Clinical overview
Almost every obstetric decision you will ever make hangs off one number: the gestational age. Whether you offer aspirin prophylaxis, whether a fetus is "growth-restricted" or simply small, whether bleeding at 22 weeks is a previable loss or a delivery, whether to give antenatal corticosteroids, whether "term" induction is safe, how you interpret a CTG, when a baby is "post-dates" and at what threshold you intervene — all of it is anchored to gestational age. Get the dating wrong and you will misclassify growth, mistime delivery, over-investigate normal pregnancies and under-treat abnormal ones. In a South African context, where a large proportion of women book late, do not know their last menstrual period (LMP) reliably, and where first-trimester ultrasound is not universally available, accurate dating is one of the highest-yield skills a registrar develops.
This is a HOTS ("higher-order thinking") objective: the examiner is not testing whether you can recite that pregnancy lasts 40 weeks, but whether you can demonstrate the ability to assess gestational age — that is, choose the right method for the woman in front of you, apply it correctly, recognise when methods disagree, and act on the result. The discipline is simple to state and easy to get wrong: establish the estimated date of delivery (EDD) as early as possible, ideally by first-trimester ultrasound, fix it, and do not move it later in pregnancy without a very good reason. A late-booking woman with no early scan, common in district clinics, forces you to reason from less reliable data and to say so honestly in your documentation.
Core knowledge
Definitions and the convention
Gestational age is counted in completed weeks and days from the first day of the last menstrual period (LMP), not from conception. This is a convention: it assumes a regular 28-day cycle with ovulation and conception around day 14, so gestational age is roughly two weeks more than the true embryonic (post-conceptional) age. A pregnancy is term from 37+0 to 41+6 weeks; preterm is birth before 37+0; post-term is 42+0 weeks or beyond. The EDD by convention is 280 days (40 weeks) from the LMP — Naegele's rule (LMP + 7 days – 3 months + 1 year) is the arithmetic shortcut, though it embeds the 28-day-cycle assumption and is only as good as the recalled LMP.
These week boundaries are not arbitrary trivia; they are decision thresholds. Viability counselling, antenatal corticosteroid windows, magnesium sulphate for neuroprotection, the definition of growth restriction, the timing of "term" elective delivery, and the post-dates surveillance/induction decision all key off them. An error of even one to two weeks can move a fetus across a management boundary.
Why menstrual dating is unreliable
LMP dating fails for predictable reasons, and you should be able to list them: irregular or anovulatory cycles, recent hormonal contraception (especially recent depot medroxyprogesterone acetate, where ovulation may be delayed for months), lactational amenorrhoea, early-pregnancy bleeding mistaken for a period, and simple recall error. Conception does not reliably occur on cycle day 14; in long or irregular cycles it is later, so LMP systematically over-estimates gestation in those women. Standard teaching is that menstrual dating, even when the LMP is "certain," is materially less accurate than a first-trimester scan, and across a population it tends to over-diagnose post-term pregnancy.
Why ultrasound dating works — and its ceiling
Ultrasound dates the pregnancy by measuring how big the embryo or fetus is and reading off a reference chart that maps size to gestational age. It works best early because in the first trimester biological variation in size is small — every normal embryo grows at almost the same rate — so size predicts age tightly. As pregnancy advances, normal fetuses diverge in size (genetics, sex, growth pathology), so a measured biometry could represent a normal fetus of one age or an abnormally grown fetus of another. This is the single most important concept in this objective: ultrasound dating accuracy degrades steadily with advancing gestation, so the earliest adequate scan wins.
The measurements, in order of when they are used:
- Mean sac diameter (MSD) — earliest, before an embryo is visible, least accurate; used to confirm intrauterine pregnancy more than to date.
- Crown–rump length (CRL) — the gold standard for dating. Used roughly from 6–7 weeks until the CRL reaches about 84 mm (around 13–14 weeks). A correctly measured CRL dates to within a few days.
- Biparietal diameter (BPD) and head circumference (HC) — used from the second trimester once the CRL is out of range.
- Femur length (FL) and abdominal circumference (AC) — AC is the most growth-sensitive measurement and the least reliable for dating, but is essential for growth assessment.
In the second trimester, dating typically uses a composite of HC, BPD and FL. By the third trimester, biometry-based dating is unreliable (commonly quoted as accurate only to within roughly ±3 weeks) precisely because of biological size divergence — which is also why you must never "re-date" a small third-trimester fetus and thereby convert genuine growth restriction into reassuring small-for-dates. (Specific gestation-window error margins are standard teaching/guideline figures — see notes; confirm against the local protocol before quoting exact numbers.)
Figure I6.1 — Gestational-age dating hierarchy: LMP is a convention, CRL is the early gold standard, second-trimester biometry is less precise and third-trimester scans assess growth rather than age.
Assessment
This section is the heart of a "demonstrate the ability to assess" objective. Work through it as a clinician would, in order.
Step 1 — History
Take a focused menstrual and reproductive history:
- LMP: first day, and crucially how certain she is. Was it a normal period in timing and flow, or light/atypical (which may have been implantation bleeding)?
- Cycle: usual length and regularity. A 35-day cycle pushes ovulation and conception later, so a "certain" LMP would over-estimate gestation by about a week.
- Contraception: recent combined pill, and especially recent injectable progestogen (DMPA/Nur-Isterate), which is common in South Africa and can delay return of ovulation for months — LMP dating after recent injectables is treacherous.
- Conception clues: known date of intercourse, a positive home test date, assisted-conception cycles (IVF dates the pregnancy almost exactly and overrides everything else).
- Booking history: any prior scan, and at what gestation it was performed — an earlier scan, if adequate, is the better dater even if a later one is "more recent."
Step 2 — Examination
Clinical examination is a crude dater and should never override an adequate early scan, but it remains essential in the late-booking, no-scan woman:
- Symphysis–fundal height (SFH) in centimetres approximates gestational age in weeks from about 24 weeks (the classic "±2–3 cm" rule). It is affected by maternal habitus, liquor volume, fetal lie, multiple pregnancy, fibroids and a full bladder, so it is a screening tool for growth, not a precise dater.
- Bimanual/abdominal uterine size in the first and early second trimester: a uterus "the size of" a given gestation (e.g. reaching the pelvic brim at ~12 weeks, the umbilicus at ~20 weeks) gives a rough estimate when no scan is available.
- Quickening (first felt fetal movements, classically ~18–20 weeks in a primigravida, ~16–18 in a multigravida) and the gestation at which the fetal heart was first audible by Pinard/Doppler are weak corroborating signs only — record them, but do not date a pregnancy on them.
Step 3 — The dating ultrasound (the decisive step)
Where available, an early scan is the definitive method. The clinical reasoning you must demonstrate:
- Offer a dating scan early. antenatal-booking should trigger a first-trimester scan wherever feasible. The earliest adequate scan is the one that dates the pregnancy for the rest of the pregnancy.
- Use the right measurement for the gestation: CRL up to ~84 mm; thereafter a HC/BPD/FL composite. Measure correctly — a mid-sagittal CRL in a neutral fetal position, calipers on true crown and true rump, no flexion or hyperextension.
- Generate the EDD and FIX it. Once an adequate early scan has set the EDD, that EDD stands. A later scan that gives a "different" gestation in the third trimester is telling you about growth, not age.
- Multiple pregnancy: date by the larger CRL (to avoid dating from an already-compromised twin), and record chorionicity at the same time — both are dating-window decisions, and chorionicity is hardest to assign late.
Step 4 — Reconciling LMP with scan
This is the judgement the examiner wants to see. The principle, consistent with antenatal-care guidance:
- If the scan-derived EDD and the LMP-derived EDD agree closely, the LMP is corroborated — either may be quoted.
- If they disagree beyond an accepted margin, the early scan dates the pregnancy and the LMP is discarded. The margin for "redating" is tightest in the first trimester (a small difference redates) and widens with gestation; standard teaching uses a several-days threshold around the CRL window and a larger window in the second trimester. (Exact discrepancy thresholds for redating are guideline-specific — see notes; do not quote a precise day-count without checking the local/NICE protocol.)
- Document explicitly which method you used and why, so the next clinician inherits a single agreed EDD rather than two competing dates.
Step 5 — The neonate (dating after birth)
When a baby is born with no reliable antenatal dating — a real scenario after unbooked or precipitate delivery — gestational age is estimated postnatally. The New Ballard Score (neuromuscular + physical maturity criteria, usable from extreme prematurity) and the older Dubowitz assessment are the recognised tools; they carry an error of roughly ±2 weeks and are a fallback, not a substitute for antenatal dating. (Postnatal scoring error margins are standard teaching — see notes.)
Figure I6.2 — Gestational-age assessment workflow: history, examination, best available dater, LMP-scan reconciliation and documentation of one agreed EDD with uncertainty where needed.
Management
"Management" of gestational-age assessment means converting an accurate EDD into safe action and recognising when uncertainty itself must be managed.
Apply the EDD to every downstream decision
A fixed, accurate EDD drives:
- Aspirin prophylaxis for pre-eclampsia risk, started in the first trimester — see hypertension-in-pregnancy and pre-eclampsia-and-hellp; the window closes if dating is late.
- Aneuploidy screening windows (combined first-trimester test, NIPT/cfDNA timing) — see antenatal-screening and down-syndrome-counselling — which are gestation-bounded.
- Viability and corticosteroid/neuroprotection counselling at the margins of viability — accurate dating is the difference between offering active resuscitation and palliative care, see high-risk-pregnancy-risks.
- Growth surveillance: a correct EDD is the denominator against which biometry is plotted — get it wrong and you misclassify intrauterine-growth-restriction or macrosomia.
- Timing of delivery in complicated pregnancy and the post-term decision below.
The late-booking / no-early-scan woman (the SA default)
In much of South African practice the woman presents in the second or third trimester with no early scan and an uncertain LMP. The disciplined approach:
- Obtain the best available scan now and date from it, accepting the wider error margin and recording that margin explicitly (e.g. "dated by second-trimester biometry, EDD ±2 weeks").
- Do not silently invent certainty. If dating is poor, say so in the notes and let downstream decisions inherit the uncertainty — never quietly assign a false precise EDD that later clinicians will trust. (This is the project hard rule: never silently corrupt the data the next clinician depends on.)
- Use the earliest scan ever performed, not the latest, as the dater.
- The South African NDoH National Integrated Maternal and Perinatal Care Guideline (NDoH, 2024) structures booking and dating within the antenatal schedule and levels of care; basic obstetric ultrasound is increasingly available at district level but is not universal, so SFH-based clinical dating and growth monitoring remain frontline tools at primary care — see sa-maternity-guidelines and partogram-use.
Post-term pregnancy — managing the EDD at the late end
Once a fixed EDD reaches the post-dates window, gestational-age assessment becomes an active management decision. Standard practice is to offer membrane sweeping and to offer induction of labour around 41 weeks to reduce the risk of stillbirth and meconium aspiration associated with prolonged pregnancy, with increased fetal surveillance (liquor volume, CTG) for those who decline or await spontaneous labour. Because post-term diagnosis is so dependent on accurate dating, the single most effective intervention against unnecessary post-dates induction is a good early scan — over-estimation of gestation from LMP is a classic cause of "false" post-dates. (Exact induction-offer gestation and surveillance schedule are guideline-specific — see notes; align to the local NDoH protocol.)
Figure I6.3 — Dating pitfalls that change management: late redating, unreliable LMP, IVF dating, twin dating, false post-dates and false precision all alter downstream care.
Red flags / pitfalls
- Re-dating a small third-trimester fetus. The commonest and most dangerous error. A fetus that "measures small" late is not "younger" — moving the EDD later converts genuine intrauterine-growth-restriction into reassuring "small-for-dates" and removes the surveillance that might prevent a stillbirth. Never re-date in the third trimester.
- Letting a late scan override an adequate early scan. The early scan dates; the late scan assesses growth. Quoting the most recent scan's "gestation" as the age is wrong.
- Trusting a "certain" LMP after recent injectable contraception. DMPA/Nur-Isterate is widespread in SA and delays ovulation; the LMP can be months out. Corroborate with a scan.
- Dating IVF pregnancies from LMP or scan. Assisted-conception pregnancies are dated from the known fertilisation/transfer date, which is essentially exact and overrides ultrasound.
- Dating multiples from the smaller twin. Date from the larger CRL, and capture chorionicity in the same early scan — both windows close fast.
- Mistaking an over-estimated EDD for true post-dates. Inducing a woman who is actually 40 weeks but "dated 42" by an unreliable LMP exposes her to the harms of intervention for no benefit.
- Recording two competing EDDs. Decide, document the method and reason, and carry one agreed EDD forward so no one downstream guesses.
- Manufacturing false precision in the unbooked woman. When dating is genuinely uncertain, document the uncertainty and its margin — do not assign a clean EDD the record cannot support.
Evidence anchors
- South African NDoH — National Integrated Maternal and Perinatal Care Guideline (NDoH, 2024). The SA obstetric source of truth: structures antenatal booking, the dating/booking visit, ultrasound availability across levels of care, and growth monitoring; anchors dating within the local resource context.
- NICE NG201 — Antenatal Care (2021). Offer an early dating ultrasound; use crown–rump length up to ~84 mm and a head-measurement composite thereafter; redate where scan and LMP disagree beyond the accepted margin. (Confirm exact CRL/HC thresholds and redating day-counts against the current guideline text before quoting them numerically.)
- RCOG GTG 31 — Small-for-Gestational-Age and Growth-Restricted Fetus, with ISUOG Doppler/biometry standards. Underpins the discipline of separating dating (early, fixed) from growth (serial, late) and the prohibition on third-trimester redating.
- Saving Mothers / NCCEMD (latest triennium). Frames why accurate dating matters in the SA mortality context — correct gestational classification underlies timely intervention for the leading direct causes (hypertension, haemorrhage) where delivery timing is decisive.