Despite improvements in component design and surgical technique, some patients still require lateral retinacular release during TKA to improve patella tracking. We studied 148 fixed-bearing TKAs to identify parameters in pre-operative knee radiographs that would predict intraoperative patellar maltracking. Digital radiographs and software were used to measure coronal alignment, distal femoral valgus angle, proximal tibia varus angle, patellar tilt, patellar shift, Insall-Salvati ratio, and patellar component placement and alignment. Patellar tracking was assessed after all components had been cemented, using both no-touch and modified "towel clip" techniques. The only radiographic parameter independently associated with maltracking was patellar shift. The median pre-operative patellar lateral shift in patients who had maltracking was 4.1 mm compared to 0.0 mm in those who did not. Patients who had a patellar shift of more than 3.0 mm had a high likelihood of maltracking, with estimated positive and negative predictive values of 78 and 95%, respectively. Pre-operative patellar shift may thus be clinically relevant for identifying osteoarthritic patients who have a higher likelihood for patellar maltracking during TKA. Variations in the intrinsic risk for maltracking within patient study populations may account for the widely differing reported rates of patellar maltracking, and our data suggest that information on pre-operative patellar shift may be helpful in stratifying these sample populations.

No abstract