Wednesday, September 26, 2018 10:35 AM – 12:00 PM Cervical Spine Trauma: 20. Traumatic cervical spinal cord injuries with fracture: an investigation of early versus delayed surgery among 6,636 propensity-matched patients

Abstract

BACKGROUND CONTEXT Currently, optimal timing of surgical intervention for traumatic cervical spinal cord injuries (SCIs), a life-altering event and enormous economic burden, is unclear. PURPOSE Our goal was to investigate the perioperative and postoperative outcomes in patients with traumatic SCIs. STUDY DESIGN/SETTING Retrospective review. PATIENT SAMPLE A total of 28,414 SCI patients. OUTCOME MEASURES Perioperative complications, mortality, length of stay, total hospital charges, discharge disposition. METHODS HCUP's NIS was queried for ICD-9 codes pertaining to cervical SCI w/ fracture (806-806.19). Patients without traumatic etiology, neuromuscular conditions (eg, Down syndrome), with SCIWORA, and without complete time to procedure (TTP) data were excluded. Patients were stratified into 7 groups by TTP: same-day as admission (SD), 1-day delay (1D), 2-day delay (2D), 3-day delay (3D), 4-7 days delay (4-7D), 8-14 days delay (8-14D), >14 days delay (>14D). In an attempt to reduce covariate bias, groups were propensity score matched (PSM) by age, comorbidity index (CCI), mechanism of injury (MOI) (fall, MVA, pedestrian), trauma status at admission (hypotension, shock, hemorrhage, intubation), and concurrent injuries. Surgical details, perioperative complications, length of stay, total charges, and discharge disposition was compared. Binary logistic regressions determined independent predictors of varying complications (reference: same-day). RESULTS A total of 28,414 patients were included. After PSM, 6,636 patients remained (948 per group). Overall age 49.3, gender: 25.4%F, 67.4% white, 15.1% black, 11.4% hispanic, CCI: 1.2. Most common MOIs were 38.2% MVAs, 32% falls, 19.9% pedestrian accidents, 5.7% assaults, 4.2% sports. Procedure rates were 64.2% spinal fusion (36.8% 2-3 lvls, 16.1% 4-8 lvls, 1.4% >8 lvls), 32.5% decompressions, 14.7% halo/traction. SD was associated with the highest mortality (28.8% vs. 6.1-11.6%), lowest LOS (15.14 vs. 15.24-54.2days) and total hospital charges ($172,086.93 vs. $204,931.12-$545,797.14), all p 14D-OR: 10.2[3.9-26.7]), discharging to another care facility (1D-OR: 3.0[2.1-4.3] a 8-14D-OR: 3.1[2.2-4.3]), or discharging with quadriplegia (1D-OR: 1.1[0.8-1.4] a >14D-OR: 1.5[1.2-1.9], exception 3D-OR: 0.9[0.6-1.1]). 2D and 3D were significantly less likely to develop ARDS (OR: 0.8[0.6-1.0], OR: 0.9[0.7-1.1]) and any complication (OR: 0.9[0.8-1.1], OR: 0.7[0.6-0.9]), while all delay groups were less likely to develop sepsis (ORs: 0.6-0.9, exception >14D OR: 2.3[1.7-3.1]) and paraplegia (ORs 0.0-0.8) compared to SD. CONCLUSIONS Patients operated on the same day as admission were significantly less likely to develop infection, respiratory complications, or discharge to another care facility. Same-day operative patients were also less likely to discharge with quadriplegia, and more likely to discharge with paraplegia, indicating early intervention may significantly benefit discharge neurologic status. 2-day and 3-day operative patients exhibited significantly less risk of developing ARDS, and complication, or sepsis. While immediate and 2-3 day delayed operations appear to have unique advantages, patients who underwent procedures >14 days after admission were associated with poor outcomes and discharge disposition. FDA DEVICE/DRUG STATUS This abstract does not discuss or include any applicable devices or drugs.