Can lumbar spine radiographs accurately determine fusion in postoperative patients? Correlation of routine radiographs with a second surgical look at lumbar fusions.

Controversy centers on the determination of surgical fusion in lumbar spinal patients. What method best determines the nature of surgical arthrodesis remains unanswered. Numerous studies have investigated the accuracy of different radiologic tests. Although the best method has not been determined, plain radiography is certainly the most widely used in many centers and reported in scientific articles. In most of the literature a poor agreement between radiographic interpretation and surgical findings was observed. The main reasons seemed to be the lack of an accurate method to assess the radiographs. The authors rely on a radiologist or spinal surgeon to estimate the success of bony fusion by reviewing a two-dimensional radiograph. The purpose of this study was to determine the accuracy of plain radiographs to predict the presence of a surgical fusion. Forty-nine patients underwent fusion site exploration in the course of hardware removal. All patients had a one- or two-level posterolateral fusion and posterior lumbar interbody fusion with pedicle screw/link rod instrumentation. Immediate preoperative anteroposterior and lateral radiographs were taken before hardware removal that included both visual assessment and a Kocher mechanical test. Two spinal surgeons and two musculoskeletal radiologists blindly judged the preoperative radiographs as to the absolute presence or absence of successful arthrodesis. A second review was repeated at 3 months. The overall agreement between radiographic assessment and actual surgical findings was 69%. The range among observers was 57-77%. The overall false positive rate was 42% (0-75%), while the false-negative rate was 29% (20-51%). Success of observed surgical arthrodesis at the time of the second look was 90% and this number was used as the standard in the agreement process. In comparing the radiographic observations with the surgical findings it is suggested that in one of five cases the plain radiographs underestimate the degree of fusion. This finding agrees with the authors' knowledge of osteoid and mineralized bone. The premineralized osteoid may be functionally fused, but appear radiolucent on radiographic film. Once solid trabecular bony bridging occurs radiographic identification of fusion is easier to determine.