Assessment of knee alignment with varus and valgus force through the range of flexion with non-invasive navigation.

In image-free total knee arthroplasty (TKA) navigation, infra-red markers are attached to bony landmarks to provide kinematic data intra-operatively, with the aim of improving the precision of implant placement. In non-invasive navigation, infra-red markers are attached to the skin surface, with recent evidence suggesting that this can give repeatable measurements of lower limb mechanical alignment. The aim of our study was to evaluate the use of a non-invasive navigation system in the assessment of mechanical alignment with applied coronal force through the range of flexion. A previously validated non-invasive system (Physiopilot™) was tested on 23 volunteers with healthy knees. Two users performed two registrations of the software workflow on each participant's right and left knees. A force was manually applied to the end-point of varus and valgus knee laxity and the measured change in mechanical alignment was recorded. Force was applied with the knee positioned in increments of flexion from 0 to 90°. In keeping with previous studies, satisfactory values of coefficient of repeatability (CR) of 1.55 and 1.33 were found for intra-observer repeatability in measurement of supine mechanical femoro-tibial angle (MFTA) in extension, with a good inter-observer correlation of intraclass correlation coefficient (ICC) .72. However, when flexion was introduced, intra-observer and inter-observer reliability fell out with acceptable limits. Therefore, the trial did not support use of the Physiopilot™ system as a measure of MFTA when flexion is introduced. It was felt that learning-curve, soft tissue artefacts and lack of force standardisation equipment may have accounted for significant levels of error, with further studies required to address these issues.