Classifications and level of evidence trends from the most influential literature on thoracolumbar burst fractures: A bibliometric analysis

Background: There are known classifications that describe thoracolumbar (TL) burst type injury but it is unclear which have the most influence on management. Our objective is to investigate the association of classification publications with the quantity and type of the most influential articles on TL burst fractures.

Methods: Web of Science was searched, and exclusion and inclusion criteria were used to extract the top 100 cited articles on TL burst fractures. The effects on type, number, and other variables were separated into four eras as defined by four major classification publications.

Results: 30 out of the top 100 articles represent level 1 or 2 evidence. The most influential journal was Spine, accounting for 35 articles and 4,537 citations. The highest number of articles (53) was published between the years 1995-2005, culminating with the Thoracolumbar Injury Severity Classification Score (TLICS) paper. After 2005, there was an increase in average citations per year. Following 2013, the number of highly influential articles decreased, and systematic reviews (SRs) became a larger proportion of the literature. There was a statistically significant increase in the level of 1 and 2 evidence articles with time until the publication of TLICS. The predictive value of time for higher levels of evidence was only seen in the pre-2005 years (AUC: 0.717, 95% CI 0.579-0.855, p = 0.002).

Conclusions: In 1994, two articles marked the beginning of an era of highly influential TL burst fracture literature. The 2005 TLICS score was associated with a preceding increase in LOE and productivity. Following 2005, the literature saw a decrease in productivity and an increase in systematic review/meta-analysis (SR-MAs). These trends represent an increase in scholarly discussion that led to a systematic synthesis of the existing literature after publication of the 2005 TLICS article.