Abstract
OBJECTIVE: To identify independent risk factors for placenta accreta spectrum among pregnancies conceived with assisted reproductive technology. DESIGN: Retrospective cohort study. SETTING: Tertiary hospital. PATIENTS: Individuals who conceived with assisted reproductive technology and reached 20 weeks' gestation or later from 2011 to 2017. INTERVENTIONS: Patient and cycle data was abstracted from hospital records and supplemented with state-level data. Poisson regression was used for multivariate analyses and reported as adjusted relative risks (aRR). MAIN OUTCOME MEASURES: Clinical or histologic placenta accreta spectrum. RESULTS: Of 1,975 qualifying pregnancies, 44 (2.3%) met criteria for accreta spectrum at delivery. In the multivariate model, significant risk factors included low-lying placenta at delivery (aRR, 15.44; 95% CI 7.76-30.72), uterine factor infertility or prior uterine surgery (aRR, 4.68; 95% CI, 2.72-8.05), initial low-lying placentation that resolved (aRR, 3.83; 95% CI, 1.90-7.73), and use of frozen embryos (aRR, 3.02; 95% CI, 1.66-5.48). When the fresh vs frozen variable was replaced with controlled ovarian hyperstimulation, the final model did not change (aRR, 2.40 for unstimulated cycles, 95% CI, 1.32-4.38). With frozen transfers, the accreta rate was 16% when the endometrial thickness was < 6mm vs 3.8% with thicker endometrium (P=.02). CONCLUSIONS: Among pregnancies conceived with assisted reproductive technology, accreta spectrum is associated with low placental implantation (even when resolved), uterine factor infertility and prior uterine surgery, and the use of frozen embryo transfer or unstimulated cycles.