Reproducibility study of cervical cytopathology in Mexico: a need for regulation and professional accreditation.

Due to the high rate of false negative results in diagnosis of cervical cytopathology, in many countries its practice has been transformed through the application of several interventions aimed at medical regulation to improve diagnostic accuracy. Diagnostic reproducibility of gynecological cytopathology was evaluated in a series of 20 cytology specimens [Papanicolaou (Pap)] and 20 cervical biopsy. (CB) studies in different clinical stages, during 1994. The observation unit consisted of 30 pathologists who observed 2 groups of 20 Pap and 20 CB specimens. The standard was a cytopathologist certified by the Pathological Anatomy Council of Mexico. Intraclass reproducibility in gynecological cytopathology is low in Mexico. In a group analysis, concordance increased as clinical status of the cervical lesion increased. For moderate dysplasia, concordance in Pap was kappa = 0.04, compared to 0.23 in CB. Concordance of diagnosis of invasive cancer was 0.29 for Pap and 0.64 for CB. Using weighted kappa at the individual level for all possible diagnoses, concordance varied from 0.29 to 0.59 for Pap, and 0.42 to 0.65 for CB. The problem of reproducibility in cervical cytopathology in Mexico emphasizes the need for continuing education, uniform diagnostic criteria, and the advantages of a single operational classification-possibly the Bethesda System-since current classification systems are obsolete.

[Diagnostic concordance in gynecologic cytology]. / Concordancia diagnóstica en citología ginecológica.

OBJECTIVE: To evaluate the diagnostic agreement in Papanicolaou of pathologists and cytotechnologists using kappa values for concordancy. METHODS: The diagnostic variation was estimated in 20 gynecological cytology (Pap) specimens by 30 pathologists and 7 cytotechnologists attending the XXXVII Congress of the Mexican Association of Pathologists in 1994. RESULTS: The best concordancy versus an expert was in tumoral diathesis (pathologists' kappa = 0.36; cytotechnologists' kappa = 0.35) and kollocytos (pathologists' kappa = 0.55; cytotechnologists' kappa = 0.36). The least concordancy was observed in anisonucleosis (pathologists' kappa = 0.11; cytotechnologists' kappa = 0.02), nuclear hyperchromasia (pathologists' and cytotechnologists' kappa = 0.11) and dyskeratosis (pathologists' kappa = 0.11; cytotechnologists' kappa = 0.16). The kappa values for cervical neoplasia showed poor agreement, and in invasive cervical cancer it was 0.30. CONCLUSIONS: There was a low concordancy of Pap diagnosis in the study. It is convenient to try to improve the concordancy of cytologic diagnosis in Mexico. One strategy could be the use of a uniform diagnostic criteria and the adoption of a single nomenclature.

Control de calidad interno en citología cérvico-vaginal mediante revisión rápida: evaluación de la capacidad del personal del laboratorio para utilizar el procedimiento / Rapid rescreening as internal quality control in cervical citology. Personal proficiency testing of the laboratory personnel

Entre los procedimientos de control interno de calidad para evitar falsos negativos en citología ginecológica, la segunda revisión rápida de todos los casos negativos ha demostrado su eficacia. Para poner en práctica el procedimiento, antes de utilizarlo evaluamos la capacidad del personal del laboratorio. Se estudiaron 66 casos escogidos aleatoriamente revisándose cada laminilla en un tiempo de 30, 50 segundos, utilizando el objetivo 10x; las laminillas anormales, se separaron para ser evaluadas posteriormente. Hubo tres casos anormales que correspondieron, por diagnóstico de consenso, a cambios atípicos de etiología incierta, favoreciendo una lesión escamosa intraepitelial de bajo grado. Un observador, discrepó de los resultados obtenidos por el resto del grupo, ya que sólo identificó uno de los tres caso anormales; por esta razón, nuestros valores globales promedio disminuyeron: la sensibilidad fue de 91.6 por ciento, la especificidad de 100 por ciento, el valor predictivo positivo fue de 100 por ciento y el negativo de 99.6 por ciento, la tasa global de falsos negativos subió a 8.3 por ciento. Este estudio identifica al personal del laboratorio que está capacitado para poner en práctica el procedimiento, por lo que el observador que falló en sus apreciaciones deberá tener un adiestramiento ulterior.