Standard Radiographs Are Not Sufficient for Assessing Knee Mechanical Axis in Patients With Advanced Osteoarthritis.

Abstract

The use of standard radiographs, and measured tibiofemoral angle (TFA), to assess lower extremity alignment is commonly practiced despite limited knowledge of its relationship to the mechanical axis (MA), as measured on hip-to-ankle (HTA) radiographs. This study assessed the predictive accuracy of previously developed equations, developed gender-specific regression equations using predictors from standard radiographs, and the clinical effectiveness of these equations in a large sample of cases using HTA radiographs as a gold standard. The MA was measured on HTA radiographs, whereas TFA and femoral angle were measured on standard radiographs in 788 cases diagnosed with knee osteoarthritis. Multiple regression analyses indicated that TFA, femoral angle, and height were the strongest factors associated with the predicting MA, accounting for 83% of the variance for men and 86% for women, but were able to predict only the actual MA within ±3° in 66% of men and 69% of women. When applied to previously reported regression equations with similar results, the best predicative accuracy obtained within ±3° was 61% and 63% of men and women, respectively. Standard radiographs are not sufficient for determining MA, and HTA radiographs should be used while making surgical decisions aimed at correcting alignment to within ±3° or for assessing alignment post-total knee arthroplasty. In addition, surgical alignment outcomes reported in previous research using standard radiographs should be viewed with caution.