Reliability of roentgenogram evaluation of pedicle screw position

Study Design. This was a human cadaver study of the accuracy of biplanar roentgenography in determining pedicle screw position. Objective. To determine the independent accuracy of radiologic evaluation of screw placement and to determine if there are any particular screw malpositions that are most likely to produce a false sense of acceptable screw position. Summary of Background Data. Other investigators have reported the correlation between radiologic evaluation and anatomic dissection. However, in those studies the radiologic evaluation was not independent of the surgeons placing the screws. There has been no comment in the literature regarding particular screw malpositions that would lead the surgeon into a false sense of successful screw placement. Methods. Pedicle screws were placed in cadaver spines, and biplanar roentgenograms of the specimens were evaluated by independent observers. The results of the roentgenogram evaluation then wee compared to those of the anatomic dissection. Results. The accuracy of roentgenogram evaluation varied from 73% to 83%, depending on the experience of the surgeon grading the roentgenograms. Screws misplaced medically into the spinal canal are more likely to give the surgeon a false sense of successful screw placement. Conclusions. The surgeon must not rely solely on the roentgenograms, but instead continue to use tactile sensory skills, anatomic knowledge, and additional modalities such as electromyography monitoring.