Kimberly Hamilton1, Darnell T Josiah2, Mark Tierney3, Nathaniel Brooks2. 1. Neurosurgery Department, University of Wisconsin Hospital and Clinics, Madison, Wisconsin, USA. Electronic address: k.hamilton@neurosurgery.wisc.edu. 2. Neurosurgery Department, University of Wisconsin Hospital and Clinics, Madison, Wisconsin, USA. 3. Metro Health, Radiology Department, University of Michigan Metro Health, Wyoming, Michigan, USA.
Abstract
OBJECTIVES: Spine surgeons at a Level 1 Trauma Center have observed a high incidence of spine and spinal cord injuries owing to falls from tree stands. These injuries have been retrospectively reviewed in the context of the Thoracolumbar Injury Classification and Severity (TLICS) and the Subaxial Cervical Injury Classification and Severity (SLICS) classification systems to assess inter-user reliability and validity. We hypothesize that the inter-rater reliability will be similar between neuroradiology and neurosurgery raters and validity of the scoring system will be maintained at our institution. METHODS: The University of Wisconsin Hospital and Clinics' trauma database was reviewed for tree stand-related injuries from 1999 to 2013, with a focus on patients suffering from spine and spinal cord injuries. The TLICS and SLICS scores were then independently determined for these injuries by a neurosurgeon and a neuroradiologist. RESULTS: When cases were grouped by management recommendation (operative, equivocal, and nonoperative) reviewer agreement was 12/15 (80%) of SLICS and 38/52 (73%) of TLICS scores. Operative SLICS positive predictive value reached 100%, however, with a wide confidence interval. Conversely, the SLICS negative predictive value was poor at 54%-60%, with frequent operative treatment for patients assigned nonoperative scores. TLICS scores reached 77.8% and 93.3% positive predictive value per reviewer, whereas negative predictive values reached 93.9% and 89.2%, respectively. CONCLUSIONS: The TLICS and SLICS systems provide good-to-excellent inter-rater reliability. SLICS validity was poor, whereas TLICS was reasonable for nonoperative cases and moderate for operative cases. Systems such as the TLICS and the SLICS may be best applied in the educational setting to confirm the fracture morphology and presence or absence of ligamentous injury between surgeons and radiologists.
OBJECTIVES: Spine surgeons at a Level 1 Trauma Center have observed a high incidence of spine and spinal cord injuries owing to falls from tree stands. These injuries have been retrospectively reviewed in the context of the Thoracolumbar Injury Classification and Severity (TLICS) and the Subaxial Cervical Injury Classification and Severity (SLICS) classification systems to assess inter-user reliability and validity. We hypothesize that the inter-rater reliability will be similar between neuroradiology and neurosurgery raters and validity of the scoring system will be maintained at our institution. METHODS: The University of Wisconsin Hospital and Clinics' trauma database was reviewed for tree stand-related injuries from 1999 to 2013, with a focus on patients suffering from spine and spinal cord injuries. The TLICS and SLICS scores were then independently determined for these injuries by a neurosurgeon and a neuroradiologist. RESULTS: When cases were grouped by management recommendation (operative, equivocal, and nonoperative) reviewer agreement was 12/15 (80%) of SLICS and 38/52 (73%) of TLICS scores. Operative SLICS positive predictive value reached 100%, however, with a wide confidence interval. Conversely, the SLICS negative predictive value was poor at 54%-60%, with frequent operative treatment for patients assigned nonoperative scores. TLICS scores reached 77.8% and 93.3% positive predictive value per reviewer, whereas negative predictive values reached 93.9% and 89.2%, respectively. CONCLUSIONS: The TLICS and SLICS systems provide good-to-excellent inter-rater reliability. SLICS validity was poor, whereas TLICS was reasonable for nonoperative cases and moderate for operative cases. Systems such as the TLICS and the SLICS may be best applied in the educational setting to confirm the fracture morphology and presence or absence of ligamentous injury between surgeons and radiologists.