Wednesday, September 26, 2018 7:35 AM–9:00 AM ePosters: P117. Cerebral vascular attacks in cervical spinal fusion

Abstract

BACKGROUND CONTEXT Surgical fusion of the cervical spine has become an increasingly popular treatment modality to addresscommon pathologies of the spine. Furthermore, the past two decades have seen a significant change in theinstrumentation and surgical approach decisions of the spine surgeon that has improved fusion rates and surgicaloutcomes. The advancement of cervical fusion surgery, combined with an increasingly aging population, has led to asubstantial surge in the number of procedures performed. The anterior cervical fusion (ACF), posterior cervical fusion (PCF) and fusion specific to the atlas-axis (AASF) have been shown to present their own specific technical challengesand associated risks, including the catastrophic cerebral vascular attack (CVA). The National Inpatient Sample (NIS) serves as the single largest all-payer inpatient database in the United States and provides weights for nationallyrepresentative estimates. For this reason it serves as an appropriate approximation of the overall perioperative morbidityand mortality statistics for this and other procedures. PURPOSE Previous NIS studies have described perioperative morbidity and mortality in spinal fusions but none of have examined the incidence and independent risk factors for CVA in cervical spine fusions with the most recent relevant data presented in this study of 2005-2013. STUDY DESIGN/SETTING Database study. PATIENT SAMPLE A total of 294,575 patients undergoing elective cervical fusion surgery from 2005 to 2013. OUTCOME MEASURES (1) Outcomes were assessed by frequencies of CVA, (2) various predictors of CVA and, (3) demographic variables associated with each cervical approach METHODS The NIS database was reviewed from 2005 to 2013. Patients undergoing elective cervical fusion surgery were identified by their discharge ICD-9-CM code, and with indications specific to the cervical spine. Patients were diveded into cohorts based on surigcal approach. Demographics of the study population were assessed including age, gender, hospital region and teaching status, insurance status, race or ethnicity, median household income and Charlson Comorbidity Index (CCI) score. Incidence of CVA was identified utilizing appropriate ICD-9 CM codes (997.02 or 434.91).Other complications related to the procedure including pulmonary embolism (PE), deep venous thrombus (DVT), infection, cardiac, hematoma, durotomy, dysphagia and mortality were identified. Various predictors of CVA were examined as well. Statistical analysis involved T tests, χ 2 analysis, and binary logistic regression with P RESULTS We identified 294,575 patients which represented an estimated 1,450,248 weighted patients hospitalized forprimary cervical spine fusion. There were 423 CVAs (1.4 per 1000).Patients had a mean age of 53.6 with 50.8% being female. The ACF cohort was significantly younger (52.9), predominantly white (81.3%), privately insured or self-pay (56.1%, 10.6%), preformed in the South region (43.8%) at teaching hospitals (50.5%), at private, for-profit hospitals (20.4%) with a smaller CCI score (0.45) than the other 2 cohorts. The PCF cohort had significantly more men (58.2%) and black individuals (13.2%) than theother cohorts. The AASF cohort used significantly more BMP (20.2%) than the other two cohorts. The AASF cohort had significantly more DVTs, Infections, cardiac and neurologic complications, incidences of dysphagia and overall mortality. The overall charges were a mean 51,539 (max 3,991,172) with a mean length of stay of 2.9days (max 336 days). Independent risk factors for CVA were age over 61, male sex, black race, insured by Medicare, location at a teaching hospitals, medianhousehold income of $39,000–$47,999, diabetes, diabetes with complications, metastatic disease, solid tumors and morethan 2 chronic diseases and increased CCI score (mean 2.65). These patients experienced an increased length of stay (18.91 days) and increased total charges ($233,154). CONCLUSIONS The incidence of CVA in cervical spine surgery (1.4 per 1000) is slightly higher than the published incidence of CVA in lumbar spine fusions (1.3 per 1000). Our study demonstrates that cervical spinal fusions are becoming increasingly common nationally in teaching and nonteaching hospitals with low rates of complications and overall mortality. We expect, with the advancements in instrumentation, success rate of fusions and an aging population that this specific procedure will continue to see an increase in utilization by the spinal surgeon across the country. The presence of CVAs in these patients warrants thorough investigation into the risk factors associated to prevent catastrophic consequences and undue cost upon the health-care system. Our study found that patients with more than 2 complications and those with a high CCI score demonstrated the most significant risk factors for CVA. Furthermore, future studies should utilize the improved data collection capabilities of the recently implemented ICD-10 codes to paint a clearer picture of this common procedure.