Trochanteric prominence angle test (TPAT)! What is it good for?

Previous work has identified the trochanteric prominence test (TPAT) as an unreliable method of assessing femoral torsion [1]. However, despite its poor performance, many laboratories and clinics continue to utilize the TPAT to assess femoral torsion. We hypothesize that difficulty in rationalizing the previously demonstrated TPAT error magnitudes has caused many centers to underestimate the poor performance of the TPAT, lead to continued use of the TPAT as a screening tool for diagnosing femoral malrotation. How much error can be expected from the TPAT when including variability in the anatomy of the proximal femur, variability in the anatomy of the proximal tibia, and variability in assessor hand placement? Simulated data for 100,000 subjects aged 4-20 was generated using the following criteria: Femur length -> Sample femur lengths from a pool of age matched peers [N = 4,416] Femoral neck length -> Scale based on femur length [Mean: 8.5% of femur length; SD: 2%] Femoral neck-shaft angle -> Natural progression and variability [Mean: 145 deg at age 3, mean 120 deg at age 20; SD: 15 deg] Trochanteric prominence angle -> Mean value and variability from [1] [Mean: 27 deg internally rotated about the long axis of the femur; SD: 12 deg] Proximal tibial varus/valgus -> Natural variability [Mean: 0 deg; SD: 2 deg] Observer hand placement -> Scale based on subject size [Mean: 0; SD: 2% of femur length] Errors due to variability of the proximal femur, variability of the proximal tibia, and variability in observer hand placement were calculated individually. Total error was calculated as the sum of all the individual for each simulated subject. Errors are reported as the 5th to 95th percentile of simulated values. Overall, the simulated TPAT resulted in large errors that decreased somewhat with age (Fig. 1). Across all ages, the total error was 12 ± 15 deg (mean ± SD). Observer hand placement contributed to the majority of errors (60 deg range). Variability at the proximal femur contributed 30-50% of the total error (~40 deg range), while variability of the tibia contributed the least (7 deg range). Download : Download high-res image (95KB)Download : Download full-size image Given the large error potential due to the intrinsic variability of the proximal femur, the TPAT does not appear to be able to provide reliable assessments of femoral torsion even if hand placement errors were eliminated. As currently performed, TPAT errors appear to be too large to be used for routine clinical purposes.