Neurological Complications of Posterior Spinal Surgery: Incidence and Clinical Features / 대한척추외과학회지

ABSTRACT

STUDY DESIGN: Retrospective study. OBJECTIVES: To identify clinical features and risk factors helpful for the prevention and early diagnosis of neurological complications. OVERVIEW OF LITERATURE Previous studies have investigated postoperative complications only for specific disease entities and did not present distinctive clinical features. MATERIALS AND METHODS: This was an observational study of patients who underwent posterior thoracolumbar spinal surgery in the orthopedic department of a single hospital over the course of 19 years (1995-2013). The incidence, cause, onset time, and risk factors of complications were investigated. Neurological deterioration was graded on a 5-point numeric scale G1, increased leg pain or sensory loss, G2, unilateral motor weakness; G3, bilateral motor weakness; G4, cauda equina syndrome; and G5, complete paraplegia. RESULTS: Sixty-five cases out of 6574 (0.989%) developed neurological complications due to the following causes epidural hematoma, 0.380%; instrumentation with inadequate decompression, 0.213%; mechanical injury, 0.167%; inadequate discectomy, 0.061%; and unknown cause, 0.167% (p=0.000). The grade of neurological deterioration was G1 in 0.167% of patients, G2 in 0.517%, G3 in 0.228%, G4 in 0.046%, and G5 in 0.030%. Neurological deterioration was most severe in patients who experienced epidural hematoma, followed by those in whom complications occurred due to instrumentation with inadequate decompression, unknown causes, mechanical injury, and inadequate discectomy, in order (p=0.009). Revision surgery was a significant risk factor (p=0.000; odds ratio, 2.741). The time that elapsed until symptom development was as follows, in order unknown cause, 0.6 hours; epidural hematoma, 5.4 hours; mechanical injury, 6.6 hours; inadequate discectomy, 18.0 hours; and instrumentation with insufficient decompression, 36.0 hours (p=0.001). CONCLUSIONS: The incidence of neurological complications in our cohort was 1%. Revision surgery increased the risk by 3 times. Severe cases (cauda equina syndrome or complete paraplegia) rarely developed, occurring in 0.08% of patients. The major causes of neurological decline were epidural hematoma and instrumentation with inadequate decompression. Close observation in the early period was important for the diagnosis because most patients developed symptoms within 12 hours. Delayed diagnosis was most common in complications caused by instrumentation with inadequate decompression.