This study aims to evaluate interobserver agreement on visual analysis of technetium-99m mercaptoacetyltriglycine (99mTc-MAG3) renal tissue transit used for the evaluation of antenatal hydronephrosis.Materials and methods
Thirty-eight 99mTc-MAG3 diuretic renograms were retrospectively collected between 1 and 31 December 2015. The 1-min reframed images were presented to four nuclear medicine consultants and to two nuclear medicine residents, one in the first year of the training program and the others in their fourth and final year. These observers were asked to classify the radiotracer cortical transit (normal/delayed) based solely on visual assessment of the images. For the interobserver agreement, modified Fleiss’ kappa (κ) analysis for multiple raters was carried out. For both groups, percentages of agreement were also calculated.Results
A total of 69 kidneys were evaluated. All four nuclear medicine consultants agreed on the classification of 88.4% of the kidneys. When the agreement of at least three of the four observers was considered, the percentage of agreement reached 98.6%. The two nuclear medicine residents agreed on the classification of 69.6% of the kidneys. The modified Fleiss’ κ-value was 0.88 (95% confidence interval: 0.79–0.95) for the group of nuclear medicine consultants, indicating almost perfect agreement. For the residents, it was 0.39 (95% confidence interval: 0.16–0.59), suggesting fair agreement.Conclusion
Our results seem to indicate that there is an almost perfect agreement in the qualitative identification of delayed cortical transit among physicians with experience at observing renographic images.