Researcher: X-rays show poor sensitivity for assessing reduction after ankle fixation

Issue: September 2013

Nearly 25% of articular congruencies in patients who undergo open reduction and internal fixation for ankle fractures are incorrectly assessed using X-rays, according to a recently presented study.

Perspective from Ivan S. Tarkin, MD

“Plain radiographs postoperatively had good interobserver reliability and specificity, however the sensitivity was poor,” Matthew R. Garner, MD, said in the presentation.

CT scan data

The researchers prospectively collected data on ankle fracture patients who underwent open reduction and internal fixation for medial malleolar or posterior malleolar fractures and underwent AP, mortise and lateral radiographs and CT scans. The investigators excluded patients with isolated lateral malleolar fractures. One hundred and twenty-two patients fit the database search criteria, and two independent observers reviewed postoperative radiographs.

“For the eight patients where the two reviewers had not had their consensus agreement, they went back together and looked at the images to reach a consensus to provide a comparison to measure against CT, which we used as the gold standard for articular reduction,” Garner said.

Although this patient’s postoperative X-ray was read as normal, the CT shows malreduction of the medial malleolus.

Images: Garner MR

The researchers defined malreduction as articular step-off >2 mm or the presence of an intra-articular loose body.

There were 51 fracture-dislocations, of which four were open dislocations. The researchers identified 15 malreductions on plain radiographs. There were a total of 30 malreductions seen on CT. Of these, 23 were not seen on X-ray. Of those 23 malreductions seen on CT, 70% were in patients who had a dislocation, Garner said. The researchers also found one loose body on plain radiograph and four loose bodies on CT scan. They found the interobserver kappa coefficient for postoperative radiographs was 0.63. The sensitivity for radiographs was 0.23 and the specificity was 0.91.

Limitations

Garner said orthopedic residents acted as the imaging reviewers, and that may have limited the study. He suggested musculoskeletal radiologists or attending orthopedic traumatologists as reviewers for future studies. The researchers were not able to compare fluoroscopic images at the time of surgical fixation because their fluoro machine does not permit calibration, he said.

“Given the ease of attainment, plain radiographs continue to be the standard method for assessment of articular reduction and alignment postoperatively, whether we do that early in the postoperative period or later,” Garner said. “The surgeon should keep in mind that despite good interobserver reliability, the sensitivity tends to be poor for this modality. Twenty-five percent of the time, the X-rays are incorrect assessing articular congruency.” – by Renee Blisard Buddle

Reference:

Garner M. Paper #18. Presented at: American Academy of Orthopaedic Surgeons Meeting; March 19-23, 2013; Chicago.

For more information:

Matthew R. Garner, MD, can be reached at the Department of Orthopaedic Surgery, Hospital for Special Surgery, 535 East 70th St., New York, NY 10021; email: garnerm@hss.edu.
Disclosure: Garner has no relevant financial disclosures.

Perspective

Ivan S. Tarkin, MD

As a moderator at the 2013 American Academy of Orthopaedic Surgeons Annual Meeting, I eagerly awaited the presentation entitled, “Assessment of ankle articular reduction after surgical fixation: Sensitivity and specificity of plain radiographs.” Fortunately, neither I, nor the audience, was disappointed. The presentation from the trauma service at Hospital for Special Surgery was not only well done, but thought-provoking.

The authors used advanced imaging to critically examine the quality of postoperative reduction after operative reconstruction of unstable ankle fracture. Surprisingly, they found that plain radiographs grossly underestimated the rate of malreduction as compared to CT scan. In their large series of 122 cases, 23 cases of malreduction were diagnosed on CT scan that were not readily apparent on postoperative plain films. Further, an additional three cases of intra-articular loose body were identified on postoperative CT scan alone.

Although, the authors did not present outcome data, orthopedic trauma dogma equates quality of joint reduction to success after periarticular fracture. Thus, their sobering data suggests that there is room for improvement in approximately 20% of cases. Perhaps the current operative paradigm for unstable ankle fracture can be improved. Thoughts for further research include innovative ideas to directly visualize/reduce the fractured chondral surface vs. relying on extra-articular reads, improvements in recognizing and addressing articular impaction associated with fracture/dislocation patterns, and refining reconstruction of associated injury to the distal tibial/fibular articulation.

Ivan S. Tarkin, MD

Division Chief Orthopedic Traumatology

University of Pittsburgh Medical Center

Pittsburgh

Disclosures: Tarkin has no relevant financial disclosures.