1 Explanation of search strategy

Our intent was to identify spine-specific disease and trauma severity measures reported in the literature. We intended to identify the most commonly used measures. Therefore, if a measure had a history of consistent use it was included. The disease severity measures include these spine-specific domains: instability deformity degenerative disease, osteoporosis, infection, tumor, stenosis, and heterotopic ossification. The trauma severity measures include spinal cord injury spine fracture or injury classification, instability, and stenosis. The fracture and injury classification measures include fracture or injury to the occipital area, atlanto-occipital area, atlas, axis, subaxial cervical spine, whole cervical spine, thoracolumbar spine, lumbosacral spine, and whole spine. Both disease and trauma severity include a general domain.

We employed a systematic approach in identifying the spine severity measures. We began our process with a group of spine surgeons who created a list of the most common severity measures they use or encounter in the literature. We then proceeded to search all of MEDLINE^® with a relatively generic search code to identify as many possible references in the literature that may contain a spine severity measure. We placed no limits on the date of publication because some severity measures developed several decades ago are the most commonly used spine severity measures today (eg, Nurick Scale for Cervical Myelopathy Severity and the Classification of Occipital Condyle Fractures by Anderson and Montesano). From a large list of initial references, additional severity measures were identified by reviewing the text and bibliographies of full-text articles. Additional severity measures were identified by using generic internet search engines (ie, Google™). A list of severity measures per domain was compiled and built upon throughout the search process. From this list, we first sought to locate the original article, then all studies in the literature that evaluated the severity measure’s predictive validity and reliability (intra- and interobserver). For some, such evaluations were not reported in the literature. For others, there may have been one to several studies reported. Each severity measure was reviewed, summarized, and scored with respect to four major categories:

• 
  

  content

  
  
• 
  

  methodology

  
  
• 
  

  clinical utility

  
  
• 
  

  overall score

Finally each severity measure received an overall score on a scale from 0 to 10 by summing the scores from each major category.

2 Summary of severity measure content

Purpose

The purpose of the severity measure’s content is to give the user a quick reference for understanding what the measure is attempting to assess and how the author(s) recommends it be quantified and interpreted. The content section provides a quick appraisal of the measure’s success at including pertinent anatomy biomechanics, and clinical status (face/content validity), as well as whether it distinguishes between various degrees of severity. Content validity is defined and discussed in Chapter 3.

Method

We divided each measure into five major content areas:

Scoring

Content of the severity measure was scored based on the inclusion of the important diagnostic items in the severity measure’s scale. The measure is awarded half a point for each item that is included in the severity measure (anatomical, biomechanical, clinical, and degree of severity). As a result, each measure could earn from 0 to 2 points for content.

3 Summary of severity measure methodological evaluation

Purpose

The purpose of the spine severity measure’s methodological evaluation is to give the user a quick reference to what populations the measures were evaluated in and how the measures performed when subject to formal evaluation in the literature.

Method

The concepts defined in Chapter 3 (“What makes a quality severity measure?”) were the subject of this evaluation. For some severity measures, there was no record of evaluation. For those that were subject to evaluation, studies reporting results were referenced. We report results based on the interpretation of the author(s) who evaluated the severity measure’s predictive validity and reliability.

Scoring

Scoring for methodology is divided between validity and reliability. One point is awarded if the severity measure was developed by predicting outcome in the index population. A second point is given if the measure was further tested for prediction in a second or test population. Points are given if the measure was tested for intraobserver (one point) or interobserver (one point) reliability. As a result, each measure could earn from 0 to 4 points for methodology.