Establishment of a prediction model for the impact of endometrial thickness on the day of embryo transfer on ectopic pregnancy in frozen-thawed embryo transfer cycle

OBJECTIVE: This study aims to investigate the factors affecting the ectopic pregnancy (EP) rate in the frozen-thawed embryo transfer (FET) cycle. METHODS: This study retrospectively analyzed 5606 FET cycles, including 5496 cycles resulting in intrauterine pregnancy and 110 cycles resulting in EP. Smooth curve fitting and piece-wise linear regression were utilized to evaluate a non-linear association between endometrial thickness (EMT) and EP. Multiple logistic regression analysis was used to study the effect of EMT on the embryo transfer (ET) day and other indexes on EP rate after adjusting for confounding factors. A nomographic prediction model was employed to predict EP occurrence. The predictive efficacy of the model was assessed using the area under the receiver operating characteristic (ROC) curve (AUC), utilizing the bootstrap sampling method for internal validation. RESULTS: After accounting for the confounding factors, the segmented linear regression analysis indicated that the EMT inflection point was 9 mm; the EP rate significantly decreased by 28% with each additional millimeter of EMT up to 9 mm (odds ratio (OR) = 0.72; 95% confidence interval (CI), 0.53–0.99; P = 0.0412) while insignificantly decreased when the EMT was greater than 9 mm (OR = 0.91; 95% CI, 0.76–1.08; P = 0.2487). Multivariate logistic regression analysis revealed that after adjusting for confounders, EP risk significantly increased in the number of previous EPs ≥ 1 (OR = 2.29; 95% CI, 1.26–4.16; P = 0.0064) and tubal factor infertility (OR = 3.86; 95% CI, 2.06–7.24; P < 0.0001). Conversely, EP risk was significantly reduced by the increased EMT (OR = 0.84; 95% CI, 0.74–0.96; P = 0.0078) and the blastocyst transfer (OR = 0.45; 95% CI, 0.27–0.76; P = 0.0027). These variables were used as independent variables in a nomogram prediction model, resulting in an AUC of 0.685. The nomination models were internally verified using self-sampling (bootstrap sampling resampling times = 500). This validation yielded an AUC of 0.689, with a sensitivity of 0.6915 and a specificity of 0.5790. The internal validation indicated minimal fluctuations in the AUC, signifying a relatively stable model. CONCLUSION: Undergoing EMT on the day of ET poses a separate EP risk in the FET cycle; to mitigate the EP incidence, the EMT should exceed 9 mm before ET. Furthermore, previous EPs and tubal factor infertility were additional factors independently increasing EP risk. Furthermore, implementing blastocyst transfer demonstrated that EP incidence was significantly reduced. Utilizing a nomogram predicting system enables EP risk evaluation before ET for individual patients, establishing a basis for devising clinical strategies for ET.