Which Malpositioned Pedicle Screws Should Be Revised?

BACKGROUND

Up to 10% of free-hand pedicle screws are malpositioned, and 1 in 300 patients may undergo return to surgery for revision of malpositioned screws. The indications for revision of asymptomatic malpositioned screws have not been carefully examined in the literature. We sought to evaluate the threshold among spinal deformity surgeons for revision of malpositioned screws.

METHODS

Twelve experienced spine surgeons reviewed x-ray and computed tomographic images of 32 malpositioned pedicle screws with variable degrees of anterior, medial, and lateral breeches. The surgeons were asked whether based on the image they would revise the screw: (1) intraoperatively before rod placement; (2) intraoperatively after rod placement; (3) in clinic with an asymptomatic patient. For each scenario, we assumed stable neuromonitoring and no neurological changes. Agreement and multirater κ was calculated.

RESULTS

There was good agreement as to which screws were malpositioned (80% agreement, κ=0.703). After the rod was placed or postoperatively (scenarios 2 and 3), surgeons less frequently recommended screw revision, and there was greater variability among the surgeons' recommendations. For return to surgery from clinic for asymptomatic screw revision, % agreement was only 65% (κ=0.477). The majority recommended revision surgery for screws which approached the dura (10/12) or the aorta (7/12 surgeons). Half of the surgeons recommended revision surgery for an asymptomatic screw if the entire screw diameter was in the canal. Revision surgery was not recommended for asymptomatic patients with screws partially violating the canal (<½ the screw diameter), malpositioned laterally in the rib head, or with small anterior cortical violations remote from a vascular structure.

CONCLUSIONS

There is significant variability of opinion among surgeons regarding which malpositioned screws can be safely observed in an asymptomatic patient. Given the frequency of malpositioned screws and morbidity of surgical return to surgery, more long-term data are needed to develop practice guidelines for determining which screws require revision surgery.

LEVEL OF EVIDENCE

Level III-retrospective comparative study.