Objective: The aim of this study was retrospective analysis of how accurate was per-operational visual evaluation of malign process in an uterus cavity during hysteroscopy. And to evaluate whether increasing experience of hysteroscopiers leads to significant accuracy considering the neoplasm of an uterus cavity. Setting: Department of Gynaecology and Obstetrics, Havířov. Method: In Havířov Hospital, 1.200 hysteroscopies altogether were performed in the period from December 1995 to March 1999. In this group, there were 26 cases of histologically verified endo- metrial cancer. The authors retrospectively attempted to evaluate how accurately the suspected disorder was already stated during the per-operational hysteroscopy. The advantage of compa- ring the sub-group was taken in the first 690 hysteroscopies, of which the complex analysis was publisched in Čs. Gynekologie 5/98, and in the sub-group of 510 hysteroscopies performed in the following period, to state whether experience can more precisely define the per-operational ma- lignity recognition. The statistical analysis was performed by means of the Fischer exact test of numerical charts. Among other things, the MEDLINE database was used during discussion. Results: The endometrial cancer was encountered 26 times altogether, it means in 2.2% cases of hysteroscopies. Carcinoma in situ occured three times, the stage IA three times, IB 17 times, IC three times. A hysteroscopier described the negative finding incorrectly 13 times altogether, it means 50% of all cases. The sensitivity and the specificity of hysteroscopy for endometrial cancer prediction was 50% and 99.5% (P < 0.01). The comparison of the first sub-group results (16 cases of endometrial cancer, sensitivity 75%, specificity 99.7%, (P < 0.01) and the second sub-group (10 cases of endometrial cancer, sensitivity 10%, specificity 99.2%, P = 0.09%) indicates that even increasing experience of a hysteroscopier does not more precisely define per-operation malign consideration. Conclusion: The authors have come to the conclusion that the pre-operation consideration of intrauteral pathology during hysteroscopy does not allow to assess precisely whether there is a neoprocess of an uterus cavity, or not. Even growing experience does not define with more precision verification of malign dosorders espetially at early stages of this illness. Hysteroscopy always has to be supplemented with endometrium biopsy.

        Key words: hysteroscopy, endometrial neoplasm, endometrium, adenocarcinoma-diagnosis