Study design: Prospective assessment of the reliability of determining cervical fusion success based on plain radiographs.
Objectives: Determination of the reliability of plain static radiographs in predicting the presence or absence of fusion.
Summary of background data: The ability of plain radiographs to assess the presence of fusion is limited. In addition, variations in the definition of "fusion" make this entity an important aspect for study.
Methods: A study was carried out to determine the reliability of plain radiographs in predicting bony fusion. Cases of cervical spondylosis undergoing a single or multilevel corpectomy with an allograft fusion and anterior instrumentation were chosen for the model. The definition of "bony fusion" was obtained from the literature. Bony fusion was defined by the presence of bony trabeculation across the graft-host interfaces, the assessment of the change in strut height over time, and the development of a kyphotic angulation over time. Data were collected at a tertiary care institution over a 5-year period. Descriptive statistics regarding baseline patient characteristics, the underlying disease process, and the surgical intervention, were obtained. Reliability of plain static radiographs in assessing fusion was evaluated by two independent neuroradiologists blinded to any subsequent clinical outcome. The Cohen Kappa statistic was used to determine the degree of agreement regarding the presence or absence of fusion at the superior and inferior aspect of the graft at the 6-week and the 12-week follow-up.
Results: The study involved 57 patients (30 males and 27 females), with a median age of 49 years. The number of levels decompressed was 1, 2, and 3 in 36, 20, and 1 patients, respectively. Fourteen patients had a history of smoking. The Cohen Kappa statistic revealed variable results depending on the time period and aspect evaluated. The degree of agreement at 6 weeks was 0.61 (95% confidence interval = 0.32-0.89) and 0.44 (95% confidence interval = 0.017-0.86) and at 12 weeks was 0.18 (95% confidence interval = -0.21-0.58) and 1.00 for the superior and inferior aspect of the graft, respectively.
Conclusions: Plain radiographs are generally quite unreliable in predicting fusion based on presence or absence of trabeculation.