Patellar tilt and subluxation in total knee arthroplasty. Relationship to pain, fixation, and design.

Two hundred thirty-four primary total knee prostheses were evaluated with a 45 degrees merchant view to examine patellar position and fixation. Despite a standardized technique for obtaining roentgenographs, patellar position varied with leg position. Overall, 54.7% tracked centrally, 31.2% tilted, and 14.5% displaced. The incidence of these findings was the same whether the patella was domed or congruent. Neither pain scores nor fixation were affected by position. Postoperative tilt and displacement were more common in patellae that were tilted preoperatively. The incidence of postoperative tilt or displacement was not significantly different in knees in which a lateral release was performed. Patellar tilt in some patients seems inevitable despite careful technique because intraoperative tests are static and postoperative function is dynamic. This does not bode well for wear in a metal-backed patella. Tilt caused increased loading at the periphery of the component where most metal-backed prostheses have thin polyethylene.