Management of Displaced Femoral Neck Fractures: Does Fellowship Training Matter?

Introduction Femoral neck fractures (FNFs) are a common source of morbidity and mortality in older adults, often leading to loss of independence and functional decline. Arthroplasty is the gold standard for displaced FNFs, yet the choice between hemiarthroplasty and total hip arthroplasty (THA) remains debated and case specific. We sought to evaluate training bias in treatment selection as well as evaluation of patient outcomes for arthroplasty (AR) surgeons vs non-AR surgeons. We hypothesized that AR surgeons would perform more THA and have improved patient outcomes when treating FNFs. Methods Patients aged ≥50 years with FNFs treated with hemiarthroplasty or THA between January 2014 and May 2025 were retrospectively identified. Demographics, comorbidities, OR time, and complications were compared based on practice type (AR vs non-AR). A sub-analysis of 30 FNF cases performed by non-AR surgeons were blinded to the treatment, and AR surgeons were surveyed for hemiarthroplasty vs THA based on radiographic data and clinical presentation to evaluate treatment differences between groups. Results 626 patients were evaluated, including 476 with hemiarthroplasty and 150 with THA. AR-trained surgeons were over 4x more likely to perform THA for FNF compared with non–AR-trained surgeons (65.7% vs 15.3%; p<0.001). For THA, operative time was significantly shorter among AR-trained surgeons (115.3 ± 31.4 vs 135.2 ± 36.4 minutes; p=0.001). 1-year mortality, readmissions, revisions, and infections were higher in Non-AR but did not reach statistical significance. In sub-analysis survey, AR-trained surgeons selected THA in 65% of cases originally treated with hemiarthroplasty by non–AR-trained surgeons (p=0.023). Conclusions The practice of AR influenced treatment choice for FNF as well as OR time. Management of these patients may warrant larger review prior to intervention or call assignments generated for surgeons more comfortable with THA procedures. Further study of patient outcomes in the setting of treatment bias is required.