Prevalence and Clinical Impact of Incidental Findings on Preoperative 3D Planning Computed Tomography for Total Shoulder Arthroplasty

INTRODUCTION: 3D planning software for shoulder arthroplasty recently emerged for aiding in intraoperative determination of native glenoid. These protocols often require increased scan resolution, however, raising the question of an increased prevalence and clinical impact of incidental findings (IFs) from preoperative imaging. METHODS: A retrospective review of preoperative shoulder CT reports was conducted for 333 consecutive patients planning anatomic or reverse total shoulder arthroplasties. Patients with thin-sliced CT scans (1.25 mm) were compared with those with standard CT scans (2.5 mm). Poisson regression was performed with baseline characteristics and potentially pathologic IFs (PPIFs). RESULTS: IFs were present in 131 of the 333 scans (39.3%), and 38 of the 333 scans (11.4%) included PPIFs. Only 8 of the 333 scans (2.4%) required workup, with 2 of the 333 (0.6%) leading to new cancer diagnoses. Thin-sliced CT scans detected a higher mean number of IFs (1.12 versus 0.22, P < 0.001) while the mean number of PPIFs remained similar (0.13 versus 0.10, P = 0.43). CONCLUSION: IFs are frequent; however, only 0.6% scans led to new cancer diagnoses. Comparison of thin-sliced with standard CT scans revealed a higher frequency of IFs but similar PPIFs, indicating increased burden of IFs without the benefit of identifying additional malignancies. As demand rises for shoulder arthroplasties, surgeons should consider the potential hidden costs of IFs when using 3D planning programs.