Initial hCG Levels in Ectopic Pregnancies After Embryo Transfer: Use of the Bayes Classifier for Risk Assessment

OBJECTIVE: To determine whether the Bayes classifier can be used to distinguish between an ectopic and intrauterine pregnancy following embryo transfer based on early human chorionic gonadotropin (hCG) levels. STUDY DESIGN: Retrospective chart review of patients undergoing in vitro fertilization and diagnosed with a singleton intrauterine or with an ectopic pregnancy. Blood was drawn for hCG levels between days 12 and 20 after transvaginal oocyte aspiration. Statistical analysis was performed using a mixed effects model and the Bayes classifier. RESULTS: Singleton intrauterine (n=91) and ectopic gestations (n=14) were analyzed. hCG levels increased by 51% daily in both groups, but levels in ectopic pregnancies were only 14% of those from the control group on the same day (p<1×10-15). Using the Bayes classifier, an hCG value <18 IU/L indicated a large probability (>75%) that the pregnancy was ectopic. There was no statistically significant difference in regards to endometrial thickness (p=0.77), fresh or frozen embryo transfer (p=0.53), number of embryos transferred (p=0.13), donor or autologous oocytes (p=0.76), or the day of hCG draw (p=0.13 and 0.43 for first and second measurement). CONCLUSION: The Bayes classifier can be used as a tool to alert the healthcare provider of a possible ectopic gestation.