Analysis of factors predicting bleeding during endometrial preparation with estrogens for oocyte donation.

Abstract

Objective: To identify and evaluate the relationship between potential risk factors and the hazard rate of bleeding during endometrial preparation with estrogens for oocyte donation.Design: Retrospective study. Follow up of endometrial preparation cycles with statistical methods for survival data analysis.Materials/Methods: 1348 cycles of endometrial preparation with estrogens performed in our institution between January 1st, 1999 and December 31st, 2000 were included. Incidence of cancellation because of bleeding was evaluated. Previous administration of a GnRHa, age of the patient, endometrial thickness, E2 serum level and female etiology were related to the hazard rate of cancellation because of bleeding using the proportion hazard model (Cox regression).Results: 384 cycles (28.5%) were cancelled because of bleeding; 26.6% of cycles with GnRHa and 36.3% of cycles without GnRHa (p = 0.0002). Cancelled cycles were longer time under estrogens therapy than cycles that did not present bleeding (39.5 vs 36.7 days; p = 0.008) and showed a thicker endometrium after 14 days of treatment (9.3 vs 8.9 mm; p = 0.006). The relations of the analyzed factors with the hazard rate of cancellation because of bleeding determined with Cox regression are shown in Table 1. The analysis of female etiology showed that patients with endometriosis had a higher hazard rate of bleeding when compared with patients whose indication for oocyte donation was their age (35.3% vs 25.5%; O.R. = 1.65 CI 95%; 1.11–2.47; p = 0.013). When a proportional hazard regression model was constructed using all the predictive variables analyzed, the selected model through backward elimination included the two significant variables shown in the table, defining the predictive equation of bleeding as follows: h(t;X)=h0(t)x e0.08×Endometrial thickness−0.25×GnRHa(p=0.0006). The prognostic index is defined with the equation: PI=1.08×Endometrial thickness+0.78×GnRHa.Table 1Cox regression.legendRelations of the analyzed factors with the hazard rate of cancellation because of bleeding determined with Cox regression.(O.R.; CI 95%)pUse of aGnRH0.78; 0.61-0.990.038Age0.99; 0.98-1.010.6Endometrial thickness1.08; 1.04-1.130.0001E2 level1.0; 1.0-1.00.31legend Relations of the analyzed factors with the hazard rate of cancellation because of bleeding determined with Cox regression. Open table in a new tab Conclusions: The use of a GnRHa prior to endometrial preparation with estrogens for oocyte donation showed to protect of bleeding. In the other hand, patients that developed a thicker endometrium and patients with endometriosis showed a higher hazard rate of cancellation because bleeding.Supported by: IVI. Objective: To identify and evaluate the relationship between potential risk factors and the hazard rate of bleeding during endometrial preparation with estrogens for oocyte donation. Design: Retrospective study. Follow up of endometrial preparation cycles with statistical methods for survival data analysis. Materials/Methods: 1348 cycles of endometrial preparation with estrogens performed in our institution between January 1st, 1999 and December 31st, 2000 were included. Incidence of cancellation because of bleeding was evaluated. Previous administration of a GnRHa, age of the patient, endometrial thickness, E2 serum level and female etiology were related to the hazard rate of cancellation because of bleeding using the proportion hazard model (Cox regression). Results: 384 cycles (28.5%) were cancelled because of bleeding; 26.6% of cycles with GnRHa and 36.3% of cycles without GnRHa (p = 0.0002). Cancelled cycles were longer time under estrogens therapy than cycles that did not present bleeding (39.5 vs 36.7 days; p = 0.008) and showed a thicker endometrium after 14 days of treatment (9.3 vs 8.9 mm; p = 0.006). The relations of the analyzed factors with the hazard rate of cancellation because of bleeding determined with Cox regression are shown in Table 1. The analysis of female etiology showed that patients with endometriosis had a higher hazard rate of bleeding when compared with patients whose indication for oocyte donation was their age (35.3% vs 25.5%; O.R. = 1.65 CI 95%; 1.11–2.47; p = 0.013). When a proportional hazard regression model was constructed using all the predictive variables analyzed, the selected model through backward elimination included the two significant variables shown in the table, defining the predictive equation of bleeding as follows: h(t;X)=h0(t)x e0.08×Endometrial thickness−0.25×GnRHa(p=0.0006). The prognostic index is defined with the equation: PI=1.08×Endometrial thickness+0.78×GnRHa. Conclusions: The use of a GnRHa prior to endometrial preparation with estrogens for oocyte donation showed to protect of bleeding. In the other hand, patients that developed a thicker endometrium and patients with endometriosis showed a higher hazard rate of cancellation because bleeding. Supported by: IVI.