Total hip replacement for neck of femur fracture: comparing outcomes with matched elective cohort

Introduction: Current literature suggests that total hip replacement (THR) is superior to hemiarthroplasty (HA) for neck of femur fracture in selected group of patients. The outcomes of THR undertaken for trauma setting remain unclear when comparing with elective THR. We compared the outcomes of THR trauma cohort with best-matched elective cohort.

Methods: We retrospectively reviewed 102 patients that underwent THR due to trauma from 2011 to 2013. We had access to 90 cases with complete records. Another 90 matched elective cases were obtained from local arthroplasty database. The elective cases were matched for gender, surgical approaches, surgeon's grade, types of implant, patient's age at operation date of ±5 years and operation date of ±60days. Subsequently, the selection criteria were relaxed to patient's age at operation date of ±10 years and operation date of ±60days. Unmatched cases were excluded. Complications and death rate were compared.

Results: The average age for both cohorts was 70 years. The trauma cohort had statistically significant lower BMI and longer hospital stay (p=0.001). The Functional Comorbidity Index (FCI) and Charlson Age Comorbidity Index (CACI) were the same for both cohorts, reflecting an active patient selection for THR in our centre. The trauma cohort had higher surgical complication rate (9% vs 4%), particularly higher dislocation rate (7% vs 1%); and higher medical complication rate (32% vs 6%). These were consistent with the literature. Contrary to literature, the trauma cohort had six dislocations that five of them were done via anterolateral approach. Among the eight trauma cases with surgical complications, six cases were performed by trainees. The cause of surgical complications remains unclear due to the nature of retrospective study. The trauma cohort had higher death rate than the elective cohort (14% vs 4%), with one post-operative cardiac arrest in the trauma cohort. The rest were non-orthopaedic related deaths, ranging between four months to four years.

Conclusion: A more robust way of selecting trauma patients for THR is warranted to reduce morbidity and mortality. Follow-up for the trauma cohort is warranted, as the patients are likely to outlive the implants.