The influence of age on hysterectomy rates after transcervical endometrial resection (TCER) is well known (1). Other suggested, but inconsistently shown, risk factors are uterine size and the presence of leiomyomas (2). The influence of the Body Mass Index (BMI) on hysterectomy rates after TCER is unknown. The purpose of the present study was to investigate whether BMI was related to the frequency of hysterectomy after endometrial resection.
Material and methodsFrom 1990 to 1996, a total of 367 women had a TCER procedure at the Department of Gynecology, Holstebro Hospital, Denmark. Women having a repeat TCER were only included with their first TCER. The criteria for TCER were at that time subjectively experienced meno-metrorrhagia, a uterine length B12 cm, and only minor submucous leiomyoma (maximum diameter 3Á4 cm) with more than half of the mass projecting into the uterine cavity. No account was taken of the women's BMI status. GnRH-agonists were not used. Resection was carried out using a standard 9-mm resectoscope with glycine as distension medium. As part of another investigation on urological symptoms (3), all patient records were reviewed in 2005 using the central computerized patient registration system in Denmark, to clarify whether the women had a later hysterectomy.Statistical analysis was carried out with the SOLO software (BMPD, Los Angeles, USA). The Mann ÁWhitney test for unpaired analysis was used to test differences between two means, and the Kruskal ÁWallis test for group means. Regression analysis was carried out with hysterectomy as the dependent variable, and BMI, age and operative data (uterine length, haemoglobin, glycin deficit, operative time, weight of the resected endometrium) as independent variables. Analysis of time until hysterectomy was by the Kaplan ÁMeier method. Potential variables affecting the outcome were compared using the PetoÁWilcoxon test after appropriate patient data grouping. Patient data investigated were BMI, age, cyclicity, parity, previous vaginal delivery, cesarean section, tubal sterilisation, preoperative knowledge of leiomyomas or adenomyosis, and histopathological endometrial phase. Values are given as medians with range, unless otherwise stated. A two-sided p-value B0.05 was considered significant.
ResultsPatient characteristics and perioperative data are shown in Table I. A total of 96 women (26%) underwent hysterectomy after TCER.The association between BMI and hysterectomy rate is shown in Figure 1. Obese women had a significantly lower risk of hysterectomy than normal and overweight women. Obese women had a median age of 47 years (31 Á73) compared to 44