Multiplanar reconstruction computed tomography for diagnosis of scaphoid waist fracture union: a prospective cohort analysis of accuracy and precision

To examine reliability and validity concerning union of scaphoid fractures determined by multiplanar reconstruction computed tomography randomized at 6, 12, and 24 weeks after injury. We used Fleiss' kappa to measure the opinions of three observers reviewing 44 sets of computed tomographic scans of 44 conservatively treated scaphoid waist fractures. We calculated kappa for the extent of consolidation (0-24 %, 25-49 %, 50-74 %, or 75-100 %) on the transverse, sagittal and coronal views. We also calculated kappa for no union, partial union, and union, and grouped the results for 6, 12, and 24 weeks after injury. As the reference standard for union, CT scans were performed at a minimum of 6 months after injury to determine validity. Overall inter-observer agreement was found to be moderate (kappa = 0.576). No union (kappa = 0.791), partial union (kappa = 0.502), and union (kappa = 0.683) showed substantial, moderate, and substantial agreement, respectively. The average sensitivity of multiplanar reconstruction CT for diagnosing union of scaphoid waist fractures was 73 %. The average specificity was 80 %. Our results suggest that multiplanar reconstruction computed tomography is a reliable and accurate method for diagnosing union or nonunion of scaphoid fractures. However, inter-observer agreement was lower with respect to partial union.