Abstract
Screw loosening remains a prominent problem for osteoporotic patients undergoing pedicle screw fixation surgeries and is affected by screw parameters (e.g., diameter, pitch, and thread angle). However, the individual and interactive effects of these parameters on screw fixation are not fully understood. Furthermore, the current finite element modeling of a threaded screw is less computationally efficient. To address these issues, we (1) explored a novel “simulated threaded screw” approach (virtual threads assigned to the contact elements of a simplified screw) and compared its performance with threaded and simplified screws, and (2) examined this approach the individual and interactive effects of altering screw diameter (5.5–6.5 mm), pitch (1–2 mm) and half-thread angle (20–30 deg) on pullout strength of normal vertebrae. Results demonstrated that the “simulated threaded screw” approach equivalently predicted pullout strength compared to the “threaded screw” approach (R2 = 0.99, slope = 1). We further found that the pullout strength was most sensitive to the change in screw diameter, followed by thread angle, pitch, and interactions of diameter*pitch or diameter*angle. In conclusion, the “simulated threaded screw” approach can achieve the same predictive capability compared to threaded modeling of the screw. The current findings may serve as useful references for planning of screw parameters, so as to improve the complication of screw loosening.