Prediction of mechanical complications in adult spinal deformity surgery-the GAP score versus the Schwab classification.

Background Context: Surgery for adult spinal deformity is a challenging and complex procedure with high reported complication (8.4%-42%) and revision rates (9%-17.6%). Failure to achieve or maintain adequate postoperative sagittal alignment has been reported to be the main cause of mechanical complications. In order to define appropriate surgical targets, the Scoliosis Research Society-Schwab classification and the Global Alignment and Proportion (GAP) score were established. In the literature, no study has yet compared these classification systems with respect to the risk of developing mechanical complications.

Purpose: To assess and compare the ability of the Schwab classification and the GAP score to predict mechanical complications following adult spinal deformity surgery.

Study Design: Two-center, retrospective cohort study.

Patient Sample: Thirty-nine patients suffering adult spinal deformity who underwent long segment spinal fusion (≥4 levels), minimum follow-up of 2years.

Outcome Measures: The ability of the Schwab classification and GAP score to predict mechanical failure was determined by computing the Area Under the receiver operating characteristic curve.

Methods: Full-spine pre- and postoperative radiographs of all patients were analyzed for mechanical complications. Subsequently, the pre- and postoperative Schwab and GAP score were determined. Logistic regression analysis was used to assess the ability of both systems to determine which was the most appropriate for the prediction of mechanical failure. Correlations between the various factors constituting the GAP score and Schwab classification were estimated using the Spearman rank order correlation coefficient.

Results: The results demonstrated that both classification systems are capable of predicting radiographic evidence of mechanical failure; however, the GAP score proved to be significantly better (p=.003). The relative pelvic version of the GAP score serves a similar role as the pelvic tilt modifier from the Schwab classification (ρ=-0.84, p<.01). The relative lumbar lordosis from the GAP score functions much like the PI-LL modifier from the Schwab classification (ρ=-0.94, p<.01). The GAP score is most significantly dependent on relative spinopelvic alignment, relative lumbar lordosis, and relative pelvic version (ρ=0.85, ρ=0.84, and ρ=0.84, respectively, p<.01). Correlation with the lordosis distribution index was also significant but was not as strong (ρ=0.65, p<.01). Age, on the contrary, showed poor correlation with the GAP score (ρ=0.17, p=.300).

Conclusions: Both the Schwab classification and the GAP score are capable of predicting mechanical complications. The GAP score proved to be significantly more appropriate. This difference is probably attributed to the fact that in the GAP score all parameters are related to the patient's individual pelvic incidence.