Timing of Delivery in Preeclampsia: Time to Reconsider?

Preeclampsia affects 2%–8% of pregnancies and remains a major cause of maternal morbidity and mortality, perinatal mortality and iatrogenic prematurity worldwide [1]. The incidence of preeclampsia is rising, likely attributable to the shift in clinical context; pregnant patients now present with greater burdens of chronic hypertension, diabetes and obesity, and are older [1]. The timing of delivery in preeclampsia has been guided by a delicate balance between maternal safety and neonatal maturity, but many pivotal studies were conducted in the 1990s–early 2000s, before universal uptake of antenatal corticosteroids, magnesium sulphate for neuroprotection, and today's neonatal intensive care (NICU) capabilities [2, 3]. At the same time, early-onset disease continues to confer substantial maternal risk, and potential for abrupt clinical deterioration [1, 4]. The net result is that the evidence for the timing of delivery decisions is less certain than many guidelines imply. Contemporary guidance recommends delivery at ≥ 37 weeks for preeclampsia without severe features and, expectant management until 34 weeks for preeclampsia with severe features with expedited delivery for maternal or foetal deterioration [4-6]. As obstetricians, we must re-evaluate whether current management strategies continue to balance the maternal versus the neonatal risk, or whether we have reached a point where the incremental improvement in neonatal outcomes from expectant management plateaus earlier than decade-old studies showed. We therefore propose that the current thresholds for empiric delivery for preeclampsia be reexamined. A growing body of evidence demonstrates that planned delivery during the late preterm period (34–36 + 6 weeks' gestation) for patients with preeclampsia without severe features in high-income settings confers maternal benefit, increases NICU admissions but, most importantly, does not increase short-term neonatal morbidity [2, 7]. However, this data must be interpreted with some caution. The Cochrane review was not limited to patients with preeclampsia but also included patients with chronic hypertension [2]. The authors concluded that there was insufficient information to draw conclusions about the effects of planned delivery in the late preterm period on composite infant mortality and severe morbidity [2]. The PHOENIX trial concluded that there is strong evidence to suggest planned delivery decreases maternal morbidity, with more NICU admissions related to prematurity but no indication of greater neonatal morbidity [8]. At 2 years, the PHOENIX follow-up study found no difference in infant neurodevelopmental outcomes between groups, reinforcing that planned delivery from 34 to 36 + 6 weeks confers maternal benefit without evidence of long-term harm [9]. A 2022 individual patient data meta-analysis, which included 1790 patients, demonstrated clear maternal benefit with delivery for preeclampsia at 34 + 0 and 36 + 6 weeks' gestation, a decreased risk of the infant being small for gestational age, and a small increase in short-term neonatal respiratory morbidity [7]. The finding of clear maternal benefit to delivery for preeclampsia at > 34 weeks' gestation was replicated in the CRADLE-4 trial, published in 2023, which examined the timing of delivery in late preterm preeclampsia in low- and middle-income countries [10]. Planned delivery was noted to reduce the risk of severe maternal hypertension and stillbirth with no increase in NICU admissions or neonatal morbidity [10]. For early-onset pre-eclampsia with severe features, the best available synthesis remains the 2018 Cochrane review of six randomised trials (n = 748, recruitment 1990–2013) comparing early delivery (an interventionist approach) to expectant management for severe preeclampsia between 24 and 34 weeks' gestation [3]. An interventionist approach did not clearly demonstrate an impact on reducing the risk of eclampsia (RR 0.98, 95% CI 0.06 to 15.58), pulmonary edema (RR 0.45, 95% CI 0.07 to 3.00) or HELLP syndrome (RR 1.09, 95% CI 0.62 to 1.91). Neonates in the interventionist group had higher risks of intraventricular haemorrhage (RR 1.94, 95% CI 1.15 to 3.29), respiratory distress caused by hyaline membrane disease (RR 2.30, 95% CI 1.39 to 3.81) and need for invasive ventilation (RR 1.50, 95% CI 1.11 to 2.02) [3]. The mean difference in gestational age at delivery between the two groups was 9.91 days (95% CI −16.37 to −3.45 days). No significant difference was found between the two groups regarding the likelihood of NICU admission (RR 1.19, 95% CI 0.89 to 1.60), but the interventionist group was less likely to be small-for-gestational age (RR 0.38, 95% CI 0.24 to 0.61) [3]. The authors concluded there was insufficient evidence to draw reliable conclusions regarding the impact of early delivery on maternal outcomes. There are several important considerations when interpreting the findings of this Cochrane review. Firstly, the review noted that the quality of evidence overall was low to very low, indicating the need for further high-quality research to better understand the comparative benefits and risks of interventionist versus expectant management for preeclampsia with severe features in this gestational window. In addition, recruitment for the six included trials spanned from 1990 to 2013 with four out of six trials undertaken prior to 2005. Pros and cons of expectant management versus delivery at < 34 weeks' gestation in preeclampsia with severe features is summarised in Table 1. The older randomised evidence does not robustly establish maternal benefit for either strategy, and neonatal trade-offs reflect a pre-modern era. Given the worsening trends in maternal outcomes and improving outcomes of premature infants, we have an obligation to re-examine whether expectant management of patients with preeclampsia with severe features at gestational ages of < 34 weeks' gestation remains the preferred strategy. This question has not been examined in large-scale randomised controlled trials in recent decades when high-resource countries have witnessed significant advances in neonatal care with NICU units capable of managing preterm infants with consistently improving outcomes [11]. We acknowledge that at the limits of viability (22–25 weeks), where neonatal survival increases dramatically with each week of gestation and survival without major morbidity also improves, equipoise for empiric delivery is more difficult to justify [11]. Therefore, reevaluation of expectant management strategies is most applicable beyond 28–30 weeks where gains in neonatal outcomes per week are smaller, and maternal risk continues to accrue. A contemporary trial examining the timing of delivery in preeclampsia must reconcile the fact that interventionist and expectant strategies push maternal and neonatal outcomes in opposite directions; the primary endpoint should therefore integrate both domains. One approach would be to utilise a Desirability of Outcome Ranking (DOOR) analysis. A DOOR analysis evaluates and compares the outcomes of different treatment strategies by ranking outcomes based on their overall desirability from the patient's perspective, taking into account both the benefits and harms associated with each intervention [12]. Additional key trial design considerations should include prespecified subgroup analyses for patients with chronic hypertension, foetal growth restriction and incorporation of cost-effectiveness to inform policy. Trials to date examining the timing of delivery in preeclampsia have focused on short-term neonatal outcomes and fail to account for long-term outcomes of the offspring into adulthood, thus overlooking the potential impact of in utero programming [2, 3]. Foetal programming, known as the Barker Hypothesis, refers to the concept that the intrauterine environment can have long-lasting effects on the development and health of the foetus, influencing outcomes well into adulthood [13]. This concept suggests that adverse conditions, like preeclampsia, can lead to developmental adaptations that are necessary for foetal survival but may predispose the individual to chronic conditions such as cardiovascular disease, diabetes and hypertension later in life [13]. Thus, in the context of severe preeclampsia, expectant management might not always be beneficial in the long term. While it can potentially reduce neonatal morbidity associated with prematurity, it also exposes the foetus to a hostile intrauterine environment, and the potential detrimental long-term effects of the in utero environment in preeclampsia necessitate careful consideration. Beyond theoretical programming, longitudinal cohorts show that children born preterm have higher risks of neurodevelopmental impairment and cardiometabolic vulnerability into childhood and beyond, with risk gradients that worsen at lower gestations. In the context of late-preterm pre-eclampsia, the PHOENIX 2-year follow-up found no difference in neurodevelopment between planned delivery and expectant management [9]. For earlier gestations, population cohorts demonstrate dose-dependent risks for motor, cognitive and metabolic outcomes that must be weighed against risks of prolonged exposure to severe disease [11]. Together, these data highlight that the timing of delivery decisions should consider both immediate perinatal trade-offs and the child's lifelong outcomes. When evaluating the timing of delivery in preeclampsia, it is equally important to consider the impact of maternal cardiovascular reprogramming, which was underappreciated in older trials. Maternal reprogramming refers to the series of adaptive changes occurring in response to the hypertensive environment of preeclampsia [14]. In severe preeclampsia, there is impaired vascular remodelling, increased vascular resistance, and a dysregulated balance between vasoconstriction and vasodilation. These vascular alterations, driven by endothelial dysfunction and heightened inflammatory response, contribute significantly to the maternal morbidity associated with the disease, including risks of end-organ injury and cardiovascular complications, with risks more profound in early onset preeclampsia prior to 34 weeks of gestation [14]. Prolonged exposure to this pathological vascular state increases maternal cardiovascular risk both acutely and long term, raising critical questions about the benefits of expectant management versus the risks of continued maternal vascular injury [14]. Just as foetal programming can predispose the offspring to adverse long-term health outcomes, ongoing maternal vascular compromise may similarly result in increased short-term and long-term cardiovascular risks for the mother. Thus, while prolonging pregnancy may benefit foetal maturation, it must be balanced against the potential worsening of maternal vascular health, supporting a nuanced, individualised approach to delivery timing in preeclampsia with severe features. As personalised medicine advances, particularly with Artificial Intelligence (AI)-enabled risk stratification [15], contemporary management of preeclampsia is unlikely to remain a one-size-fits-all model. Instead, the timing of delivery may increasingly be individualised by integrating each patient's evolving trajectory (e.g., blood pressure control, laboratory results, foetal growth/Doppler, antenatal surveillance) to guide individualised timing of delivery. Any AI tools should be transparent, externally validated and routinely recalibrated, functioning within clear safety guardrails. There is a growing body of evidence suggesting that the timing of delivery in both preeclampsia with and without severe features should be reassessed. However, the balance between prolonging pregnancy to enhance foetal maturity and the risk of severe maternal morbidity and mortality must be considered. In addition, the long-term impact of expectant management of severe preeclampsia on foetal programming and health into adulthood should be examined. We feel there is equipoise with regard to the timing of delivery, particularly in severe preeclampsia and new approaches to analyzing outcomes to justify a large-scale randomised trial to test whether earlier delivery in pregnant patients with severe preeclampsia achieves a more desirable outcome than expectant management. R.H.: writing of original draft, literature review, drafting of reviewer responses and revising manuscript. A.M.: literature review, review/editing of manuscript. M.M.: literature review, contributed to writing of original draft. T.K.: review and editing of manuscript. B.S. & G.S.: formation of hypothesis, supervision, review/editing of manuscript. We thank Grace Spencer for her assistance in the preparation of this manuscript. The authors declare no conflicts of interest. Data sharing not applicable to this article as no datasets were generated or analysed during the current study.