RESUMEN

Antecedentes: la histeroscopia es el patrón de referencia para diagnosticar anormalidades intrauterinas, su aplicación en ciclos de reproducción ha incrementado la tasa de embarazo; a la fecha existen pocos estudios que evalúen la concordancia diagnóstica del estudio histeroscópico.

Objetivo: evaluar, en residentes de endoscopia ginecológica, la concordancia diagnóstica interobservador de imágenes digitalizadas de estudios histeroscópicos.

Material y métodos: estudio transversal, descriptivo y observacional al que se incluyeron 28 imágenes de estudios histeroscópicos seleccionadas por al menos 2 de 3 expertos en histeroscopia, quienes determinaron que eran adecuadas para realizar una impresión diagnóstica. De los 4 residentes en endoscopia ginecológica se eligieron 2 por muestreo aleatorio simple. Se les presentaron las imágenes de los estudios histeroscópicos de forma aleatoria a cada uno de los residentes y se determinó su concordancia diagnóstica entre ambos utilizando la prueba estadística de kappa con intervalo de confianza de 95%.

Resultados: la concordancia obtenida para las diferentes impresiones diagnósticas, fue: cavidad uterina normal κ = 0.81 (IC 95%, 0.56-1.00), pólipo endometrial κ = 0.71 (IC 95%, 0.33-1.00), mioma submucoso k = 0.71 (IC 95%, 0.33-1.00), adherencias intrauterinas κ = 0.84 (IC 95%, 0.52-1.00), septo uterino 0.76 (IC 95%, 0.43-1.00) y para hiperplasia endometrial-probable cáncer de endometrio κ = 0.87 (IC del 95%, 0.61-1.00).

Conclusiones: la concordancia diagnóstica interobservador en imágenes digitalizadas de estudios histeroscópicos en residentes de cirugía endoscópica ginecológica fue de “buena” a “casi perfecta”.

Palabras clave: variación interobservador, acuerdo, residencia, comprensión de imágenes, diagnóstico.

ABSTRACT

Background: Hysteroscopic studies are of the most used methods to examine the uterine cavity in patients that present abnormal uterine hemorrhage as well as those patients with infertility. The use of hysteroscopic studies during reproductive cycles has increased the successful of pregnancy rates. Up to date, there are not many studies evaluating the inter-observer agreement in the diagnosis of different uterine pathology when using a hysteroscopic study. 

Objective: To evaluate the inter-observer agreement in the diagnosis of uterine pathology when using digitalized images in hysteroscopic studies made by residents of gynecological endoscopy. 

Materials and methods: A cross-sectional, descriptive and observational study was made including 28 images of hysteroscopic studies selected by at least two of three experts in hysteroscopy, who determined that they were adequate to do a diagnostic impresion. From a total of four residents, two were selected using a randomized sampling. The images were shown to each resident in a randomized presentation and the diagnosis agreement was evaluated. Kappa test was used to evaluate the interobserver agreement with 95% confidence interval.

Results: The interobserver agreement obtained by Kappa test for different images for diagnosis of uterine pathology were: normal uterine cavity (κ = 0.81 with CI 95%, 0.56-1.00), endometrial polypus (κ = 0.71 with CI 95%, 0.33-1.00), submucous myoma (κ = 0.71 with CI 95%, 0.33-1.00), intrauterine adherences (k = 0.84 with CI 95%, 0.52-1.00), uterine septum (κ = 0.76 with CI 95%, 0.43-1.00) and endometrial hyperplasia or potential endometrial cancer (κ = 0.87 with CI 95%, 0.61-1.00).

Conclusions: The interobserver agreement using digitalized images in the diagnosis of different uterine pathology from hysteroscopic studies made by residents of endoscopic surgery was high and very high in all cases.

Key words: Interobserver variation, agreement, residency, image compression, diagnosis.