Errors in visual estimation of flexion contractures during total knee arthroplasty

To quantify and reduce the errors in visual estimation of knee flexion contractures during total knee arthroplasty (TKA).

This study was divided into two parts: Quantification of error and reduction of error. To quantify error, 3 orthopedic surgeons visually estimated preoperative knee flexion contractures from lateral digital images of 23 patients prior to and after surgical draping. A repeated-measure analysis of variance was used to compare the estimated angles prior to and following the placement of the surgical drapes with the true knee angle measured with a long-arm goniometer. In an effort to reduce the error of visual estimation, a dual set of inclinometers was developed to improve intra-operative measurement of knee flexion contracture during TKA. A single surgeon performed 6 knee extension measurements with the device during 146 consecutive TKA cases. Three measurements were taken with the desired tibial liner trial thickness, and 3 were taken with a trial that was 2 mm thicker. An intraclass correlation coefficient (ICC) was calculated to assess the test-retest reliability for the 3 measurements taken with the desired liner thickness, and a paired t test was used to determine if the knee extension measurements differed when a thicker tibial trial liner was placed.

The surgeons significantly overestimated flexion contractures in 23 TKAs prior to draping and significantly underestimated the contractures after draping (actual knee angle = 6.1° ± 6.4°, pre-drape estimate = 6.9° ± 6.8°, post-drape estimate = 4.3° ± 6.1°, P = 0.003). Following the development and application of the measurement devices, the measurements were highly reliable (ICC = 0.98), and the device indicated that 2.7° ± 2.2° of knee extension was lost with the insertion of a 2 mm thicker tibial liner. The device failed to detect a difference in knee extension angle with the insertion of the 2 mm thicker liner in 9/146 cases (6.2%).

We determined the amount of error associated with visual estimation of knee flexion contractures, and developed a simple, reliable device and method to improve feedback related to sagittal alignment during TKA.