Risk Factors for Adverse Maternal Outcomes among Patients with Severe Preeclampsia Before 34 Weeks

Objective This study aimed to characterize rates of maternal morbidity associated with early (<34 wk) preeclampsia with severe features and to determine factors associated with developing these morbidities. Study Design Retrospective cohort study of patients with early preeclampsia with severe features at a single institution from 2013 to 2019. Inclusion criteria were admission between 23 and 34 weeks and diagnosis of preeclampsia with severe features. Maternal morbidity defined as death, sepsis, intensive care unit (ICU) admission, acute renal insufficiency (acute kidney injury [AKI]), postpartum (PP) dilation and curettage, PP hysterectomy, venous thromboembolism (VTE), PP hemorrhage (PPH), PP wound infection, PP endometritis, pelvic abscess, PP pneumonia, readmission, and/or need for blood transfusion. Death, ICU admission, VTE, AKI, PP hysterectomy, sepsis, and/or transfusion of >2 units were considered severe maternal morbidity (SMM). Simple statistics used to compare characteristics among patients experiencing any morbidity and those not. Poisson regression used to assess relative risks. Results Of 260 patients included, 77 (29.6%) experienced maternal morbidity and 16 ( 6.2%) experienced severe morbidity. PPH (n = 46, 17.7%) was the most common morbidity, although 15 (5.8%) patients were readmitted, 16 (6.2%) needed a blood transfusion, and 14 (5.4%) had AKI. Patients who experienced maternal morbidity were more likely to be advanced maternal age, have preexisting diabetes, have multiples, and deliver nonvaginally (all ps<0.05). Diagnosis of preeclampsia <28 weeks or longer latency from diagnosis to delivery were not associated with increased maternal morbidity. In regression models, the relative risk of maternal morbidity remained significant for twins (adjusted odds ration [aOR]: 2.57; 95% confidence interval [CI]: 1.67, 3.96) and preexisting diabetes (aOR: 1.64; 95% CI: 1.04, 2.58), whereas attempted vaginal delivery was protective (aOR: 0.53; 95% CI: 0.30, 0.92). Conclusion In this cohort, more than 1 in 4 patients diagnosed with early preeclampsia with severe features experienced maternal morbidity, whereas 1 in 16 patients experienced SMM. Twins and pregestational diabetes were associated with higher risk of morbidity, whereas attempted vaginal delivery was protective. These data may be helpful in promoting risk reduction and counseling patients diagnosed with early preeclampsia with severe features.