Outcomes and Complications After Elective Thoracic and Lumbar Spinal Fusion in Elderly Patients: A Comparison of Methods to Predict Adverse Events

Abstract

Study Design: Retrospective study. Objectives: The current study compares the ability of the modified Frailty Index (mFI), the American Society of Anesthesiologists (ASA) classification, the modified Charleston Comorbidity Index (mCCI), the American College of Surgeons Surgical Risk Calculator (SRC), and the Fusion Risk Score (FRS) to predict perioperative outcomes. Methods: Comorbidity indices were calculated for patients undergoing elective thoracic and lumbar spinal fusion at a single institution and assessed for their discriminative ability in predicting the desired outcomes using an area under the curve (AUC) analysis. Results: 393 patients met the inclusion and exclusion criteria. Patients being treated for adult spinal deformity (ASD) had the highest rate of complications (44.4%). The FRS had acceptable discrimination (AUC >0.7) and the highest ability among the methods studied to predict any adverse effects, new neurological deficit, return to OR within 90 days, and surgical site infection. It had good discrimination ability (AUC >0.8) predicting durotomy, respiratory failure (RF) requiring intubation, hemodynamic instability, and sepsis. The SRC had acceptable discrimination and highest ability to predict deep venous thrombosis (DVT). The mCCI had excellent and the highest ability to predict acute renal failure (ARF). For the other outcomes, the indices had either poor predictive ability (AUC 0.6-0.7) or no discriminative ability (AUC <0.6). Conclusions: The FRS had a better ability than the ASA, mCCI, mFI, and SRC to predict the most perioperative adverse events and reoperation. Further study is needed to develop preoperative indices with better predictive ability of postoperative outcomes.