Propagation of registration error into maximum total point motion to analyze tibial baseplate stability at six months using marker-based and model-based RSA.

Maximum total point motion (MTPM) of a tibial baseplate at 6 months is used to predict long-term aseptic loosening after total knee arthroplasty. However, the propagation of registration error into MTPM for stable baseplates (i.e. baseplates with MTPM < 0.5 mm) manifested as bias (i.e. systematic error) and precision (i.e. random error) has not been quantified and compared to the 6-month stability limit for marker-based and model-based RSA, which have different magnitudes of registration error. To determine the bias and precision in MTPM for stable baseplates, registration errors in six degrees of freedom reported in the literature for marker-based and model-based RSA were applied to an example baseplate using computer simulations. Results revealed that the bias in MTPM for stable baseplates with model-based RSA is three to four times that of marker-based RSA, and that the precision in MTPM for stable baseplates with model-based RSA is double that of marker-based RSA. This assessment of bias and precision in MTPM for stable baseplates led to a method for adjusting the 6-month stability limit for model-based RSA where half the width of the 95% confidence interval on the mean MTPM and the bias in MTPM for marker-based RSA were subtracted from 0.5 mm to compute true MTPM. The bias in MTPM and half the width of the 95% confidence interval on the mean MTPM for model-based RSA were then added to the true MTPM to obtain the adjusted stability limit for model-based RSA which ranged from 0.57 mm to 0.64 mm.