Periprosthetic fractures have become one of the main reasons for revision of total hip arthroplasty, with the treatment strategy guided by the Vancouver and SOFCOT classification systems. The main objective of this work was to compare the functional outcomes and mortality rate in patients who suffered a periprosthetic hip fracture with femoral stem loosening and who were treated either by internal fixation with locking plate only, or in combination with revision of the existing femoral stem with a cemented stem. The secondary objective was to compare the blood loss, procedure duration, length of hospital stay, living arrangements, time to weight bearing and major complications between these two treatment strategies. The starting hypothesis was that there is no difference in the functional outcomes between these two groups.

This was a single center, retrospective study of patients treated between 2013 and 2019. Thirty-six patients (54%) underwent femoral stem revision (FSR group) and 31 patients (46%) were treated by locking plate fixation only (ORIF group). In this population, 19 of the patients had been residing at a nursing home before the surgical treatment: 12 (33%) in the FSR group and 7 (23%) in the ORIF group (p=0.33). The mean preoperative Parker score was 7.03 (± 1.48) in the FSR group versus 5.19 (± 1.92) in the ORIF group. The functional outcomes at the final review were evaluated using the Parker score, HHS (Harris Hip score), WOMAC (Western Ontario and McMaster Arthritis Index) and Oxford score. Fracture healing was determined on radiographs centered over the hip joint with the entire femur visible on AP and lateral views. All patients were reviewed at a minimum of 1 year for clinical and radiological assessments.

At a mean follow-up of 28 months [12–85], the Oxford score was 40.4 (± 10.6) in the FSR group versus 37.1 (± 13.5) in the ORIF group (p=0.34), WOMAC of 45.5 (± 26.9) versus 50.5 (± 32.4) (p=0.56), and HHS of 62.7 (± 25.5) versus 60.3 (± 28.6) (p=0.79). The Parker score at the final assessment was 5.14 (± 1.96) in the FSR group versus 4.64 (± 2.17) in the ORIF group (p=0.33). In the FSR group, 10 patients were at a nursing home (34%), 6 of whom were not there initially, versus 6 in the ORIF group (43%), 5 of whom were not there initially (p=0.59). The mean operating time was 152±41.4 minutes (minimum: 85; maximum: 254) in the FSR group versus 89.2 (± 44.1) minutes (minimum: 39; maximum: 280) in the ORIF group (p<0.001). Resumption of full weight bearing occurred at 6 weeks in 21 patients (58%) in the FSR group versus 14 patients (47%) in the ORIF group (p=0.023). There were 5 major complications (14%) in the FSR group versus 7 (23%) in the ORIF group (p=0.35). Among the reviewed patients, the union rate was 86% on the latest radiographs in the FSR group versus 78% in the ORIF group.

There was no statistically significant difference between groups in the various functional scores, confirming our study hypothesis. Internal fixation with a locking plate appears to be an acceptable solution for treating periprosthetic hip fractures in which the femoral stem has loosened.

III, comparative retrospective study.