Independent value of serum β-human chorionic gonadotropin in predicting early pregnancy loss risks in IVF/ICSI cycles

BACKGROUND: Early pregnancy loss (EPL) is the most prevalent complication, particularly in couples undergoing assisted reproductive technology treatment. The present study aimed to determine whether the serum β-human chorionic gonadotropin (β-hCG) level after 14 days of embryo transfer, either alone or in conjunction with other parameters in IVF/ICSI cycles, could be used to predict subsequent EPL. METHODS: This was a retrospective cohort study of all couples who received clinical pregnancy and underwent fresh IVF/ICSI cycles at a single large reproductive medical center between January 2013 and June 2020. The research involved a total of 6600 cycles. For risk variables, we conducted the least absolute shrinkage and selection operator (LASSO) analysis, and for risk scoring, we used logistic regression coefficients. To analyze relevant risk factors for EPL, univariate and multivariate logistic regression analyses were employed. Areas under the curve (AUC) were determined and compared between β-hCG and other factors using receiver operating characteristic (ROC) curves. RESULTS: β-hCG level was considerably lower in women who had EPL than in those who were ongoing pregnancy (564.03 ± 838.16 vs 1139.04 ± 1048.72 IU/L, p< 0.001). Univariable and multivariable logistic regression revealed that β-hCG levels were significantly correlated with the probability of EPL, independent of other risk factors. More importantly, the β-hCG level could independently predict the occurrence of EPL and was comparable to the model that combined other risk factors. The optimal serum β-hCG cut-off value for predicting EPL was 542.45 IU/L. CONCLUSIONS: Our results suggest that the serum β-hCG level has a strong independent predictive value for EPL occurrence in fresh IVF/ICSI cycles.