Portal de Periódicos da CAPES

Re: Prevalence of endometrial polyps and abnormal uterine bleeding in a Danish population aged 20–74 years

2009; Wiley; Volume: 33; Issue: 3 Linguagem: Inglês

10.1002/uog.6345

1469-0705

L. Valentin,

Endometriosis Research and Treatment

I read with great interest the important article by Dreisler et al.1, in which the authors tried to determine the prevalence of endometrial polyps among Danish women and the association between endometrial polyps and abnormal vaginal bleeding. The results are surprising, because they go against previously published results2, 3 as well as against common clinical opinion. They show that polyps are more common among asymptomatic women than among women with abnormal vaginal bleeding, 38% of women without polyps having abnormal vaginal bleeding versus 19% of women with endometrial polyps (P < 0.01). The article left me wondering whether the reported results reflect reality or whether they could be explained by selection bias. Of 1660 women invited to participate, 686 (41%) were included and 619 (37%) underwent full evaluation: 393 (24%) of the women who received an invitation letter to participate did not respond, and 273 (28%) of the 959 eligible responders declined to participate. Who were the responders? Who were those of the eligible responders that agreed to participate? Could it be that symptomatic women were more likely to respond and to accept participation than asymptomatic women? Could it be that those symptomatic women who did not respond or who declined participation were those who had recently been scheduled for treatment of a (suspected) polyp? If this were the case, it could explain why the prevalence of polyps was lower among the symptomatic than among the asymptomatic women included in the study. Could it be that the asymptomatic women who participated were those who knew they belonged to a risk group for polyps or who intuitively felt that ‘something was wrong’? If this were the case, it could explain why the prevalence of polyps was higher among the asymptomatic than among the symptomatic women included in the study. Some kind of selection bias seems likely because the prevalence of abnormal vaginal bleeding in the study population was high (of the postmenopausal women, 15% had postmenopausal bleeding; of the premenopausal women, 36% had intermenstrual bleeding and 10% had menorrhagia), as was the prevalence of endometrial cancer (1.3% of the postmenopausal women had endometrial cancer, a prevalence at least 10 times higher than the incidence of endometrial cancer among postmenopausal women). If 36% of premenopausal women had intermenstrual bleeding, then cannot intermenstrual bleeding (as defined in this study) be considered normal? What would be the prevalence of polyps among symptomatic and asymptomatic premenopausal women in this study if women with intermenstrual bleeding were included in the asymptomatic group? I would encourage the authors to estimate the likelihood of selection bias, explaining their results. Do they have the possibility to check how many of the non-responders (n = 393) and how many of the eligible women who declined to participate (n = 273) were diagnosed with endometrial polyps shortly before or shortly after they were invited to participate and whether they had abnormal bleeding before the polyp was diagnosed? I understand there is a national Danish hysteroscopy registry and I wonder whether it is complete and whether it is possible to extract reliable information from that registry. The results of this study are important. If they reflect reality, they imply that polyps should not be regarded as a cause of abnormal vaginal bleeding and that polyp resection should not be the first-line treatment for abnormal vaginal bleeding. Indeed, the authors cite two studies in which hysteroscopic resection of endometrial polyps did not always lead to symptom relief4, 5. They do not cite a study which suggests the opposite6, findings which are corroborated in an article in press in this Journal7. However, there has been no randomized controlled trial evaluating whether polyp resection relieves abnormal vaginal bleeding. L. Valentin*, * Department of Obstetrics and Gynecology, Malmö University Hospital, Lund University, Malmö, Sweden