A prospective randomized study of a collar versus a collarless femoral prosthesis.

A prospective, randomized study of 44 collared versus 40 collarless cemented HD-2 total hip arthroplasties (THAs) was performed in 84 patients. Seventy patients were examined at an average of 4.6 years after surgery. Five hips have required, or are scheduled to have, revision arthroplasty: two with collared prostheses and three with collarless prostheses. Good collar contact with the proximal medial femur was achieved in 47% of the collared prostheses. No patient with good contact has required or is scheduled to have revision hip surgery. Radiolucent lines were greater in both frequency and width in Gruen Zones 2 and 7 in patients with a collarless prosthesis. Loss of endosteal height of the femoral neck was 3.5 mm in patients with collarless prostheses and 1.7 mm in patients with collared prostheses. Loss in height of the femoral neck was 2.6 mm in patients with collared prostheses with poor bone contact and 0.7 mm in patients with collared prostheses with good bone contact. The mean preoperative and postoperative Harris hip scores for the collarless group and for the collared group were not statistically different. There were no statistically significant differences in acetabular position, the incidence of acetabular radiolucent lines, or femoral subsidence between patients treated with collared or collarless prostheses. This study demonstrates that it is possible to stress the proximal medial portion of the femur with a collared prosthesis. Furthermore, a significant increase is identified in both the frequency and width of radiolucent lines in patients with collarless prostheses in Zones 2 and 7.(ABSTRACT TRUNCATED AT 250 WORDS)