Validity of clinical measures of frontal plane knee alignment: data from the Osteoarthritis Initiative

Clinicians frequently assess frontal plane knee alignment of patients suspected of having knee osteoarthritis. However, there is little data to indicate whether simple to obtain clinical measures frontal plane alignment indicate the actual bony alignment of the knee or the extent and type of knee osteoarthritis. The purpose of this study was to assess the validity of goniometric measures of frontal plan knee position as a simple clinical tool to assess lower limb alignment and tibiofemoral OA status. Data from the Osteoarthritis Initiative were used to assess concurrent and diagnostic validity of goniometric measures. Data from 142 persons with symptomatic knee OA were used to examine concurrent validity, based on comparison to whole limb radiographs. Diagnostic validity was examined using data from 1390 persons with symptomatic knee OA, based on comparison to tibiofemoral joint space narrowing obtained from standardized radiographs. Associations between goniometric measures and whole limb radiographs were moderate (r = 0.43 for right knee and r = 0.46 for left knee). Goniometric measurements explained approximately 20% of the variation in lower limb alignment as measured by whole limb radiography suggesting that simple clinical measures of lower limb alignment contain too much error for inferring actual lower limb alignment. Diagnostic validity indicated positive likelihood ratios of 3.3 or less for detecting isolated severe medial or lateral joint space narrowing and a high rate of misclassification. Simple clinical assessments of frontal plane lower limb alignment or tibiofemoral OA status, in isolation, contain substantial error and likely do not inform clinical practice.