The Reliability of the Ball-tipped Probe for Detecting Pedicle Screw Tract Violations Prior to Instrumenting the Thoracic and Lumbar Spine

Abstract

STUDY DESIGN Cadaveric. OBJECTIVE To determine the confidence with which surgeons should rely on a flexible ball-tipped probe to detect pedicle breeches in the thoracic and lumbar spine. SUMMARY OF BACKGROUND DATA The reliability of a ball-tipped probe for detecting cortical violations of the pedicle tract has not been studied among fellowship-trained surgeons. METHODS A total of 134 pedicles were randomized to have pedicle screw tracts with one of six possible options: no violation, anterior, superior, inferior, medial, or lateral violations. Five fellowship-trained spine surgeons examined each pedicle, using a standard flexible ball-tipped probe on three nonsequential occasions. The percentage of correctly identified violations, sensitivity, specificity, positive predictive value, and negative predictive value were calculated for the surgeons as a group and individually. The Cohen kappa coefficient was used to assess the accuracy of the observers and the interobserver and intraobserver agreement. Finally, we analyzed our results by spinal region to see whether this impacted the surgeons' ability to detect a pedicle violation. RESULTS The surgeons were able to correctly identify 81% of intact pedicles, 39% of superior, 68% of medial, 74% of lateral, 62% of anterior, and 50% of inferior violations. The sensitivity varied considerably by breech location and surgeon with a range of 18% to 85%. Positive predictive value for each breech location ranged from 12% to 20%. The specificity was 81% and negative predictive value 98% overall. The intraobserver reliability was moderate and interobserver reliability was low in this series. The ability to detect a pedicle violation was significantly better in the lower thoracic region (T6-T12) than in other areas of the spine. CONCLUSION The standard ball-tipped probe was much less reliable than expected. This technique can be used to confirm an intact pedicle but has an unacceptably high false-positive rate and should be used with caution. Our study suggests that overconfidence in pedicle probing might be dangerous.