Static Evaluation of Pedicle Screw Spinal Constructs to the ASTM Standard: A Comparison of Multiple Test Laboratories

Newly acquired mechanical test data along with predicate device data—sometimes from the designer’s historical records—are a requisite part of the regulatory submission process for new pedicle screw spinal constructs. Several studies have compared the mechanical properties of various designs of these medical devices, which are commonly inserted during spinal fusion operations. However, a rigorous comparison of the results coming from different laboratories testing the same devices under controlled conditions has not been performed. Six different test labs performed a series of static bending compression and static torsion tests on identically prepared pedicle screw and rod test constructs (n = 5 for both tests). Sufficient data were acquired to uncover and understand differential interpretations of the methodology described in the standard test method ASTM F1717. For ultimate displacement in compressive bending, the mean values ranged between 47.64 mm and 71.64 mm, and every single laboratory’s data were statistically significantly different from those of every other laboratory. Significant differences in the mean values carried through to the ultimate load data, but this trend did not continue for stiffness, yield displacement, yield load, and elastic displacement. For stiffness in static torsion, the mean values ranged between 0.38 Nm/° and 1.07 Nm/°. There were statistically significant differences among some of the labs for some of the parameters, but no strict patterns emerged. This is likely due to methodological and interpretive differences among the labs, such as the depth of the clevis fixtures and the direction of rotation during torsion testing. These differences in the test labs’ methodology have caused the ASTM subcommittees to clarify the standard so that fewer aspects are open for interpretation, but more work is needed. With appropriate refining of F1717 (and adherence to the new methodology by the test laboratories), test results from different laboratories would likely be more directly comparable than in the current situation.