Purpose of the study: An unstable hip prosthesis is a therapeutic challenge. The prevalence of revision is 5 to 26.6% in the literature. We evaluated the contribution of double-mobility implants for revisions of unstable hip implants.

Material and methods: This series was composed of 45 patients who underwent revision between January 2000 and December 2003 for hip instability (44 dislocations, 1 subluxation). The same implant was used for all patients, either for the first-intention version (press-fit or cemented), or for the revision version (press-fit). For certain patients, the first-intention implant was cemented in an armature. The series included 28 females and 17 males, mean age 66.5 years (range 36–48 years). The initial diagnosis was osteoarthritis in 34 cases (76%), dysplasia in seven (16%), osteonecrosis in two (4%), Paget’s disease in one (2%) and rheumatoid disease in one (2%). The patients had had 2.8 dislocations on average (range 1 – 10). Time from first dislocation to the first-intention operation was 45.6 months (range 15 days – 20 years). Mean time from the first-intention operation to revision was 64.3 months (range 3 weeks – 20 years). Risk factors for instability were repeated hip surgery (> 3 operations) for 13 patients, wear for seven, nonunion of the greater trochanter for five, neurological and cognitive impairment in five, and malposition in three.

Results: Mean follow-up was 25.2 months. None of the patients were lost to follow-up. Two patients died late after the operation. Among the complications observed, there were: two cases of recurrent dislocation, one case of subluxation, two cases of infection (one with favorable outcome after surgical cleaning and antibiotics the other followed by patient death), two cases of deep vein thrombosis, one case of popliteal paresia with favorable outcome, one case of delirium tremens. Surveillance was the therapeutic option for the patient with subluxation. For patients with dislocation, revision surgery was performed using the same implant. For one of these patients, the dislocation occurred following early loosening.

Conclusion: Use of double-mobility implants for prosthetic revision undertaken because of prosthesis instability provides encouraging results, with a rate of dislocation (4%) close to that observed with first-intention implants.

Purpose: Hip joint involvement is a frequent complication of Paget’s disease. We conducted a multicentric retrospective study to analyse perioperative problems and outcome after total hip arthroplasty in patients with Paget’s disease.

Material and methods: Thirty-nine total hip arthroplasties were implanted between 1979 and 1998 in 35 patients with Paget’s disease of the hip (four bilateral cases). The series included 20 men and 15 women, mean age 74 years (55–86). The pre and postoperative status was evaluated with the Harris score and radiographically. We recorded operative time, blood loss, and events noted in the operative report. Among the 35 patients, 24 were retained for analysis (three deaths, eight lost to follow-up) at mean 62 months.

Results: The mean preoperative Harris score was 46/100 (18–67). Eighteen patients had leg length discrepancy. Nine had permanent hip flexion associated with external rotation and seven had coxa vara. Twenty-nine patients were given anti-osteoclastic treatment preoperatively. We implanted 20 cemented cups and 19 press-fit cups. Thirty-one femoral stems were cemented and eight were not. On average, operative time was 130 minutes and blood loss was 830 cc. Difficult operative events involved luxation of the femoral head, remodelled sclerous bone (greater trochanter fractures, difficult reaming, narrow canal), cam effect related to bone hypertrophy and bleeding. Venous thrombosis occurred in four patients, pulmonary embolism in one, and one psoas haematoma. There were three luxations. At last follow-up (mean 71 months), the clinical outcome was excellent for 13 patients (48%), good for eight (29%), fair for two (7%) and poor for four (15%). Implants were cemented at the pelvis and femur level in three cases and noncemented in one.

Discussion: Prosthetic hip surgery in patients with Paget’s disease is difficult and raises the risk of postoperative complications. Prior medical treatment is needed before surgery to limit the risk of bleeding. In our series, fixation modalities were very variable. Non-cemented implants on Paget’s diseased bone performed comparably with non-cemented implants.

Purpose: There has only been one reported series of 30 cases of greater trochanter fracture during total hip arthroplasty and 26 of these were postoperative discoveries. We evaluated the frequency of this event and its postoperative consequences.

Material and methods: Among our series of 1171 total hip arthroplasties performed between 1985 and 2000, 38 patients (3.2%) with greater trochanter fracture were identified (mean age 63 years). Osteosynthesis was performed in all cases. Thirty-one fractures were observed during primary arthroplasty and seven during revision procedures.

Results: Eighteen patients had a favouring condition: corti-costeroid therapy,alcoholism,osteoporosis,diabetes,Paget, ablation of trochanteric material, periprosthetic osteolysis. The anterolateral approach was used in 22 and the posterolateral approach in 16. The fracture occurred along the access route in four (material removal or prefracture situtation), at removal of a previously implanted stem in two, and during implantation in 32. Twelve different stems were involved but a screwed stem was involved in 18 cases, i.e. 10% of all implanted screwed stems, while this complication only occurred in 1.2% of other implanted stems. Immediate weight bearing was authorised in 27 patients and deferred three weeks to three months in eleven. There were two deaths, so follow-up data was available for 36 hips: we observed anatomic bone healing in 22, deformed calluses in five and nonunion in nine, including two cases with infection (three revision procedures were required). Pain persisted at two months for eleven hips and limping persisted for ten (eight nonunions).

Discussion: Prostheses with a large metaphyseal component were involved in the majority of the fractures. The surgical approach was not incriminated. When well stabilised, trochanter fractures healed well. Nonunion, often announced by persistent pain, is an important risk in patients with osteoporosis and a poorly stabilised fracture. Although all cases of nonunion were observed in patients with deferred weight bearing, this criterion is simply the expression of the surgeon’s apprehension in case of less than satisfactory fixation.

Conclusion: Because of the deficient bone stock, which explains the higher rate of nonunion, fracture of the greater trochanter cannot be considered in the same light as a planned osteotomy. Prevention requires choosing a less cumbersome metaphyseal component in patients with favouring conditions. Osteosynthesis must be performed with particular care in order to obtain rapid healing and good functional outcome.