Cervical Fracture Patterns Associated with Blunt Cerebrovascular Injures When Utilizing Computed Tomographic Angiography: A Systematic Review and Meta-Analysis.

BACKGROUND CONTEXT: Prior studies have demonstrated an association between cervical spine fractures and blunt cerebrovascular injuries (BCVI) due to the intimate anatomic relationship between the cervical spine and the vertebral arteries. Digital subtraction angiography (DSA) has historically been the gold standard, but computed tomography angiography (CTA) is commonly used to screen for BCVI in the trauma setting. However, there is no consensus regarding which fracture patterns mandate screening. Over aggressive screening may lead to increased radiation, increased false positives, and overtreatment of patients which can cause unnecessary patient harm, and increased healthcare costs.

PURPOSE: The aim of this meta-analysis is to analyze which cervical spine fracture patterns are most predictive of BCVI when utilizing CTA.

STUDY DESIGN/SETTING: Systematic review and meta-analysis OUTCOME MEASURES: Odds ratios for specific cervical fracture patterns and risk of developing a BCVI METHODS: A systematic literature review of all English language studies from 2000-2020 was conducted. The year 2000 was chosen as the cut-off because use of CTA prior to 2000 was rare. Ovid MEDLINE®, Embase, Cochrane Central Register of Controlled Trials, Scopus, Global Index Medicus, and ClinicalTrials.gov were queried. Studies were included if they met the following criteria: 1) the diagnostic imaging modality was CTA; 2) investigated blunt cervical trauma; 3) noted specific cervical spine fracture patterns associated with BCVI; 4) odds ratios for specific cervical spine fracture patterns or the odds ratio could be calculated; 5) subjects were 18 years old or older. Studies were excluded if they: 1) included DSA or magnetic resonance imaging; 2) included penetrating cervical trauma; 3) included pediatric patients less than 18 years of age; 4) were not written in English. All statistical analysis was performed using R Studio (RStudio, Boston, MA, USA).

RESULTS: The initial search, after duplicates were removed, resulted in 10,940 articles for independent review. Six studies met the criteria for inclusion in the meta-analysis. Specific fracture patterns mentioned are isolated C1, C2, C3 fractures, any C1-C3 fracture, any C4-C7 fracture, two-level fractures, subluxation/dislocations, and transverse foramen (TF) fractures. Three studies were included in the meta-analysis for C1, C2, C1-C3, subluxations/dislocations, and TF fractures. Two studies were included in the meta-analysis for C3, C4-C7, and two-level fractures. The pooled odds ratio with 95% confidence interval for: C1 fractures and BCVI is 1.3 (0.8-2.1); C2: 1.6 (0.9-2.8); C3: 1.8 (0.9-3.6); C1-C3: 2.2 (1.1-4.2); C4-C7: 0.7 (0.3-1.7); Two-level: 2.5 (1.4-4.6); Subluxation/Dislocation: 2.9 (1.8-4.5); TF: 3.6 (1.4-8.9).

DISCUSSION/CONCLUSION: This study found that when utilizing CTA for screening of BCVI only fractures in the C1-C3 region, two-level fractures, subluxations/dislocations, and transverse foramen fractures were associated with increased incidence of a BCVI. Further refinement of protocols for CTA in the setting of blunt cervical trauma may help limit unnecessary patient harm from overtreatment and reduce healthcare costs.