-Thomas shunt portended higher risk than Sano conduit (7.8% vs 4.4%, p = 0.0002). Non-modifiable risk factors identified with increased risk of interstage mortality included female sex, lower gestational age, and lower birth weight. Post-operative risk factors included the requirement of unplanned catheterization and more severe atrioventricular valve insufficiency at discharge. Light gradient boosting machines demonstrated the best performance with an area under the receiver operative characteristic curve of 0.642. Advanced machine learning algorithms highlight a number of modifiable and non-modifiable risk factors for interstage mortality following stage I palliation. However, model performance remains modest, suggesting the presence of unmeasured confounders that contribute to interstage risk.

Risk Stratification for Congenital Heart Surgery for ICD-10 Administrative Data (RACHS-2).

BACKGROUND: As the cardiac community strives to improve outcomes, accurate methods of risk stratification are imperative. Since adoption of International Classification of Disease-10th Revision (ICD-10) in 2015, there is no published method for congenital heart surgery risk stratification for administrative data. OBJECTIVES: This study sought to develop an empirically derived, publicly available Risk Stratification for Congenital Heart Surgery (RACHS-2) tool for ICD-10 administrative data. METHODS: The RACHS-2 stratification system was iteratively and empirically refined in a training dataset of Pediatric Health Information Systems claims to optimize sensitivity and specificity compared with corresponding locally held Society of Thoracic Surgeons-Congenital Heart Surgery (STS-CHS) clinical registry data. The tool was validated in a second administrative data source: New York State Medicaid claims. Logistic regression was used to compare the ability of RACHS-2 in administrative data to predict operative mortality vs STAT Mortality Categories in registry data. RESULTS: The RACHS-2 system captured 99.6% of total congenital heart surgery registry cases, with 1.0% false positives. RACHS-2 predicted operative mortality in both training and validation administrative datasets similarly to STAT Mortality Categories in registry data. C-statistics for models for operative mortality in training and validation administrative datasets-adjusted for RACHS-2-were 0.76 and 0.84 (95% CI: 0.72-0.80 and 0.80-0.89); C-statistics for models for operative mortality-adjusted for STAT Mortality Categories-in corresponding clinical registry data were 0.75 and 0.84 (95% CI: 0.71-0.79 and 0.79-0.89). CONCLUSIONS: RACHS-2 is a risk stratification system for pediatric cardiac surgery for ICD-10 administrative data, validated in 2 administrative-registry-linked datasets. Statistical code is publicly available upon request.

Predictors and outcomes of heart block during surgical stage I palliation of patients with a single ventricle: A report from the NPC-QIC.

BACKGROUND: Mortality in cohorts with a single ventricle remains high with multiple associated factors. The effect of heart block during stage I palliation remains unclear. OBJECTIVE: The purpose of this study was to study patient and surgical risks of heart block and its effect on 12-month transplant-free survival in patients with a single ventricle. METHODS: Patient, surgical, outcome data and heart block status (transient and permanent) were obtained from the National Pediatric Cardiology Quality Improvement Collaborative single ventricle database. Bivariate analysis was performed comparing patients with and without heart block, and multivariate modeling was used to identify variables associated with block. One-year outcomes were analyzed to identify variables associated with lower 12-month transplant-free survival. RESULTS: In total, 1423 patients were identified, of whom 28 (2%) developed heart block (second degree or complete) during their surgical admission. Associated risk factors for block included heterotaxy syndrome (odds ratio [OR] 6.4) and atrial flutter/fibrillation (OR 3.8). Patients with heart block had lower 12-month survival, though only in patients with complete heart block as opposed to second degree block. At 12 months of age, 43% (12/28) of patients with heart block died and were more likely to experience mortality at 12 months than patients without block (OR 4.9; 95% confidence interval 1.4-17.5; P = .01). CONCLUSION: Although rare, complete heart block after stage I palliation represents an additional risk of poor outcomes in this high-risk patient population. Heterotaxy syndrome was the most significant risk factor for the development of heart block after stage I palliation. The role of transient block in outcomes and potential rescue with long-term pacing remains unknown and requires additional study.

Center Reputation and Experience Relate to Industry Payments to Congenital Heart Disease Providers.

BACKGROUND: The Physician Payments Sunshine Act was enacted to understand financial relationships with industry that might influence provider decisions. We investigated how industry payments within the congenital heart community relate to experience and reputation. METHODS: Congenital cardiothoracic surgeons and pediatric cardiologists were identified from the Open Payments Database. All payments from 2013 through 2017 were matched to affiliated hospitals' U.S. News & World Report (USNWR) rankings, The Society of Thoracic Surgeons-Congenital Heart Surgery Public Reporting Star Ratings, and Optum Center of Excellence (COE) designation. Surgeon payments were linked to years since terminal training. Univariable analyses were conducted. RESULTS: The median payment amount per surgeon ($71; interquartile range [IQR], $41-$99) was nearly double the median payment amount per cardiologist ($41; IQR, $18-$84; P < .05). For surgeons, median individual payment was 56% higher to payees at USNWR top 10 children's hospitals ($100; IQR, $28-$203) vs all others ($64; IQR, $23-$140; P < .001). For cardiologists, median individual payment was 26% higher to payees at USNWR top 10 children's hospitals ($73; IQR, $28-$197) vs all others ($58; IQR, $19-$140; P < .001). Findings were similar across The Society of Thoracic Surgeons-Congenital Heart Surgery star rankings and Optum Center of Excellence groups. By surgeon experience, surgeons 0 to 6 years posttraining (first quartile) received the highest number of median payments per surgeon (17 payments; IQR, 6.5-28 payments; P < .001). Surgeons 21 to 44 years posttraining (fourth quartile) received the lowest median individual payment ($51; IQR, $20-132; P < .001). CONCLUSIONS: Industry payments vary by hospital reputation and provider experience. Such biases must be understood for self-governance and the delineation of conflict of interest policies that balance industry relationships with clinical innovation.

Frequency of medical emergency team activation prior to pediatric cardiopulmonary resuscitation.

BACKGROUND: Medical Emergency Teams (METs) are designed to respond to signs of clinical decline in order to prevent cardiopulmonary arrest and reduce mortality. The frequency of MET activation prior to pediatric cardiopulmonary resuscitation (CPR) is unknown. METHODS: Within the Get With The Guidelines-Resuscitation Registry (GWTG-R), we identified children with bradycardia or cardiac arrest requiring CPR on the general inpatient or telemetry floors from 2007 to 2013. We examined the frequency with which CPR outside the ICU was preceded by a MET evaluation. In cases where MET evaluation did not occur, we examined the frequency of severely abnormal vital signs at least 1hour prior to CPR that could have prompted a MET evaluation but did not. RESULTS: Of 215 children from 23 hospitals requiring CPR, 48 (22.3%) had a preceding MET evaluation. Children with MET evaluation prior to CPR were older (6.8±6.5 vs. 3.1±4.7 years of age, p<0.001) and were more likely to have metabolic/electrolyte abnormalities (18.8% vs. 5.4%, p=0.006), sepsis (16.7% vs. 4.8%, p=0.01), or malignancy (22.9% vs. 5.4%, p<0.001). Among patients who did not have a MET called and with information on vital signs, 55/141 (39.0%) had at least one abnormal vital sign that could have triggered a MET. CONCLUSION: The majority of pediatric patients requiring CPR for bradycardia or cardiac arrest do not have a preceding MET evaluation despite a significant number meeting criteria that could have triggered the MET. This suggests opportunities to more efficiently use MET teams in routine care.

Characteristics and safety of interventions and procedures performed during catheterisation of patients with congenital heart disease: early report from the national cardiovascular data registry.

OBJECTIVES: The objective of this study was to report procedural characteristics and adverse events on the data collected in the IMproving Paediatric and Adult Congenital Treatment registry. BACKGROUND: The IMproving Paediatric and Adult Congenital Treatment- registry is a catheterisation registry focussed on paediatric and adult patients with congenital heart disease who are undergoing diagnostic catheterisations and catheter-based interventions. This study reports procedural characteristics and adverse events of patients who have undergone selected catheterisation procedures from January, 2011 to June, 2013. METHODS: Demographic, clinical, procedural, and institutional data elements were collected at participating centres and entered via either a web-based platform or software provided by the American College of Cardiology-certified vendors, and were collected in a secure, centralised database. For the purpose of this study, procedures that were not classified as one of the 'core' IMproving Paediatric and Adult Congenital Treatment procedures originally chosen for additional data collection were identified and selected for further data analysis. RESULTS: During the time frame of data collection, a total of 8021 cases were classified as other procedures and/or multiple procedures. The most commonly performed case types - isolated or in combination with other procedures - were right ventricular biopsy in 3433 (42.8%), conduit/MPA interventions in 979 (12.3%), and systemic pulmonary artery collateral occlusion in 601 (7.5%). For the whole cohort, adverse events of any severity occurred in 957 (12.0%) cases, whereas major adverse events occurred in 113 (1.4%) cases; six patients (0.1%) died in the catheterisation laboratory. CONCLUSIONS: The IMproving Paediatric and Adult Congenital Treatment registry has provided important data on the frequency and spectrum of cardiac catheterisation procedures performed in the present era. For many procedures, more data and work are needed to identify more subtle differences between case categories, especially as it relates to the incidence of major adverse events, and to further develop a risk-adjustment methodology to allow equitable comparisons among institutions.