Multi-state analysis of hemi- and total hip arthroplasty for hip fractures in the Swedish population—Results from a Swedish national database study of 38,912 patients

Introduction: Hip fractures are a common problem of the elderly population with significant mortality and morbidity. The choice between total hip arthroplasty (THA) and hemiarthroplasty depends on multiple factors including comorbidity. The Swedish Hip Arthroplasty Register (SHAR) provides a unique opportunity to study mortality and revision rates in this population. Linkage with government databases allow for in-depth research into the factors that influence risk of revision surgery and death in the hip fracture patient. Patients and methods: Data was linked between SHAR, Statistics Sweden and the National Board of Health and Welfare. Data was collected on 38,912 patients who received a fracture-related hip arthroplasty between 2005 and 2012. A multistate analysis was performed and three states were identified: primary hip surgery and alive (state 1), revision after primary hip surgery (state 2) and death (state 3). These were marking points in the longitudinal outcome study. Results: 38,912 patients who received an arthroplasty for an acute hip fracture were included. By the end of the study period 1309 (3.4%) of these patients underwent a revision and 17,365 (45.1%) patients died. Patients with THA had a reduced risk of death from primary operation compared to hemiarthroplasty (HR = 0.49) and a decreased revision risk (HR = 0.69). Female patients had a statistically significant reduced mortality (HR = 0.6) compared to men. There was no statistically significant difference in risk of revision surgery between direct lateral and posterior approach. Conclusion: We identified an influence of type of surgery, sex, age and Elixhauser Comorbidity Index (ECI) on risk of revision and mortality. Males, greater comorbidity burden and older patients had higher mortality risks. The posterior approach did not have a significant influence on revision risk. Further research could include all patients who had reoperation(s) to further strengthen our findings. Patients who had a THA had lower revision rate and mortality. The latter is likely due to selection