MRI for Scaphoid Nonunion: Utilization Rates, Factors Associated with Utilization, and Subsequent Vascularized Bone Graft Use

Purpose: The use of magnetic resonance imaging (MRI) for evaluation of scaphoid nonunion may be an example of low-value imaging for the treatment of scaphoid nonunion. The purpose of this study was to investigate variation in MRI use for scaphoid nonunion, the association of MRI with a vascularized bone graft (VBG) and to develop consensus on MRI use for scaphoid nonunion.

Methods: We identified patients >18 years of age who underwent scaphoid nonunion surgery between 2010 and 2019 using a claims database. Patients who had, and did not have, an MRI within 90 days prior to their diagnosis of scaphoid nonunion were included and a multivariable analysis was performed to evaluate variation in MRI and VBG use. Subsequently, a literature review was performed, and a preliminary consensus statement was developed pertaining to the routine use of MRI for scaphoid nonunion. A consortium of nine hand surgeons evaluated the importance, feasibility, usability, and scientific acceptability of the statement through a modified RAND Coroporation/University of California, Los Angeles Delphi. Panelists evaluated the statement in two voting rounds with an intervening face-to-face discussion.

Results: We identified 1,324 eligible patients with surgical repair of a scaphoid nonunion. Two hundred and sixty-three (19.9%) underwent an MRI within 90 days prior to surgery. Differences in age, insurance type, and comorbidity burden existed between patients who received MRI and those who did not. The MRI cohort was more likely to receive VBG (10.6%) compared to those without an MRI (4.7%). Panelists agreed on the voting domains of the consensus statement and therefore the statement, "There is no benefit of routine MRI/MRA in the treatment of scaphoid nonunion with or without presumed avascular necrosis," was considered valid.

Conclusions: MRI use within 90 days of surgical repair of scaphoid nonunion varies, is associated with greater rates of VBG use, and may represent low-value imaging given the lack of sufficient evidence on this topic.

Clinical relevance: As MRI use for scaphoid nonunion varies and may represent low-value imaging, a validated consensus statement may help guide the evaluation of patients with scaphoid nonunion.