Accuracy and reliability of the AO/OTA classification for tibial shaft fractures

Background
Available literature lacks information regarding the accuracy and reliability of the AO/OTA classification for tibial shaft fractures. This study aimed to assess the inter- and intra-observer agreement and accuracy of the AO/OTA 42 classification (4-signs) for adult patients with tibial shaft fractures.

Materials and Methods
The study design is an accuracy, inter- and intra-observer agreement study. Anterior posterior (AP) and lateral X-rays of the fracture were used in the examination. The raters comprised of two junior doctors and two orthopedic trauma consultants, who underwent patient scoring twice in a blinded and randomized set-up. A committee including two consultant orthopedic trauma surgeons, one consultant orthopedic radiologist, and one associate professor was established to represent the “gold standard.” The 3- and 4-signs AO classification was used for classification.

Results
A total of 101 patients were included. X-rays were available for all 101 patients. Based on the gold standard classification, AO/OTA 42-A1 (56%) was the most common fracture type, followed by AO/OTA 42-A3 (14%). The agreement at 4-signs, when comparing the four raters and the gold standard classification was between 75% and 86% (Choen’s kappa 0.53 to 0.79). Choen’s kappa coefficient at 4-sgns for intra-and inter-observer agreement was between 0.47 and 0.74 and 0.31 and 0.60, respectively.

Conclusion
This study showed substantial to moderate accuracy of the 4-signs AO/OTA 42- classification for tibial shaft fractures. Intra-observation agreements at 4-signs showed moderate to substantial agreement with and without available CT scans. Inter-observer agreements at 4-signs showed moderate to substantial agreement with only X-rays available. Inter-observer agreements for CT scan at 4-signs showed slight to moderate agreements.