Reliability of two different measuring techniques with computer tomography for penetration and distribution of cement in the proximal tibia after total knee arthroplasty

To evaluate the reliability of two different techniques for measuring penetration and distribution of the cement mantle in the proximal tibia after total knee arthroplasty (TKA) with Computer Tomography (CT) in vivo. Standardized CT scans of the proximal tibia were taken 1 to 2 years after total knee arthroplasties implanted with a surface cementing technique. These prospectively acquired transversal CT images of the surface of the proximal tibia were divided into four quadrants and were assessed once manually and once with a numerical computing program (MATLAB® Update 2, The MathWorks, Inc.) based on Hounsfield Units by one of the researchers. The assessments were repeated by the same and a second researcher. The ratio cement/trabecular bone was calculated 1, 3 and 5 mm distal of the tibia tray per quadrant. Kruskall-Wallis tests with multiple pairwise comparisons (Dunn's test) were used to determine differences between the quadrants. Intra- and inter-rater reliability as well as the inter method reliability were assessed with the Intraclass Correlation Coefficient (ICC) per level of depth and with Bland-Altman plots. A total of 92 CT scans were included. The intra- and inter-rater reliability of the manual method ranged from 0.22 and 0.52. The intra- and inter-rater reliability of the matlab method varied between 0.98 to 0.99. The median percentage cement measured with the matlab method 1 mm underneath the tibial tray varied between 82 and 88%; at 3 mm depth between 38 and 54% and at 5 mm between 15 and 25%. There was significantly (p