Issue: May 2006
May 01, 2006
3 min read

Surgeons quantify stability with new cervical spine injury classification system

Investigators claim the Cervical Spine Injury Severity Score system is both reliable and valid.

Issue: May 2006
You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact

CHICAGO – University of Wisconsin investigators developed a new classification system for subaxial cervical spine injuries that allows observers to quantify stability.

A recent 34-patient study testing the validity and reliability of the system found an average 0.98 intraobserver interclass correlation coefficient (ICC), an average 0.89 interobserver ICC and excellent validity, according to Paul A. Anderson, MD, of the University of Wisconsin.


Paul A. Anderson, MD [photo]
Paul A. Anderson

“We felt there had to be a change in the paradigm of how we [classify cervical spine injuries], and it took me almost 20 years to figure this out,” Anderson said at the American Academy of Orthopaedic Surgeons 73rd Annual Meeting. “Taking care of patients with fractures this bad, you really need a description so that we can communicate, and then you need validity to quantify how severe that individual injury is on a stability scale. Finally, you need to add the neurologic state.”

The Cervical Spine Injury Severity Score system divides the spine into four columns: anterior, right pillar, left pillar and posterior. Observers grade each column on an analog scale from 0 (no injury) to 5 (worst-case injury), based on fracture displacement and ligamentous injury severity, Anderson said. The total score ranges from 0 to 20.

“A [score of] 1 is a nondisplaced fracture and a 5 is the worst injury you think could happen in that column,” Anderson said. “So a 5, for instance in the facet, might be a facet dislocation with 7 mm of subluxation. In the posterior column, it would be a complete posterior ligamentous complex disruption and a wide separation of the spinous process.”

He added that intermediate values are assigned to injuries with various degrees of displacement. Fractional values can be used and, if necessary, examiners may also upgrade in particular instances, such as for patients with severe degenerative ankylosis or ankylosing spondylitis that affects spine rigidity, Anderson said.


Injuries among the 34 study patients ranged from “mild” to “terrible,” Anderson said. Fifteen observers participated in the study, including residents, fellows, musculoskeletal radiologists and attending orthopedic and neurosurgeons with experience in treating acute cervical spine injuries.

In the Cervical Spine Injury Severity Score system, reviewers rate the injury in each spinal column on an analog scale of 0 to 5. A score of 0 indicates no fracture, while a score of 5 indicates the worst type of fracture for the specific column.

Courtesy of Paul A. Anderson

The observers reviewed plain radiographs and CT scans using EFilm Lite (Merge eMed), an image viewing software program. Investigators randomly duplicated five cases to assess intraobserver reliability, Anderson said.

Intraobserver ICCs ranged from 0.97 to 0.99 with an average 0.98. Interobserver agreement ICCs ranged from 0.75 to 0.97 (average 0.89). An ICC score greater than 0.75 typically indicates excellent reliability, as shown in past studies, he said.

“If you remember back to the AO and Denis systems, they were only at 0.4 and 0.5 [ICC], respectively, so this is significantly better,” Anderson said.

Researchers also found no difference among examiners based on experience. Residents and fellows had an average ICC of 0.871, while attending surgeons and radiologists had an average of 0.894.

To assess validity, researchers evaluated the correlation between scores and treatment with the concept that the higher the score, the more necessary it is for the patient to have surgery.

“All patients who had a mean severity score over 7 had surgery,” Anderson said. “That was 14 out of the 34 cases, whereas only 15% of the patients with a score of less than 7 had surgery.”

Researchers also checked validity in relation to neurologic deficit. Of the patients with scores greater than 7, 65% had a neurologic deficit, compared to only 15% of cases with scores below 7. “Two of those three [with scores below 7] had neurological deficit radiculopathy that we ended up operating for,” Anderson said.

Anderson and his colleagues plan to conduct a study using the Cervical Spine Injury Severity Score system with MRI.

For more information:
  • Anderson P. Classifying cervical spine injuries. #21. Presented at the Federation of Spine Associations Specialty Day Meeting. March 25, 2006. Chicago.