Facebook Twitter YouTube



Exploding the Learning Curve Myth: Does Surgical Experience Influence Degree of Curve Correction In AIS?

Saturday, October 20, 2012: 3:20 PM
Melrose (Hilton Riverside)
Terry Amaral1, Vishal Sarwahi1, Meredith Steinman1, Preethi M. Kulkarni1, Elliot Harmon1, Jonathan J. Horn1, Yungtai Lo2 and Adam Wollowick1, (1)Orthopaedic Surgery, Montefiore Medical Center and the Albert Einstein College of Medicine, Bronx, NY, (2)Epidemiology, Albert Einstein College of Medicine, Bronx, NY

Purpose: Previous studies have shown a learning curve associated with the use of pedicle screws in the surgical correction of AIS. The main focus has been on the accuracy of screw placement, and there appears to be continuous improvement as surgeon experience increases. This study analyzes the effect of increasing surgeon experience on the ability to provide three dimensional deformity correction in AIS. 

Methods: 125 AIS patients who had surgery utilizing pedicle screws between 2005 and 2010 were evaluated for three-dimensional deformity correction utilizing x-rays as well as pre- and post-operative CT scans. Charts were reviewed for patient demographics, curve characteristics, length of surgery, blood loss, and number of fixation points. Patients were grouped into three contiguous time periods and analyzed using a Kruskal-Wallis analysis of variance. 

Results: The percentage Cobb correction varied between 73.22 and 86.22% and was not statistically significant different between the three time periods (p=0.217). There was no significant difference between pre- and post-operative kyphosis, sagittal balance, and coronal balance. The preoperative AVR ranged from 16.9 to 17.9 and did not change significantly. This improved to a range of 12.0 to 13.2 post operatively with a percent change ranging from 21.3 to 30.5. The difference was also not significant (p=.487). The screw density significantly increased over time (p=.007) with more screws bring placed in the latter time periods. The number of levels fused increased significantly (p=0.014). The estimated blood loss decreased significantly (p=0.019) as did the surgical time (p<0.001) and anesthesia time (p=0.001). 

Conclusion: A learning curve effect was not seen in this study for the variables investigated. Over time, three-dimensional correction of spine deformity in AIS did not change with improvement in surgical efficiency, but a significant increase in the number of screws inserted and the number of levels fused was seen. The significance of this finding on patient outcomes is unclear and further investigation is warranted.