Purpose: Metal-on-Metal (MOM) hip resurfacing is a popular alternative to conventional hip arthroplasty. The purpose of the present study is to compare patient characteristics and radiographic findings for revision versus non-revision cases treated at a single center with a MOM resurfacing device.

Method: Eighty-nine MOM resurfacing arthroplasties were performed between December 2001 and June 2006. Subjects were assessed for implant revision status, age, gender, weight, operative side, primary diagnosis, femoral head size, and time to revision. Postoperative radiographs were assessed for acetabular cup inclination and femoral stem inclination relative to the femoral neck.

Results: Thirteen of 89 hips (14.6%) have required revision to date with follow-up ranging from 4 (a revision) to 91 months. Female gender, smaller implant size, and a diagnosis of osteonecrosis were all associated with lower device survival. A significant difference in acetabular cup angle was observed between revised and non-revised hips. However, this difference can largely be attributed to two outliers in cup position associated with early migration and the difference became non-significant when these outliers were excluded. No significant difference was found in stem angle between revised and non-revised hips. The revision rate for the first 25 hips was 24% versus 8% for the next 64 hips. Females accounted for 56% of subjects 1–25 and 23% of subjects 26–89. Seven of the 8 failures for the first 25 procedures occurred beyond 4 years follow-up. Only 2 early device failures (< 2 yrs) were identified and both were in the 26–89 subgroup.

Conclusion: Despite representing only 33% of included subjects, females accounted for 62% of revision procedures. An apparent learning curve was identified by a lower device survival proportion for the first 25 hips versus hips 26–89. However, this learning curve was not associated with failures which could be attributed to acetabular or femoral component malpositioning and is likely explained by the higher proportion of females enrolled early in the study. Taken together, we propose the apparent “learning curve” exhibited in our study is not wholly technical in nature but rather influenced by changes in patient selection over time by the operative surgeon.