Predictive Value of Elevated BMI for Prolonged Hospital Stay After Multilevel Scoliosis Surgery

Introduction: Obesity is a recognized risk factor in spinal surgery, with increased body mass index (BMI) associated with higher rates of postoperative complications and extended hospital stays. This study investigates the relationship between BMI and length of stay (LOS) following multilevel scoliosis surgery, aiming to identify patients at risk for prolonged hospitalization based on BMI classification. Understanding this relationship can enhance perioperative planning and resource allocation in neurosurgical practices.

Methods: Using the Nationwide Readmissions Database (NRD), we identified 8,122 patients who underwent multilevel scoliosis surgery. Patients were binned into three BMI groups: BMI 30-35 (n=424, LOS 8d), BMI 35-40 (n=289, LOS 8d), and BMI 40+ (n=228, LOS 8d). A mixed-effects logistic regression model was employed to predict LOS in the top quartile. Fixed effects included the Elixhauser Comorbidity Index, age, sex, insurance status, NRD discharge weighting, and median income by ZIP code, while random effects included hospital characteristics. The model’s performance was assessed via receiver operating characteristic (ROC) curves, and area under the curve (AUC) values calculated for each BMI group.

Results: ROC curve analysis demonstrated high discriminative ability of the model across all BMI groups, with AUC values of 0.868 for BMI 30-35, 0.857 for BMI 35-40, and 0.857 for BMI 40+. These results indicate that the model reliably predicts extended LOS in the top quartile for all BMI categories. Although patients with a BMI of 30-35 exhibited the highest AUC, the predictive accuracy for the BMI 35-40 and BMI 40+ groups were comparable (p>0.05).

Conclusion : While patients with a higher BMI are often thought to be at greater risk for extended hospitalization, the model’s consistent performance across BMI groups suggests that other factors, such as comorbidities and hospital characteristics, may also play significant roles. These findings underscore the need for perioperative planning that considers both BMI and other patient- and hospital-specific factors.