Conclusion
Our study showed that endosonographers and cytopathologists had good concordance in EUS-FNA specimens on-site cytopathology interpretation, except in mediastinal/pulmonary cases.
Methods
A prospective blinded study comparing one endossonographer with one physician cytopathologist. The study was developed in the Santa Casa Medical School, Brazil from February to November 2012. Fifteen different cases of EUS-FNA were analysed, in a total of 50 slides. Each observer described the slides for the adequate or not of tissue sampling, and classified as benign, suspicious, malign or undefined. The analyses were then matched.
Objective
To determine the interobserver concordance of on-site cytopathology interpretation of EUS-FNA specimens by comparing endosonographers trained in cytology with a physician cytopathologist.
Results
We analyzed the concordance of 50 slides description made by the endossonographer and physician cytopathologist, according to enough material, cellular group identification and final diagnosis. Kappa (κ) indexes were: Presence of material κ = 0.480 (P < 0.001); presence of malignance κ = 0.808 (P < 0.001); in subepithelial lesions κ = 0.615 (P = 0.06); in pancreatic lesions κ = 0.675 (P < 0.001); in mediastinal lesions κ = 0.243 (P = 0.128).