Unsuspected placenta accreta spectrum at vaginal delivery: Assessment of incidence, characteristics, and maternal morbidity

Abstract

To assess incidence, characteristics, and maternal morbidity related to unsuspected placenta accreta spectrum (PAS) at vaginal delivery. This is a retrospective cohort study querying the Healthcare Cost and Utilization Project’s National Inpatient Sample. The study population was 8,694,669 term vaginal deliveries from 1/2016-12/2019. Exclusion criteria included prior uterine scar (including prior cesarean delivery and myomectomy), placenta previa, and preterm delivery. Exposure was assigned by the diagnosis of PAS. The main outcomes were (i) independent characteristics assessed with multivariable binary logistic regression model and (ii) maternal morbidity assessed with inverse probability of treatment weighting. PAS was reported in one in 3,797 vaginal deliveries. In a multivariable analysis, (i) patient characteristics of older age and White race, (ii) uterine factors such as uterine anomaly and uterine myoma, (iii) pregnancy factors including early-term delivery and prior recurrent pregnancy loss, and (iv) fetal factors of in-utero growth restriction and demise were associated with increased likelihood of encountering unsuspected PAS (all, P< 0.05). Of those, uterine anomaly had the strongest association with PAS (adjusted-odds ratio [aOR] 6.23, 95% confidence interval [CI] 4.20-9.26). In a propensity score-weighted model, patients in the PAS group were more likely to have hemorrhage (65.2% vs 4.1%), blood product transfusion (21.3% vs 0.6%), performance of manual placental removal (25.1% vs 0.6%), hysterectomy (14.9% vs < 0.1%), coagulopathy (2.9% vs 0.1%), and shock (2.9% vs < 0.1%) compared to those without. The current study suggests that encountering unsuspected PAS among patients undergoing vaginal delivery with otherwise low-suspicion for PAS is rare. However, unsuspected PAS is associated with significant maternal morbidity.