Resumo Objetivo Avaliar a acurácia do ultrassom transvaginal para o diagnóstico de lesões intrauterinas, tendo a histeroscopia como padrão de referência. Métodos Foi realizado um estudo observacional prospectivo em 307 pacientes, submetidas à histeroscopia após ultrassonografia prévia para comparação dos resultados. A histeroscopia foi realizada por duas médicas com experiência, e os exames de ultrassom foram realizados em diversas fontes, públicas ou privadas, como ocorre no cotidiano da assistência à saúde em nosso meio. Foram avaliados sensibilidade, especificidade e acurácia, tendo a histeroscopia como padrão-ouro. O nível de concordância foi avaliado pelo teste de Kappa. Resultados A idade média foi de 56,55±12,3 anos. Os resultados para pólipo endometrial foram: sensibilidade 39.8%, especificidade 72,7%, acurácia de 52,8%, e índice Kappa 0,11 (p=0,025). Para mioma, sensibilidade 46,7%, especificidade 95,0%, acurácia 87,9%, e índice Kappa 0,46 (p<0,001). Para espessamento endometrial, sensibilidade 68,7%, especificidade 41,7%, acurácia 47,6%, e índice Kappa de 0,06 (p=0,126). Para atrofia, sensibilidade 6,7%, especificidade 99,3%, acurácia 90,2%, e índice Kappa 0,10 (p=0,006). Para outros achados, sensibilidade 15,6%, especificidade 99,6%, acurácia 87,3%, e índice Kappa 0,23 (p<0,001). Conclusão Nosso estudo demonstrou baixo nível de acurácia da ultrassonografia transvaginal para o diagnóstico de lesões endometriais, quando realizada por profissional não experiente. Assim, é importante considerar o uso da histeroscopia para evitar tratamentos desnecessários e inadequados.

Abstract Objective To evaluate the accuracy of transvaginal ultrasound in the diagnosis of intrauterine lesions, using hysteroscopy as the gold standard. Methods This was a prospective observational study with 307 patients. All patients underwent hysteroscopy after a previous transvaginal ultrasound to compare the results. The hysteroscopy was performed by experienced examiners, and transvaginal ultrasounds were performed in various public and private services, which is reflective of routine healthcare practices in obstetrics and gynecology. The sensitivity, specificity, and accuracy of the transvaginal ultrasound were calculated using hysteroscopy as the gold standard. The level of agreement between the two exams was calculated using the Kappa test. Results Themean age was 56.55±12.3 years. For endometrial polyps, we observed a sensitivity of 39.8%, specificity of 72.7%, accuracy of 52.8%, and Kappa index of 0.11 (p=0.025). For fibroids, the sensitivity was 46.7%, specificity was 95.0%, accuracy was 87.9%, and Kappa index was 0.46 (p<0.001). For endometrial thickening, the sensitivity was 68.7%, specificity was 41.7%, accuracy was 47.6%, and Kappa index was 0.06 (p=0.126). For endometrial atrophy, we found a sensitivity of 6.7%, specificity of 99.3%, accuracy of 90.2%, and Kappa index of 0.10 (p=0.006). For the other findings, the sensitivity was 15.6%, specificity was 99.6%, accuracy was 87.3%, and Kappa index was 0.23 (P<0.001). Conclusion Our study demonstrated a low level of accuracy of transvaginal ultrasound for the diagnosis of endometrial lesions, when performed by a non-experienced professional. Thus, it is important to consider the use of hysteroscopy to avoid unnecessary and inappropriate treatments.