Combining early pregnancy bleeding with ultrasound measurements to assess spontaneous abortion risk among infertile patients

Approximately 15% of all clinically recognized pregnancies in patients with infertility result in spontaneous abortion. However, despite its potential to have a profound and lasting effect on physical and emotional well-being, the natural history of spontaneous abortion in women with infertility has not been described. Although vaginal bleeding is a common symptom in pregnancies conceived via reproductive technologies, its prognostic value is not well understood.This study aimed to evaluate the combination of early pregnancy bleeding and first-trimester ultrasound measurements to determine spontaneous abortion risk.This was a retrospective cohort study of patients with infertility who underwent autologous embryo transfer resulting in singleton intrauterine pregnancy confirmed by ultrasound from January 1, 2017, to December 31, 2019. Early pregnancy symptoms of bleeding occurring before gestational week 8 and measurements of crown-rump length and fetal heart rate from ultrasounds performed during gestational week 6 (6 0/7 to 6 6/7 weeks of gestation) and gestational week 7 (7 0/7 to 7 6/7 weeks of gestation) were recorded. Modified Poisson regression with robust error variance was adjusted a priori for patient age, embryo transfer day, and transfer of a preimplantation genetic-tested embryo to estimate the relative risk and 95% confidence interval of spontaneous abortion for dichotomous variables. The relative risks and positive predictive values for early pregnancy bleeding combined with ultrasound measurements on the occurrence of spontaneous abortion were calculated for patients who had an ultrasound performed during gestational week 6 and separately for patients who had an ultrasound performed during gestational week 7. The primary outcome was spontaneous abortion in the setting of vaginal bleeding with normal ultrasound parameters. The secondary outcomes were spontaneous abortion with vaginal bleeding and (1) abnormal crown-rump length, (2) abnormal fetal heart rate, and (3) both abnormal crown-rump length and abnormal fetal heart rate.Of the 1858 patients who were included (359 cases resulted in abortions and 1499 resulted in live births), 315 patients (17.0%) reported vaginal bleeding. When combined with ultrasound measurements from gestational week 6, bleeding was significantly associated with increased spontaneous abortion only when accompanied by absent fetal heart rate (relative risk, 5.36; 95% confidence interval, 3.36-8.55) or both absent fetal heart rate and absent fetal pole (relative risk, 9.67; 95% confidence interval, 7.45-12.56). Similarly, when combined with ultrasound measurements from gestational week 7, bleeding was significantly associated with increased spontaneous abortion only when accompanied by an abnormal assessment of fetal heart rate or crown-rump length (relative risk, 5.09; 95% confidence interval, 1.83-14.19) or both fetal heart rate and crown-rump length (relative risk, 14.82; 95% confidence interval, 10.54-20.83). With normal ultrasound measurements, bleeding was not associated with increased spontaneous abortion risk (relative risk: 1.05 [95% confidence interval, 0.61-1.78] in gestational week 6 and 0.80 [95% confidence interval, 0.36-1.74] in gestational week 7), and the live birth rate was comparable with that in patients with normal ultrasound measurements and no bleeding.Patients with a history of infertility who present after embryo transfer with symptoms of vaginal bleeding should be evaluated with a pregnancy ultrasound to accurately assess spontaneous abortion risk. In the setting of normal ultrasound measurements, patients can be reassured that their risk of spontaneous abortion is not increased and that their live birth rate is not decreased.