Fetal Growth Restriction: Doppler Surveillance and Delivery Timing

The Clinical Problem in Fetal Growth Restriction

Fetal growth restriction describes a fetus that fails to reach its genetically determined growth potential, usually because of placental insufficiency. It complicates up to 10% of pregnancies and ranks second only to prematurity among causes of perinatal death.

The danger concentrates in the fetuses whose placental function is actively failing. When end-diastolic flow in the umbilical artery is absent or reversed before 34 weeks, the risk of fetal death rises steeply, as a meta-analysis of 31 studies quantified.

This gradient creates the central dilemma of management. Deliver too early and the infant faces the harms of prematurity; wait too long and a salvageable fetus may die or sustain hypoxic injury. Surveillance exists to find the narrow window between these two harms.

Two phenotypes frame the problem. Early-onset restriction, before 32 weeks, is strongly placental, follows a fairly predictable Doppler deterioration, and is dominated by the prematurity trade-off. Late-onset restriction is milder but harder to detect, and contributes disproportionately to term stillbirth.

Available Approaches

Four approaches dominate how clinicians decide when to deliver a growth-restricted fetus, each anchored to a different physiological signal. The table summarises the trial and systematic-review evidence behind them.

Umbilical Artery Doppler: The Surveillance Foundation

Umbilical artery Doppler reflects placental resistance and helps separate true fetal growth restriction from constitutional smallness. It is the one modality every major guideline endorses. The Cochrane meta-analysis associated its use with fewer perinatal deaths and fewer obstetric interventions.

The trialists rate this evidence only moderate, and the number needed to treat is large with a very wide interval. Progressive deterioration — decreased, then absent, then reversed end-diastolic flow — marks escalating risk and drives the preterm delivery thresholds discussed below.

Ductus Venosus and Computerised CTG: Timing in Early-Onset Disease

Below 32 weeks, once umbilical flow is abnormal, the question shifts from whether the fetus is compromised to when to intervene. The TRUFFLE trial tested whether ductus venosus changes could time delivery better than reduced CTG short-term variation.

Its primary intention-to-treat endpoint showed no significant difference between the monitoring strategies. In the prespecified analysis restricted to survivors, more infants timed by late ductus venosus changes were neurologically intact at 2 years than those timed by CTG variation alone.

That signal is tempered by how the trial ran: many fetuses were delivered on safety-net criteria or after 32 weeks, diluting the protocol effect. The trial group concluded that its outcome is best reproduced using ductus venosus Doppler and computerised CTG together, not either alone.

Cerebroplacental Ratio and Middle Cerebral Artery: The Late-Onset Question

In late fetal growth restriction, umbilical flow is often normal while the fetus redistributes blood toward the brain, lowering the cerebroplacental ratio. A low ratio is associated with adverse outcome, but systematic reviews find its standalone predictive accuracy modest, with substantial heterogeneity.

RATIO37 tested adding the cerebroplacental ratio to a routine late-term growth scan in a low-risk population. It did not reduce perinatal mortality, but it was associated with less severe non-neurological neonatal morbidity compared with growth assessment alone.

Several caveats limit how far this extends to diagnosed fetal growth restriction: the population was low-risk rather than growth-restricted, the trial was open-label, and about 30% of women declined the recommended planned delivery. The finding is hypothesis-supporting, not a management mandate.

Gestational-Age Thresholds: Immediate Versus Expectant Delivery

The GRIT trial randomised compromised preterm fetuses to immediate versus delayed delivery. It found no significant difference in death or severe disability, with only a non-significant excess of disability after immediate delivery in the most preterm subgroup (13% vs 5%).

At term, DIGITAT compared induction with expectant monitoring and met equivalence: composite adverse neonatal outcome was similar between arms, though induction delivered infants about 10 days earlier and slightly lighter. Together these trials underpin the gestational-age thresholds that guidelines now use.

Comparative Analysis of Fetal Growth Restriction Management

Across these approaches a consistent pattern emerges. Surveillance that identifies failing placental function appears to improve outcomes. Interventions that merely change the timing of delivery, once compromise is already recognised, show remarkably little difference between strategies.

The most reliable signal in fetal growth restriction is prognostic rather than therapeutic. Umbilical and venous Doppler stratify the risk of fetal death well, which is precisely why delivery thresholds are built around them. No randomised trial demonstrates that acting earlier on cerebral redistribution improves hard outcomes.

This is where guidelines part company. ISUOG incorporates ductus venosus and middle cerebral artery Doppler into surveillance and timing, reflecting the TRUFFLE experience. SMFM suggests against routine ductus venosus, middle cerebral artery, or uterine artery Doppler for management, citing the absence of trial evidence that it changes outcomes.

The practical reconciliation is by phenotype. Early-onset disease benefits from multivessel assessment where the expertise exists to interpret it. Late-onset disease is dominated by the challenge of detection, where cerebral Doppler adds prognostic information but no proven management benefit.

Quality and Consistency of the Evidence

• Moderate. Umbilical artery Doppler surveillance reduces perinatal death in high-risk pregnancies (meta-analysis of randomised trials, graded moderate by the reviewers).
• Moderate. Absent or reversed end-diastolic flow strongly predicts fetal death, consistently across a large observational meta-analysis with low heterogeneity.
• Low to moderate. Ductus venosus and computerised CTG timing may improve 2-year neurodevelopment in early fetal growth restriction, but on a single trial whose primary endpoint was not significant.
• Low. Acting on the cerebroplacental ratio may reduce morbidity, on one low-risk trial and a secondary endpoint; diagnostic accuracy alone is modest.
• Moderate. At term, induction and expectant monitoring are broadly equivalent for the fetus, on a dedicated equivalence trial.

What the Evidence Does Not Show

Practical Implications

The following reflect what the current evidence and guidelines support, not fixed directives; individual decisions depend on the whole clinical picture.

• Confirm the phenotype. Umbilical artery Doppler distinguishes placental fetal growth restriction from constitutional smallness and anchors the surveillance plan.
• Escalate by Doppler stage. Normal flow permits interval scanning; decreased, then absent, then reversed end-diastolic flow warrants progressively intensive monitoring and earlier planned delivery.
• Match timing to gestation. Guidelines point reversed flow toward roughly 30–32 weeks, absent flow toward 33–34 weeks, and milder disease with normal umbilical Doppler toward 37–39 weeks, with corticosteroids and magnesium sulfate for neuroprotection as indicated by gestational age.
• Use multivessel Doppler selectively. It informs early-onset timing where interpreted expertly, recognising that guidelines differ on its routine role.
• Near term, share the decision. Both induction and monitored expectant management are defensible options.

Evidence Grade and Bottom Line

Article Information and References