Evaluating Alternate Registration Planes for Imageless, Computer-Assisted Navigation During Total Hip Arthroplasty

Imageless computer navigation improves component placement accuracy in total hip arthroplasty (THA), but variations in the registration process are known to impact final accuracy measurements. We sought to evaluate the registration accuracy of an imageless navigation device during THA performed in the lateral decubitus position.

A prospective, observational study of 94 patients undergoing a primary THA with imageless navigation assistance was conducted. Patient position was registered using 4 planes of reference: the patient's coronal plane (standard method), the long axis of the surgical table (longitudinal plane), the lumbosacral spine (lumbosacral plane), and the plane intersecting the greater trochanter and glenoid fossa (hip-shoulder plane). Navigation measurements of cup position for each plane were compared to measurements from postoperative radiographs.

Mean inclination from radiographs (41.5° ± 5.6°) did not differ significantly from inclination using the coronal plane (40.9° ± 3.9°, P = .39), the hip-shoulder plane (42.4° ± 4.7°, P = .26), or the longitudinal plane (41.2° ± 4.3°, P = .66). Inclination measured using the lumbosacral plane (45.8° ± 4.3°) differed significantly from radiographic measurements (P < .0001). Anteversion measured from radiographs (mean: 26.1° ± 5.4°) did not differ significantly from the hip-shoulder plane (26.6° ± 5.2°, P = .50). All other planes differed significantly from radiographs: coronal (22.6° ± 6.8°, P = .001), lumbosacral (32.5° ± 6.4°, P < .0001), and longitudinal (23.7° ± 5.2°, P < .0001).

Patient registration using any plane approximating the long axis of the body provided a frame of reference that accurately measured intraoperative cup position. Registration using a plane approximating the hip-shoulder axis, however, provided the most accurate and consistent measurement of acetabular component position.