The reliability of a ball-tipped probe for detection of breached versus non-breached pedicle drill tracts in the thoracolumbar and lumbar spine of a canine large breed cadaveric model

Background: Spinal stabilization in dogs is challenging, and improper implant placement can result in significant morbidity. Objectives were to evaluate reliability of a ball-tipped probe for detection of breached versus non-breached pedicle drill tracts in a canine large breed cadaveric model. CT scans obtained from T6 to sacrum. Fifty-six pedicles (T7-L7) were randomized to have 1 of 5 modified Zdichavsky drill tracts (non-breached, partial/full medial breach, partial/full lateral breach). All tracts were created using 3D-printed guides. Each pedicle was probed separately by 2 investigators in randomized fashion on 2 occasions, 2 weeks apart. Post-probing CT grading of tracts was performed by one investigator, blinded to probing results. Specificity, sensitivity, positive (PPV) and negative (NPV) predictive values and times to assign grade were compared between surgeons.

Results: CT identified 43 non-breached tracts, 7 medial and 5 lateral breaches. One tract was excluded because of guide misplacement. Specificity was 90.1%. PPV for detection of medial and lateral breaches was 80.0% and 57.1%. Sensitivity for medial (grade 2a/2b) and lateral (grade 3a/3b) breaches was 100.0% and 80.0%. NPV was 98.8%. Overall median time to assign grade with probe was 37.5-59.5 s. Both surgeons had very good intraobserver agreement (k,0.823 and 0.812). Interobserver agreement for round 1 was good (k,0.674) and good-to-very good for round 2 (k,0.800).

Conclusions: Use of a ball-tipped probe is associated with excellent specificity, excellent sensitivity to detect medial breaches, and good sensitivity to detect lateral breaches. This technique may be useful to confirm integrity of pedicle screw/pin tracts in dogs.