Perioperative Complications and Mortality After Spinal Fusions

STUDY DESIGN: Retrospective review. OBJECTIVE: To analyze the trends in complications and mortality after spinal fusions. SUMMARY OF BACKGROUND DATA: Utilization of spinal fusions has been increasing during the past decade. It is essential to evaluate surgical outcomes to better identify patients who benefit most from surgical intervention. Integration of empiric evidence from large administrative databases into clinical decision making is instrumental in providing higher-quality, evidence-based, patient-centered care. METHODS: This study used Nationwide Inpatient Sample data from 2001 through 2010. Patients who underwent spinal fusions were identified using the CCS (Clinical Classifications Software) and ICD-9 (International Classification of Diseases, 9th Revision) codes. Data on patient comorbidities, primary diagnosis, and postoperative complications were obtained via ICD-9 diagnosis codes and via CCS categories. National estimates were calculated using weights provided as part of the database. Time trend analysis for average length of stay, total charges, mortality, and comorbidity burden was performed. Univariate and multivariate models were constructed to identify predictors of mortality and postoperative complications. RESULTS: An estimated 3,552,873 spinal fusions were performed in the United States between 2001 and 2010. The national bill for spinal fusions increased from $10 billion to $46.8 billion. Today, patients are older and have a greater comorbidity burden than 10 years ago. Mortality remained relatively constant at 0.46%, 1.2%, and 0.14% for cervical, thoracic, and lumbar fusions, respectively. Morbidity rates showed an increasing trend at all levels. Multivariate analysis of 19 procedures and patient-related risk factors and 9 perioperative complications identified 85 statistically significant (P< 0.01) interactions. CONCLUSION: The data on perioperative risks and risk factors for postoperative complications of spinal fusions presented in this study is pivotal to appropriate surgical patient selection and well-informed risk-benefit evaluation of surgical intervention. LEVEL OF EVIDENCE: N/A.