Discriminant analysis forecasting model of first trimester pregnancy outcomes developed by following 9,963 infertile patients after in vitro fertilization

Abstract

Objective: To investigate a forecasting method developed to predict first trimester pregnancy outcomes using the first routine ultrasound scan for early pregnancy on days 27-29 after ET and to determine whether to perform a repeated scan several days later based on this forecasting method.

Design: Prospective analysis.

Setting: Infertile patients at an assisted reproductive technology center.

Patient(s): A total of 9,963 patients with an early singleton pregnancy after in vitro fertilization (IVF)-ET.

Intervention(s): None.

Main outcome measure(s): Ongoing pregnancy >12 weeks of gestation.

Result(s): The classification score of ongoing pregnancy was equal to (1.57 × Maternal age) + (1.01 × Mean sac diameter) + (-0.19 × Crown-rump length) + 25.15 (if cardiac activity is present) + 1.30 (if intrauterine hematomas are present) - 47.35. The classification score of early pregnancy loss was equal to (1.66 × Maternal age) + (0.84 × Mean sac diameter) + (-0.38 × Crown-rump length) + 8.69 (if cardiac activity is present) + 1.60 (if intrauterine hematomas are present) - 34.77. In verification samples, 94.44% of cases were correctly classified using these forecasting models.

Conclusion(s): The discriminant forecasting models are accurate in predicting first trimester pregnancy outcomes based on the first scan for early pregnancy after ET. When the predictive result is ongoing pregnancy, a second scan can be postponed until 11-14 weeks if no symptoms of abdominal pain or vaginal bleeding are present. When the predictive results suggest early pregnancy loss, repeated scans are imperative to avoid a misdiagnosis before evacuating the uterus.

Keywords: Discriminate analysis; early pregnancy loss; forecasting model; in vitro fertilization-embryo transfer; transvaginal ultrasound.