Frailty Status Enhances Prediction of Postoperative Anemia in Scoliosis Surgery

Introduction: Frailty is an increasingly recognized risk factor for adverse outcomes in surgical patients, especially in complex procedures such as multilevel scoliosis surgery. This study aims to determine whether frailty, as defined by the Johns Hopkins Adjusted Clinical Groups (JHACG) frailty-defining diagnosis indicator, improves the prediction of acute posthemorrhagic anemia (blood loss anemia) in this patient population. Identifying patients at heightened risk can help guide perioperative management and resource allocation.

Methods: We analyzed data from the Nationwide Readmissions Database (NRD), including 8,123 patients who underwent multilevel scoliosis surgery. Patients were classified as frail (n=859) or non-frail (n=7,264) based on the JHACG frailty-defining diagnosis indicator, which employs ten ICD-10 code categories (e.g., malnutrition, dementia, incontinence, difficulty walking). Two mixed-effects logistic regression models were developed to predict the occurrence of acute posthemorrhagic anemia: one incorporating frailty status and the other agnostic to it. Fixed effects included age, sex, insurance status, NRD discharge weighting, and median income by ZIP code, while random effects accounted for hospital characteristics.

Results: The model agnostic to frailty status produced an AUC of 0.639, indicating moderate predictive ability for acute posthemorrhagic anemia. Including frailty status increased the AUC to 0.682, demonstrating improved predictive performance. The p-value of 0.01785 indicates that the inclusion of frailty status provides statistically significant predictive value, suggesting that frail patients are at higher risk of developing anemia after surgery.

Conclusion : This study highlights the importance of frailty as a predictor of acute posthemorrhagic anemia following multilevel scoliosis surgery. While both models demonstrated moderate predictive capability, the inclusion of frailty status significantly improved model performance. These findings underscore the value of incorporating frailty assessments in preoperative planning to better anticipate complications and optimize patient outcomes.