Inter- and Intraobserver Reliability for the Computed Tomography Scan Assessment of Union After Surgery for Scaphoid Fractures and Nonunion.

PURPOSE: Assessment of scaphoid union after operative treatment might be associated with uncertainty, disagreement, and misinterpretation. Union is defined as consolidation on three of four X-ray views. Computed tomography (CT) scans are increasingly used to evaluate union of scaphoid fractures and nonunion. It allows multiplanar assessment of the scaphoid. Reliability in CT scan assessment of union after operative intervention is limited. We hypothesized that inter- and intraobserver reliability of the CT scan assessment of union after operative treatment for scaphoid fracture and nonunion are substantial and better in fracture cases.

METHODS

An institutional search identified 230 patients with operative intervention. According to the sample size, we randomly selected 60 patients (30 fractures and 30 nonunion), with age above 18 years and CT scans 6-26 weeks after surgery, when uncertainty about healing status is most pronounced. Three observers evaluated the CT scans on two occasions 6 weeks apart. Observers classified the scaphoid into two categories: No/partial/full union and >/<50% union.

RESULTS

Interrater agreement in no/partial/full union was overall substantial, substantial in fractures, and moderate for nonunion cases. Interobserver agreement for >/<50% union was overall moderate and moderate for both fractures and nonunion groups. The interrater agreement of no/partial/full union with CT scans ≥12 weeks after surgery was substantial versus fair for the cases below. Intrarater agreement was overall better in cases with CT scans ≥12 weeks postoperative versus earlier.

CONCLUSIONS

CT scans show substantial agreement in the assessment of no/partial/full union after operative intervention, although the presence of metal artifacts and bone grafts potentially impair visibility among observers. The agreement was better in scaphoid fractures. CT scan assessment of >/<50% union is with overall moderate agreement. Inter- and intrarater agreement is improved when CT scans are taken 12 weeks or later after operative treatment, blinded to the clinical information.

LEVEL OF EVIDENCE: Diagnostic, III.