Objectives: To better characterize operative agreement and disagreement among orthopaedic surgeons treating lateral compression type 1 (LC-1) pelvic fractures in an effort to improve communication between care providers and improve patient care.
Design: Decision analysis.
Setting: Level 1 trauma center.
Methods: Twenty-seven LC-1 cases were selected to represent a wide array of LC-1 injuries. Each case was presented with 3 plain pelvic radiographs (anteroposterior, inlet, and outlet) and a scrollable computed tomography at the OTA national meeting. Attendees were queried whether they would perform operative stabilization "yes/no." Years of surgical practice (0-5, 6-10, and >10), annual pelvic fracture case volume (0-20, 21-50, and >50), and completion of a trauma fellowship (yes/no) were also collected. Fleiss' kappa (K) was used to measure operative agreement among survey respondents, where K = 0.21-0.40 was fair and K = 0.41-0.60 was moderate agreement.
Results: One hundred eleven surgeons completed the survey where the average tendency to operate across surveys was 40%. Of the 27 cases presented, only 9 cases (33%) showed substantial agreement. There were 4 cases where nearly everyone chose operative stabilization (93.1%-94.4%) and 5 cases where nearly no one chose operative stabilization (0%-8.7%). The overall agreement was fair with K = 0.39 [95% confidence interval (CI), 0.34-0.44]. Although there was a trend for surgeons with more years of surgical practice to have a lower tendency to operate, it did not achieve statistical significance (odds ratio for >10 years vs. 0-5 years = 0.73; 95% CI, 0.48-1.11). Annual case volume and completion of a trauma fellowship were not statistically significant predictors of operative tendency.
Conclusions: Our results show only fair operative agreement (K = 0.39; 95% CI, 0.34-0.44) in a radiographic survey representing a broad range of LC-1 fracture morphologies among OTA surgeons. Only 9 of the 27 cases (33%) had substantial agreement. There was no difference in the decision to operate based on surgical volume, completion of a trauma fellowship, or time in practice. These results highlight the differing practice decisions among surgeons currently treating LC-1 injuries, and there is need for further studies to more fully understand stability after this injury pattern.