Editorial: Load-sharing score

Abstract

The management of thoracolumbar burst fractures varies considerably among spine surgeons. The loadsharing classification was initially published to predict the success of short-segment posterior fusions for thoracolumbar burst fractures.5 In the following paper, Radcliff et al. studied the relationship between the loadsharing score (LSS) and different aspects of these burst fractures.7 They concluded that the LSS did not correlate with posterior ligamentous complex (PLC) injury, neurological status, or management decision. An interesting finding in this paper is the relatively poor interobserver reliability in determining the LSS. Although the average kappa value for interobserver agreement on the assigned LSS in their report falls within the range of reported results,1,3 the kappa value for an LSS > 6 was relatively low (0.24). When the same injury is given different scores by different observers, it is expected that the assigned score will have a questionable validity and will not correlate with other variables reflecting injury severity or management decisions. Despite acknowledging the issue of interobserver agreement in their report, the authors’ conclusions about the limitations of the LSS are valid. The LSS was only meant to assess the adequacy of anterior column support to allow for a clinical and radiographically successful short-segment posterior fusion. While a higher LSS would be expected to be associated with an increase in PLC injury, this relationship was not established by most of the reviewers. The failure of the LSS to include neurological status is an additional shortcoming. This limitation also occurs in other more comprehensive classification systems.2,4 A classification system for thoracolumbar injuries that would guide management must account for both clinical and radiological variables, and recently such scoring systems have been developed.6,8 One limitation of these systems is the predilection to consolidate heterogeneous injuries into a one-dimensional severity score, and although they may accurately predict which patients will need anterior, posterior, or 360° surgery, the levels requiring treatment in those patients requiring posterior surgery are not addressed. It is in this subset that the LSS has value. If the LSS score is high, then a multilevel instrumented fusion should be considered. The treatment of spinal fractures is complex, requiring a specialist in spinal surgery, and optimal outcomes are achieved when the treating surgeon utilizes all available data. The decision process for the management of thoracolumbar burst fractures must be done on an individual basis in every case after careful consideration of all relevant clinical and radiological variables. Occasionally, even other injuries or the patient’s overall health will influence the process, and it is difficult to make all decisions based on a single scoring or grading system. We congratulate the authors on this well-written paper addressing a timely and important topic. (http://thejns.org/doi/abs/10.3171/2012.1.SPINE111007)