Eight-year outcomes from the WHiTE3 HEMI trial : comparing the Thompson hemiarthroplasty with the Exeter polished tapered stem and Unitrax modular head in the treatment of displaced intracapsular hip fractures at a single centre.

Aims: The WHiTE3:HEMI trial recruited patients with an intracapsular hip fracture to be randomized to receive the Thompson hemiarthroplasty or the Exeter stem and Unitrax modular head. Results demonstrated no difference in mortality or health-related quality of life at four months. The aim of this study was to review outcomes at eight years for patients recruited into the WHiTE3:HEMI trial at our institution. Methods: Dates of readmission, reoperation, and death were recorded from our local Trust universal clinical system for clinical noting, results, and radiology imaging. These data are concordant with the NHS Spine Personal Demographic Service and those submitted to the National Hip Fracture Database. Patient characteristics, incidence of periprosthetic fracture (PPF), reoperation, revision, and mortality were compared between the two groups. Patients were analyzed 'as randomized'. Results: In total, 135 Thompson patients were compared with 143 Exeter Unitrax patients. There were no differences in age, sex, American Society of Anesthesiologists grade, Abbreviated Mental Test Score, Nottingham Hip Fracture Score, grade of operating surgeon, or residence on admission. In the Thompson group, there were 13/135 (9.6%) surgical complications, 6/135 (4.4%) reoperations, and 1/135 (0.7%) PPF. In the Exeter Unitrax group, there were 10/143 (7.0%) surgical complications (p = 0.425), 5/143 (3.5%) reoperations (p = 0.685), and 2/143 (1.4%) PPFs (p = 0.596). Eight-year Thompson cumulative implant survivorship was 96.9% (95% CI 90.4 to 99.0) and Exeter Unitrax 96.5% (95% CI 90.6 to 98.7) (p = 0.954). There were no differences in 30-day or four-month mortality rates. One-year mortality rates were 37.0% in Thompson and 20.3% in Unitrax (difference 16.7%, 95% CI 1.1 to 31.8, p = 0.002). Pooled mortality at eight years was 86.3% with no difference between the two groups (p = 0.612). Conclusion: At one year, a significantly higher mortality rate was noted in the Thompson group. At long-term follow-up, Thompson and Exeter Unitrax demonstrated equivalent and low reoperation rates and excellent implant survivorship. Further studies may look to validate these findings.