Article Impact Level: HIGH Data Quality: STRONG Summary of Journal of the American College of Cardiology, 82(23), 2212–2221. https://doi.org/10.1016/j.jacc.2023.09.826 Dr. Natalie Jayaram et al.
Points
- Using the RACHS-2 methodology and administrative data from the Kids’ Inpatient Database 2019, we developed and validated a risk-adjustment model for congenital heart surgery (CHS).
- They identified 22,310 pediatric cardiac surgeries with assigned RACHS-2 scores out of 5,902,538 weighted hospital discharges, with an in-hospital mortality rate of 2.4%.
- The mortality model based solely on RACHS-2 demonstrated a C-statistic of 0.81, which improved to 0.83 with the addition of age and further enhanced to 0.87 with the inclusion of age, payer, and the presence of a complex chronic condition outside of congenital heart disease.
- The final multivariable model exhibited high discriminatory power, with a C-statistic of 0.87, and validation in a separate cohort yielded a C-statistic of 0.83, indicating the model’s effectiveness in risk stratification for CHS patients.
- The risk model’s comprehensive incorporation of patient and procedural characteristics offers a valuable tool for health services research and quality improvement initiatives, addressing the need for collective study of diverse CHS patients and procedures.
Summary
The study aimed to develop a risk-adjustment model for congenital heart surgery (CHS) utilizing the Risk Stratification for Congenital Heart Surgery for ICD-10 Administrative Data (RACHS-2) methodology. Analyzing the Kids’ Inpatient Database 2019, the researchers identified 22,310 pediatric cardiac surgeries with assigned RACHS-2 scores among 5,902,538 weighted hospital discharges. The in-hospital mortality rate was 2.4%, and the mortality model based solely on RACHS-2 demonstrated a C-statistic of 0.81, which improved to 0.83 with the addition of age. The final multivariable model, incorporating RACHS-2, age, payer, and the presence of a complex chronic condition outside of congenital heart disease, further enhanced discrimination to 0.87, with validation in a separate cohort yielding a C-statistic of 0.83.
The research successfully established and validated a comprehensive risk-adjustment model for CHS, accounting for patient and procedural characteristics associated with in-hospital mortality, including the newly developed RACHS-2. The model’s high discriminatory power, as evidenced by the C-statistic of 0.87, underscores its effectiveness in risk stratification for CHS patients. Moreover, incorporating age, payer, and complex chronic conditions outside of congenital heart disease further improved the model’s predictive accuracy. The findings highlight the critical utility of the risk model in health services research and quality improvement initiatives, offering a valuable tool for the collective study of diverse CHS patients and procedures.
In summary, the study’s development and validation of a robust risk-adjustment model for CHS, leveraging the RACHS-2 methodology and administrative data, represent a significant advancement in risk standardization for congenital heart surgeries. The model’s comprehensive incorporation of patient and procedural characteristics enhances its applicability in health services research and quality improvement efforts. It provides a valuable resource for studying and improving outcomes in the heterogeneous population of CHS patients.
Link to the article: https://www.sciencedirect.com/science/article/abs/pii/S0735109723077021
References Jayaram, N., Allen, P., Hall, M., Karamlou, T., Woo, J., Crook, S., & Anderson, B. R. (2023). Adjusting for Congenital Heart Surgery Risk Using Administrative Data. Journal of the American College of Cardiology, 82(23), 2212–2221. https://doi.org/10.1016/j.jacc.2023.09.826
