Estudo de reprodutibilidade inter-observador de células glandulares atípicas e interpretações

Abstract

Introduction: The diagnosis of Atypia in glandular cells (AGC) using the conventional method due to the difficulty of cytomorphological interpretation and the frequent overlap with squamous lesions, has been a challenge for Cytopathologists. Methodology: 180 conventional Pap smears with a previous diagnosis of endocervical glandular atypia were subsequently reevaluated by two cytopathologists. Kappa's correlation index was applied to analyze the agreement of the different results. Objective: To evaluate the the reproducibility of the diagnosis of AGC by the Papanicolaou method through the conventional methodology. Results: The reproducibility of the first analysis when compared with the posterior analysis of the first cytopathologist showed correlation of k=0.53 (IC95%: 0.41 to 0.65) p< 0.01. With the posterior result of the second cytopathologist the correlation was k= 0.11 (IC 95%: 0.05 to 0.17) p<0.01. Classifying the results as benign and malignant the correlation found between the first analysis and the results of the first cytopathologist was k=0.54 (95%CI: 0.32 to 0.76) p< 0.01, as the second cytopathologist the correlation found was k=0.38 (95%CI: 0.11 to 0.65) p< 0.01. Conclusion: The varied correlation between the different analyses both for the issued diagnoses of glandular atypia (AGC-SOE, AGC-NEO, adenocarcinoma in situ and invasive adenocarcinoma) and in the classification of results considered benign and malignant demonstrates that the reproducibility of diagnosis for GCA suffers from reproducibility difficulty in recognizing the morphological findings that hinders the observer's interpretation.