Reports of high dislocation rates after revision total hip arthroplasty (THA) have encouraged the widespread use of dual-mobility cups. Dislocation has been less common but not fully abolished with dual-mobility cups, and its causes have remained unidentified. The objectives of this retrospective matched case-control study were: 1) to identify risk factors for dislocation, 2) and to assess dislocation outcomes.

The causes of dislocation after revision THA with a dual-mobility cup can be identified.

Among 653 consecutive patients identified retrospectively as having undergone revision THA with a dual-mobility cup (Medial Cup, Aston, Saint-Étienne, France) between January 2007 and December 2017, 16 (2.45%) subsequently experienced dislocation, after a mean of 3.6 months (range, 0.9–19 months). For each of these 16 patients, we collected the main patient characteristics (age, sex, body mass index [BMI], ASA score, and reason for the initial arthroplasty procedure), local history (number of previous surgical procedures, reason for revision, femoral and acetabular bony defects classified according to Paprosky, and status of the abductor apparatus), and characteristics of the revision (approach, diameters of the cup and femoral head, cup inclination, femoral offset, lower limb length, and implant anteversion). Controls were patients without dislocation after revision dual-mobility THA. Each of the 16 patients was matched to 3 controls on age (±10 years), sex, year of revision, and whether revision was only acetabular or bipolar. Univariate and multivariate analyses were done to compare the cases and controls, and dislocation outcomes in the cases were evaluated.

By univariate analyses, factors associated with dislocation were BMI>30 (cases, 37.5%; controls, 10.4%; p=0.02), larger number of previous surgical procedures on the same hip (cases, 2.8; controls, 1.8; p=0.004), larger number of arthroplasties (cases, 2.3; controls, 1.5; p=0.004), history of instability (cases, 31% with chronic dislocation and 13% with recurrent dislocation; controls, 6.25% and 2.1%, respectively; p=0.004), and compromised abductor apparatus (cases, 56.25%; controls, 14.6%; p=0.002). Independent risk factors for dislocation identified by multivariate analysis were instability (odds ratio [OR], 14.5; 95% confidence interval [95%CI], 1.5–149.9) and, most importantly, abductor apparatus compromise (OR, 43.1; 95%CI, 3.18–586.3). Of the 16 cases, 1 was lost to follow-up, 1 had contra-indications to anaesthesia, 1 died after several dislocation episodes, and 1 died 3 months after surgical reduction. In 5 cases, there was a single dislocation episode. Further surgery was performed in 8 cases (surgical reduction, n=1; constrained cup, n=3; trochanteric internal fixation, n=1; exchange of the dual-mobility cup, n=2, including 1 with subsequent dislocation episodes; and femoral component exchange followed by a retentive cup due to further dislocation episodes, n=1).

Risk factors for dislocation consisted of a history of instability and, most importantly, abductor apparatus compromise. A constrained cup should be considererd in patients with impaired abductor apparatus. No further dislocations occurred after reduction of the first episode in 31.25% of cases. Recurrent dislocation should prompt measures to correct impaired abductor apparatus whenever possible, as well as correction of any component malposition. Whether a retentive cup should be implanted routinely remains unclear.

III, matched case-control study.