Letter re: Incremental Increase in Hospital Length of Stay due to Complications of Surgery for Adult Spinal Deformity

Dear Editor,

We recently read the article “Incremental Increase in Hospital Length of Stay Due to Complications of Surgery for Adult Spinal Deformity” by Lafage et al. ¹ The authors’ efforts to quantify the impact of surgical complications on hospital length of stay are commendable, and their findings provide valuable insights into postoperative outcomes. While we agree with these findings, several additional factors may influence patient outcomes, which could be considered in future research.

Firstly, the study does not describe preoperative psychological factors such as depression, anxiety, or coping mechanisms, all of which can impact recovery and, consequently, hospital length of stay. Previous research by Park et al. has shown that patients with elevated anxiety or depression levels face higher risks of complications and extended recovery periods, resulting in prolonged hospital stays. ² Incorporating psychological factors through multivariable analyses or stratification by mental health status could help clarify these influences.

Secondly, the study does not discuss socioeconomic factors, such as income level, education, occupation, and insurance coverage. Holbert et al. have demonstrated that patients with lower socioeconomic status often present later to health care, leading to advanced deformities and complications that extend hospital stays. ³ Including these variables could allow future analyses to identify and address disparities in patient outcomes.

Additionally, diabetes is a known risk factor for infections and complications following spinal surgeries. However, this study does not specify whether diabetic patients were included or excluded. As diabetes is associated with an increased risk of wound infection and delayed healing, it would be beneficial to consider this factor, possibly by analyzing outcomes for diabetic vs non-diabetic patients. ⁴

Further, surgeon experience and procedure volume significantly impact patient recovery. Studies have found that hospitals with higher volumes of specialized surgeries report fewer complications, leading to shorter hospital stays. ⁵ Addressing these variables in future work could help illuminate differences in recovery outcomes across institutions or surgeons.

Using a neural network model for predicting hospital stay length is intriguing but would benefit from more transparency regarding how outliers and data variability were managed. Although machine learning models are decisive for handling complex datasets, they often function as “black boxes,” limiting interpretability in clinical settings. We recommend supplementing the neural network model with more interpretable statistical approaches, such as multivariable regression or decision trees, to enhance clinical understanding. Clarifying how outliers were handled would improve the model’s transparency. ⁶