The pedicle width predicts an accurate screw insertion

Introduction. Correlation between pedicle screw malposition and small values of pedicle morphometric parameters has been confirmed in numerous studies. Definition of critical pedicle size for screw insertion is an actual problem for pediatric spinal surgery.

Material and methods. 29 patients, aged 3-17, with congenital or acquired spinal deformities were included in the study. All the patients had posterior surgery with pedicle screw implantation. All the screws were inserted by free hand technique. On preoperative CT, external pedicle width, internal pedicle width, and spongiosa proportion were measured. On postoperative CT, pedicle screw accuracy was evaluated. The binomial logistic regression was used to define dependence of pedicle screw accuracy on pedicle morphometric parameter values. ROC-curves were graphed, and AUC were calculated.

Results. 233 pedicle screws were implanted to 29 patients by free hand technique. On postoperative CT, 191 (82%) screws were confirmed to be accurately inserted. The logistic model confirmed significance of all the examined morphometric parameters (p<0.001). The external pedicle width possessed the maximal predictive value. Statistical indices for the prognostic model (sensitivity, specificity, and accuracy) were calculated for pedicle width 3.5; 6.0; 7.5 mm.

In the cut-off value of external pedicle width 3.5 mm, probability of accurate screw insertion is about 50%; this technique has been highly sensitive and maximally accurate. This morphometric feature is a technical limit of free hand pedicle screw insertion. Recommendations for selecting an implantation technique in different pedicle width are proposed.

Conclusion. The external pedicle width 3.5 mm is a critical one for pedicle screw insertion by the free hand technique.