Reliability and accuracy of fine-cut computed tomography scans to determine the status of anterior interbody fusions with metallic cages

Abstract 

Background context

Computed tomography (CT) scan has been shown to be more accurate than radiographs in evaluating anterior interbody fusion but may still over-read the extent of fusion.

Purpose

To assess the reliability and accuracy of fine-cut CT scans with reconstructions in evaluating anterior lumbar interbody fusion (ALIF) with metallic cages using surgical exploration as the reference standard.

Study design

Accuracy of a diagnostic test referenced to the gold standard.

Patient sample

A total of 49 patients and 69 surgical levels.

Outcome measures

Evaluation of fine-cut CT scans for evidence of fusion with subsequent surgical exploration as the reference standard.

Methods

Forty-nine patients who underwent ALIF with metallic cages over 69 levels, who had a fine-cut CT scan before revision were included. Five spine surgeons unaware of the findings on surgical exploration evaluated pre-revision CT scans, classified these as fused or not; and determined the presence of a “sentinel sign” and a “posterior sentinel sign.” Kappa coefficients for interobserver reliability, sensitivity, and specificity to detect fusion were determined.

Results

There were 26 males and 23 females with a mean age of 43 years. There were 27 smokers. Average time from index to revision surgery was 22 months. Interobserver kappa for classification as fused or not was 0.25 with 70% to 97% sensitivity and 28% to 85% specificity. The interobserver kappa for the sentinel sign was 0.34 with 13% to 33% sensitivity and 77% to 92% specificity. The interobserver kappa for the posterior sentinel sign was 0.23 with 33% to 87% sensitivity and 56% to 90% specificity.

Conclusions

Raters generally overstated fusion with low specificities across raters and low consensus specificity. Overall accuracy of the posterior sentinel sign (74%) was higher than the sentinel sign (61%). The low kappa value indicates fair reliability. In patients with metallic interbody devices, surgeons should be cautious about interpreting the findings on fine-cut CT scans whether using a general assessment of the fusion, the sentinel sign, or the posterior sentinel sign.

Keywords: Diagnostic reliability, Diagnostic accuracy, Computed tomography, Lumbar fusion, Anterior spine fusion