Reliability of Routine Radiographs for Conservatively Treated Metacarpal Shaft and Neck Fractures.

Abstract

To investigate the reliability of orthopedic hand surgeons to evaluate radiographic healing in initial and follow-up radiographs of the conservatively treated metacarpal shaft and neck fractures. The rationale for this study was to reduce the rate of unnecessary, routine radiographs when treating metacarpal fractures. Forty sets of digital x-rays,twenty at the initial visit and twenty at the 4-week follow-up,were randomly selected and reviewed. Three hand surgeons evaluated the x-rays for (1) fracture location, (2) radiograph timing, (3) healing status, (4) percentage healed, (5) angulation, and (6) confidence in healing status. Observers reviewed studies in random order and evaluated the same set of radiographs one month after the initial review. Intra- and interobserver agreements were analyzed using Fleiss' kappa (κ) for all parameters and all possible observer pairings. Interobserver and intraobserver reliability was highest when evaluating fracture location and lowest when assessing the percentage healed. The interobserver reliability was fair for radiograph timing and healing status and fair-to-moderate for angulation. The intraobserver reliability was moderate for radiograph timing and healing status and moderate-to-substantial for angulation. Observers correctly differentiated initial vs. follow-up images 62% of the time and reported to feel somewhat certain in their evaluation of healing status. When evaluating initial and 4-week follow-up radiographs, hand surgeons were somewhat confident in their assessment of healing but had less than substantial intra- and interobserver reliability following radiographic evaluation. Due to their poor reproducibility, routine radiographs may be unnecessary when evaluating conservatively treated metacarpal fractures. Further studies and guidelines that identify clear indications for the use of routine imaging in metacarpal fracture care are warranted.