In this study we demonstrated excellent reliability both within and between observers using a novel, 3D CT reconstruction-based digital templating method for assessing the total tibial coverage afforded by different prosthesis designs.
Difficulties in coding were resolved by discussion between the observers, using their notes made during the observation.
This shows that it is easier to have a high agreement between two observers using a less accurate technique [23, 28, 29].
This is only slightly lower than the ICC between different observers for either machine implying an "excellent" agreement, and that the variation in measurements between the machines was only slightly greater than the variation in measurements between two observers using an identical machine.
The kappa value for the inter-observer variability between the observers using the EAP checklist was 0.618, indicating substantial agreement between the observers [ 21].
Denoted here as percentage dense area (PDA), it can be measured reliably within and between trained observers using a computerized thresholding technique [ 2], in part a consequence of PDA's having a large variance even after adjusting its mean for age.
143 observation pairs would be sufficient to estimate an expected kappa of 0.4 and its 95% confidence interval (indicating fair agreement between different observers using the same measurement tool) for rating newborns as having feet shorter or longer than 8 cm.
Whilst operator experience is an important factor in edge delineation, the machine settings heavily influence this assessment and create a challenge for standardised reporting between observers using different equipment.
Inter-observer variability in percentage difference is generally low (less than 5%%) showing a good agreement between the different observers using the sVOI method for absorbed dose calculations.
Kappa statistics was used for assessing the agreement between observers using the NCSS 2007 statistical software (NCSS and GESS, NCSS, LLC. Kaysville, UT, USA).
