How you ended up having this conversation

Usually the tape measure. NICE guideline NG201 recommendation 1.2.30 asks for symphysis fundal height measurement at each antenatal appointment after 24+0 weeks, plotted on a growth chart. Recommendation 1.2.31 is the trigger for what happened next: "If there are concerns that the symphysis fundal height is large for gestational age, consider an ultrasound scan for fetal growth and wellbeing."

Note the verb. For small-for-dates, NICE says offer a scan; for large-for-dates, it says consider one. That difference reflects how much less the scan changes the outcome at this end of the distribution.

What "large" means numerically

ACOG Practice Bulletin No. 216 groups suspected macrosomia into three bands for the purpose of quantifying risk: "Category 1 (4,000-4,499 g); category 2 (4,500-4,999 g); category 3 (5,000+ g)", compared with a reference group of 3,000-3,999 g. Risks rise across those bands rather than switching on at a single number.

The bulletin's own summary of the problem is worth having: "As birth weight increases, the likelihood of labor abnormalities, shoulder dystocia, birth trauma, and permanent injury to the newborn increases", and its stated purpose includes addressing "the accuracy and limitations of methods for estimating fetal weight".

How accurate is the scan? Not very.

This is the part that most changes how you should weigh what you have been told. The RCOG's information on shoulder dystocia states: "Ultrasound scans are not good at telling whether you are likely to have a large baby and therefore they are not recommended for predicting shoulder dystocia, if you have no other risk factors."

Third-trimester estimated fetal weight carries a meaningful margin of error in both directions. A scan reporting 4.2 kg is not a measurement of your baby; it is an estimate with a range around it. That does not make it useless - it makes it one input among several, rather than a verdict.

If you do not have diabetes: what NICE actually offers

NICE guideline NG207 recommendation 1.2.24 is unusually explicit about uncertainty, and it is worth reading in full because it is the script your clinician should be working from. Discuss with women without diabetes and with suspected fetal macrosomia that:

"the options for birth are expectant management, induction of labour or caesarean birth"
"there is uncertainty about the benefits and risks of induction of labour compared to expectant management, but: with induction of labour the risk of shoulder dystocia reduced compared with expectant management; with induction of labour the risk of third- or fourth-degree perineal tears is increased compared with expectant management"
"there is evidence that the risk of perinatal death, brachial plexus injuries in the baby, or the need for emergency caesarean birth is the same between the 2 options"
"they will also need to consider the impact of induction on their birth experience and on their baby"

NICE closes with: "Discuss the options for birth with the woman, taking into account her individual circumstances and her preferences, and respect her decision."

Read that trade-off carefully. Induction lowers one risk (shoulder dystocia) and raises another (severe perineal tearing), and does not change the outcomes people fear most - perinatal death, nerve injury to the baby, or emergency caesarean. There is no clinically obvious answer here, which is exactly why the decision is yours.

One point of tension worth naming rather than hiding: the RCOG's shoulder dystocia information, published in 2013, states that "if you don't have diabetes, early induction of labour does not prevent shoulder dystocia, even if your baby is suspected to be large", while NICE's 2021 guideline reports a reduction in shoulder dystocia with induction. NICE is the later assessment of the evidence. Both are quoted here so you can see the disagreement rather than be handed one side of it.

If you do have diabetes: different guidance

NICE points explicitly to its diabetes guideline for this group (NG207, recommendation 1.2.25), and NG3 sets a different frame. Recommendation 1.4.7: "For pregnant women with diabetes who have an ultrasound-diagnosed macrosomic fetus, explain the risks and benefits of vaginal birth, induction of labour and caesarean section."

Timing is also already decided for most of this group. Women with gestational diabetes are advised to give birth "no later than 40 weeks plus 6 days" (NG3 recommendation 1.4.4), and women with type 1 or type 2 diabetes between 37 weeks and 38 weeks plus 6 days (1.4.2). The RCOG adds that if you have diabetes, "you will usually be offered early induction of labour or planned caesarean section. This will reduce the risk of shoulder dystocia."

Diabetes also changes the monitoring: NG3 recommendation 1.3.34 asks for ultrasound monitoring of fetal growth and amniotic fluid volume every 4 weeks from 28 to 36 weeks.

Shoulder dystocia, in proportion

This is the specific fear behind most large-baby conversations, so here are the numbers from the RCOG.

"Shoulder dystocia occurs in about one in 150 (0.7%) vaginal births."
"Shoulder dystocia is more likely with large babies but nevertheless there is no difficulty delivering the shoulders in the majority of babies over 4.5 kg (10 lb)."
"Half of all instances of shoulder dystocia occur in babies weighing less than 4 kg (about 9 lb)."
"In the majority of cases, the baby will be born promptly and safely."

That third figure is the one that reframes everything. If half of shoulder dystocia happens in babies under 4 kg, then predicting it by size alone was never going to work. Other risk factors the RCOG lists include previous shoulder dystocia, diabetes, BMI of 30 or more, induced labour, long labour and assisted vaginal birth.

Questions worth asking

If you are being offered a decision, these get you to the substance: what estimated fetal weight did the scan give, and what is the margin of error on it? Which growth chart was used? Am I being offered induction because of the estimated size alone, or because something else has changed? And, if induction is recommended: at what gestation, and what is the alternative plan if I choose to wait?

NICE's own framing gives you permission to ask all of these. The recommendation ends by telling clinicians to respect your decision, which is only meaningful if the information behind it has actually been shared.

Large for Gestational Age