[Validity of cervicovaginal cytology in a Brazilian State with high incidence rate of cervical cancer]. / Acurácia dos exames citológicos cervicovaginais em Estado de elevada incidência de câncer de colo de útero.

PURPOSE: To evaluate the validity of cervicovaginal cytology performed at LAPER, the main Laboratory of Pathology of the State of Roraima, Brazil, by interrater agreement (external monitoring) and agreement with histopathologic results. METHODS: One hundred women were included, a population-based convenience sample. Their cervical cytological exams were evaluated by the laboratory staff and reviewed by expert medical pathologists, external to the laboratory. Cohen's Kappa index, sensitivity and specificity were evaluated. The study was approved by LAPER coordination and Federal University of Roraima Research Ethics Commitee. RESULTS: Regarding the prevalence of human papillomavirus-related atypical, there was no concordance between the results issued by LAPER and by the external pathologists (k=0.21). A low sensitivity (28.5%) and specificity (89,2%) was detected for the diagnostic performance of LAPER, with a high proportion of false positive and false negative results. The cytological reports of the external pathologists showed higher sensitivity and specificity (71.4 and 98.9%, respectively), ruling out the possibility that errors related to collection methods and staining would explain the low performance of the laboratory. CONCLUSION: A low diagnostic accuracy of cervicovaginal cytology can be a barrier against the control of cervical cancer in Roraima. We emphasize the need for professional training and internal and external monitoring in Brazilian states with a high incidence of cervical cancer.

Acurácia dos exames citológicos cervicovaginais em Estado de elevada incidência de câncer de colo de útero / Validity of cervicovaginal cytology in a Brazilian State with high incidence rate of cervical cancer

OBJETIVO: Avaliar a acurácia da citologia cervicovaginal realizada no principal laboratório do Estado de Roraima, por concordância entre avaliadores (monitoramento externo) e com resultados histopatológicos. MÉTODOS: Lâminas de 100 mulheres, de uma amostra de conveniência de base populacional, foram avaliadas pelo LAPER, principal laboratório de referência do Sistema Único de Saúde do Estado de Roraima e reavaliadas por médicos patologistas experts externos ao laboratório. Foram avaliados índice Kappa, sensibilidade e especificidade. O estudo foi aprovado pela Coordenação do Laboratório Estadual e pelo Comitê de Ética em Pesquisa da Universidade Federal de Roraima. RESULTADOS: Quanto à prevalência de atipias celulares (relacionadas ao papilomavírus humano), não houve concordância entre resultados emitidos pelo LAPER e pelos patologistas externos (k=0,21). Constatou-se baixa sensibilidade (28,5%) e especificidade de 89,2% para o desempenho do LAPER, com elevada proporção de resultados falsos positivos e falsos negativos. Os laudos citológicos de patologistas externos apresentaram sensibilidade e especificidade maiores (71,4 e 98,9%, respectivamente), descartando que erros de coleta e coloração possam explicar o baixo desempenho do LAPER. CONCLUSÃO: O baixo desempenho diagnóstico da citologia cervicovaginal pode ser uma barreira para o controle do câncer de colo de útero em Roraima. Ressalta-se a necessidade de capacitação profissional e monitoramento interno e externo em Estados brasileiros com alta incidência de câncer de colo de útero. .

PURPOSE: To evaluate the validity of cervicovaginal cytology performed at LAPER, the main Laboratory of Pathology of the State of Roraima, Brazil, by interrater agreement (external monitoring) and agreement with histopathologic results. METHODS: One hundred women were included, a population-based convenience sample. Their cervical cytological exams were evaluated by the laboratory staff and reviewed by expert medical pathologists, external to the laboratory. Cohen's Kappa index, sensitivity and specificity were evaluated. The study was approved by LAPER coordination and Federal University of Roraima Research Ethics Commitee. RESULTS: Regarding the prevalence of human papillomavirus-related atypical, there was no concordance between the results issued by LAPER and by the external pathologists (k=0.21). A low sensitivity (28.5%) and specificity (89,2%) was detected for the diagnostic performance of LAPER, with a high proportion of false positive and false negative results. The cytological reports of the external pathologists showed higher sensitivity and specificity (71.4 and 98.9%, respectively), ruling out the possibility that errors related to collection methods and staining would explain the low performance of the laboratory. CONCLUSION: A low diagnostic accuracy of cervicovaginal cytology can be a barrier against the control of cervical cancer in Roraima. We emphasize the need for professional training and internal and external monitoring in Brazilian states with a high incidence of cervical cancer. .