The reliability and validity of physiotherapist visual rating of dynamic pelvis and knee alignment in young athletes

Purpose: To investigate visual rating of pelvis and knee position in young athletes during lower extremity functional tests. Methods: Pelvis and knee alignment, in 23 athletes, was visually rated by 66 physiotherapists. Peak two-dimensional (2D) and three-dimensional (3D) kinematics were also quantified. Ratings were compared to consensus visual ratings of an expert panel. The consensus ratings were also compared to peak kinematics. Reliability was determined using percentage agreement (PA) and the first order agreement coefficient (AC1). Sensitivity, specificity, diagnostic odds ratio (DOR) and differences in kinematics between groups based on the expert visual ratings were calculated to assess rating validity. Results: Mean intra-rater agreement was substantial (PA: 79-88%, AC1: 0.60-0.78). Inter-rater agreement ranged from fair to substantial (PA: 67-80%; AC1: 0.37-0.61). Sensitivity (≥80%) and specificity (≥50%) were acceptable for all tests except the Drop Jump. Experience (DOR 1.6-2.8 times better) and slower movement (4.9 times better) improved rating accuracy. Peak 3D and 2D kinematics were different between groups rated as having good versus poor alignment by the experts. Conclusions: Visual rating by physiotherapists is a valid tool for identifying young athletes with poor frontal plane dynamic pelvis and knee alignment.