RESUMO
BACKGROUND: We evaluated which treatment decisions in the management of displaced femoral neck fractures (FNFs) may associate with measures of resource utilization relevant to a value-based episode-of-care model. METHODS: A total of 1139 FNFs treated with hip arthroplasty at 7 hospitals were retrospectively reviewed. Treatment choices were procedure (hemiarthroplasty vs total hip arthroplasty [THA]), surgeon training status, admitting service, and time to surgery. Dependent variables were length of stay, discharge disposition, 30-day readmission, and in-hospital mortality. Variation across hospitals was evaluated with analysis of variance and chi-square tests. Treatment choices were evaluated for the dependent variables of interest with univariable and multivariable regression. RESULTS: There was significant variation between hospitals regarding proportion of cases treated with THA (range = 3.0%-73.2%, P < .001), proportion treated by arthroplasty fellowship-trained surgeons (range = 0%-74.9%, P < .001), proportion admitted to the orthopedic service (range = 2.8%-91.3%, P < .001), mean time to surgery (range = 0.9-2.1 days, P < .001), and proportion of discharge home (range = 63.9%-97.8%, P < .001). Multivariable analysis adjusting for age, gender, and Charlson Comorbidity Index demonstrated correlations between (1) decreased length of stay and admission to orthopedics (B = -1.256, P < .001); (2) lower 30-day readmission and THA (odds ratio [OR] = .376, P = .004), and (3) decreased discharge to a care facility and admission to orthopedics (OR = 0.402, P = <.001), THA (OR = 0.435, P = .002), and treatment by an arthroplasty fellowship-trained surgeon (OR = 0.572, P = .016). None of the treatment variables tested associated with in-hospital mortality. CONCLUSION: We observed significant variation in the treatment of displaced FNF patients across 7 hospitals and identified treatment choices that associated with resource utilization within the episode of care. Future, prospective study is necessary to understand whether care pathways that adapt some combination of these characteristics may result in more value-based care.