The effect of fracture comminution on the reliability and accuracy of radial head sizing

Abstract

Radial head implant sizing can be based on the maximum head diameter (D-MAX), the minimum head diameter (D-MIN), or the articular dish diameter (D-DISH). The purpose of this study was to assess the reliability of the different radial head sizing techniques and to investigate the effect of radial head fracture comminution on measurement accuracy. Ten observers measured 11 cadaveric radial heads with 3 radial head features (D-MAX, D-MIN, and D-DISH diameter). Radial heads were then fractured into 2, 3, and 4 parts, and the measurements were repeated. Variability was assessed by intraclass correlation. The measurements were compared with the intact state to assess the effect of radial head fracture comminution on sizing accuracy. D-MAX and D-MIN measurements were the most reliable among all observers (intraclass correlation coefficients, 0.980, 0.973). The D-DISH measurement was less reliable (intraclass correlation coefficient, 0.643). Radial head comminution did not significantly affect the reliability of any measurement (P > .2). Fracture comminution, however, significantly affected measurement accuracy with D-MAX and D-DISH. With fracture comminution, D-MAX underestimated radial head diameter (-0.4 ± 0.3 mm; P < .001), whereas D-DISH overestimated diameter (+0.5 ± 0.4 mm; P < .001). Comminution did not significantly affect D-MIN (-0.1 ± 0.3 mm; P = .13). The D-MAX and D-MIN measurements were more reliable than D-DISH for diameter sizing of intact and comminuted radial heads. Overall, increasing comminution did not significantly affect measurement reliability. However, the accuracy of the D-MIN technique was least affected by comminution, suggesting that D-MIN should be used in selecting the diameter of a radial head implant.