Deliver evidence-based antenatal care and routine screening

In one line

Antenatal care is a structured, goal-directed screening and surveillance programme that runs from early booking to delivery; its modern justification is hard-edged — a minimum of eight contacts reduces perinatal mortality compared with the older four-visit model — and in South Africa it is delivered through BANC-Plus, where universal HIV and syphilis screening, early dating, and pre-eclampsia risk assessment with aspirin matter more than the number of visits the woman actually attends.

The groundwork — what a booking history covers, how to date a pregnancy, the normal physiology of the visit — is assumed here. Revise it at the antenatal booking visit. This chapter is about the consultant judgement: which screening tests earn their place and on what evidence, where the major schedule models diverge and why, and how to run a programme that detects the women who will deteriorate before they do.

Mechanism & pathophysiology

Antenatal care does not treat a disease; it converts an asymptomatic population into a screened one, and almost every controversy in the field is a screening-theory problem wearing obstetric clothes. The logic that governs it is the logic of secondary prevention: a test is worth doing only if the condition has a detectable preclinical phase, the test performs adequately in that phase, and acting on a positive result changes outcome. Each routine investigation in pregnancy has to clear that bar, and several historically entrenched ones do not.

The deeper rationale for frequency — the eight-versus-four argument — is a Bayesian one. Pregnancy complications are not static; pre-eclampsia, fetal growth restriction, malpresentation and gestational diabetes declare themselves across the third trimester, so the pre-test probability of finding a serious abnormality rises as gestation advances. That is precisely why the WHO model loads contacts into the third trimester (30, 34, 36, 38 and 40 weeks) rather than spacing them evenly — the surveillance density should track the hazard rate. The four-visit Focused ANC model thinned the late-pregnancy net at the exact point the disease incidence was climbing, which is the mechanistic explanation for the perinatal-mortality signal that eventually overturned it.

Dating sits upstream of everything. An accurate gestational age, established by first-trimester crown–rump length, is the denominator for every later judgement: whether growth is restricted, whether labour is preterm or post-term, whether a screening result is interpretable. Get the dates wrong and every downstream surveillance decision inherits the error — a falsely "post-term" pregnancy induced unnecessarily, or a genuinely growth-restricted fetus reassured as small-for-dates because the menstrual dates were optimistic. The accuracy of CRL (±5–7 days to 13⁺⁶ weeks) versus the much wider error of later biometry or recalled last menstrual period is the reason early booking is the single highest-leverage act in the schedule.

Risk stratification is the organising principle that makes a one-size schedule safe. The model assumes a low-risk population by default and uses the booking assessment to pull out the women who do not belong in it — prior pre-eclampsia, prior stillbirth, chronic hypertension, diabetes, significant medical disease, HIV, multiple pregnancy — and route them to a more intensive, often specialist-led, pathway. The midwife-led "normal" track and the consultant-led "high-risk" track are not two services but two arms of one triage system, and the booking visit is the triage point. Stratification is also not a one-time act: a woman can enter the low-risk track and earn her way out of it at any contact — a rising blood pressure, a lagging fundal height, a new medical diagnosis — which is why the schedule re-screens rather than simply re-visits.

The final mechanistic point is that screening tests have characteristics, and a consultant reasons in those terms rather than treating a test as a binary truth. Symphysis–fundal height has poor sensitivity for growth restriction; a normal smear in a woman with a visible cervical lesion is a false reassurance; a single non-reactive HIV test at booking does not exclude an infection acquired in the third trimester. The discipline is to know each test's blind spot and to cover it — with serial measurement, a lower threshold for definitive imaging, and the scheduled repeat test — rather than to trust a single negative result to clear a whole pregnancy.

Assessment

The booking contact is the most consequential of the schedule and should happen early — NICE asks for the first appointment by 10⁺⁰ weeks; the WHO model's first contact is within the first trimester (up to 12 weeks). In South Africa the persistent reality is the opposite: a large fraction of women book in the second trimester or later, which forfeits accurate dating, first-trimester aneuploidy screening, and the window in which aspirin prophylaxis works. Late booking is not a footnote in SA practice — it is the dominant failure mode, and counselling the community to present early is a public-health intervention in its own right.

The booking history and examination establish risk, not just background. Beyond standard obstetric, medical, surgical, drug and social history, the assessment is actively looking for the stratifiers above, for previous pregnancy complications that recur, for medications that need review against teratogenicity, and — a mandatory SA element folded into BANC-Plus — sensitive enquiry for intimate-partner violence. Booking BMI, baseline blood pressure and urinalysis anchor later trend interpretation; a single late blood pressure has no baseline to be read against.

Gestational dating is by ultrasound where available: crown–rump length in the first trimester, head circumference in the early second. The dating scan reduces inductions for apparent post-term pregnancy and is the reference against which all later growth assessment is judged. Where early ultrasound is not accessible — a real constraint at many SA district facilities — symphysis–fundal height and clinical estimation carry more weight, and the uncertainty must be acknowledged rather than papered over.

Booking bloods and their interpretation:

• Full blood count / haemoglobin — to detect and treat anaemia (common in SA from iron deficiency, and compounded by HIV); ferritin clarifies iron status where the picture is mixed. Anaemia at booking is both a direct risk (cardiac reserve in the event of haemorrhage) and a marker.
• Blood group and red-cell antibody screen — identifies the rhesus-negative woman who needs anti-D and antibody surveillance, and detects clinically significant atypical antibodies that threaten the fetus. (The alloimmunisation pathway is developed in rhesus-alloimmunisation.)
• HIV — universal, opt-out testing at booking, with repeat testing later in pregnancy because seroconversion during pregnancy is a major driver of vertical transmission in a high-incidence setting. SA uses PVT (prevention of vertical transmission) framing; a positive result triggers ART and the full PVT pathway, not merely a note in the file.
• Syphilis — universal screening at booking, with a second test at 32–34 weeks if the first was negative, because untreated maternal syphilis causes stillbirth and congenital syphilis that early treatment prevents. Point-of-care dual HIV/syphilis rapid tests are increasingly used at SA primary-care antenatal facilities to close the treatment-gap that lab turnaround creates.
• Hepatitis B surface antigen — identifies the infant needing birth-dose vaccine ± immunoglobulin and the mother who may need antiviral therapy to reduce transmission.
• Rubella immunity — does not change this pregnancy (the vaccine is live and contraindicated antenatally) but identifies the susceptible woman for postpartum vaccination, closing the loop for the next pregnancy.

Ongoing surveillance at each contact is deliberately narrow and goal-directed: blood pressure and urine for proteinuria (the pre-eclampsia net), symphysis–fundal height plotted against gestation (the growth net, with growth scans triggered by a static or lagging SFH or by risk factors), enquiry about fetal movements, and from the early third trimester, determination of presentation — with malpresentation at term routed to discussion of external cephalic version. Each measurement is a screening test with its own characteristics: SFH is insensitive for growth restriction, which is why a customised growth chart and a low threshold for biometry matter in the at-risk woman.

Interpreting these trends is the consultant skill, not the measuring. A blood pressure that is "normal" at 28 weeks but 20 mmHg above the booking baseline is a developing-pre-eclampsia trajectory, not a normal reading — which is the entire argument for recording the booking baseline. A fundal height that plateaus across two contacts means more than a single point below the 10th centile, because it is the velocity of growth, not a static percentile, that distinguishes the constitutionally small fetus from the failing one. Proteinuria on dipstick is a screening trigger, not a diagnosis — a positive dip prompts quantification (protein:creatinine ratio) rather than immediate action on the strip alone. And reduced fetal movements are a symptom to be taken at face value: the reduction reported by the mother outperforms most objective surveillance as an early signal, and an algorithm that reassures her without assessment is the recurrent route to a preventable stillbirth.

A practical SA caveat runs through all of this surveillance: it assumes the woman returns. A net that depends on serial measurement leaks badly when attendance is irregular, when a woman moves between her rural home and an urban relative, and when a single contact must do the work of several because the next one may not happen. That argues for completing as much of the screening agenda as possible at each attendance rather than spreading it thinly across a schedule the woman may not keep — the opposite of the leisurely, evenly-spaced model a well-resourced low-prevalence system can assume.

Management

The "management" of antenatal care is the design of the programme itself — what happens at each contact, how risk is escalated, and where SA practice sits.

The structure of a contact: immediate → ongoing → forward-looking.

• Immediate at every visit: the surveillance set above (BP, urine, SFH, fetal movements, presentation from ~36 weeks), plus a check that scheduled screening for that gestation has been done. A visit that takes a blood pressure but does not ask whether the syphilis retest or the GTT was done is a visit half-wasted.
• Ongoing across the schedule: the timed interventions — first-trimester dating and aneuploidy screening, the mid-trimester anomaly scan (~18–22 weeks; cross-referenced to prenatal-screening-aneuploidy and structural-anomaly screening in the P domain), gestational diabetes screening in the late second trimester, anti-D prophylaxis for the rhesus-negative woman, and the tetanus and other recommended immunisations.
• Forward-looking / long-term: danger-signs education at every contact (the woman must know that headache, visual disturbance, reduced fetal movements, vaginal bleeding, abdominal pain, fluid loss or fever mean come in now), anticipatory guidance on labour and delivery planning, infant feeding counselling (central in the HIV-positive woman), and the postpartum agenda seeded antenatally — contraception, postpartum rubella vaccination, GDM follow-up.

Supplementation and immunisation are the population-level preventive layer that the surveillance set sits on top of. The WHO model recommends daily iron and folic acid through pregnancy — iron because antenatal anaemia is common and consequential in SA, folate to reduce neural-tube defects (most effective when started preconceptionally, which late booking forfeits) — and calcium supplementation in populations with low dietary intake as a pre-eclampsia-reduction measure, a recommendation with particular SA relevance. On immunisation, the live rubella and varicella vaccines are contraindicated in pregnancy and deferred to the postpartum period, while inactivated vaccines are given antenatally where indicated: tetanus-containing vaccine to prevent neonatal tetanus, and influenza and other recommended inactivated vaccines according to current national policy. The hepatitis B birth-dose programme for the infant of a surface-antigen-positive mother is set up antenatally, not improvised at delivery.

The forward-looking agenda also includes deciding, antenatally, where the woman should deliver. A risk flag changes the planned place of birth — a woman with a previous caesarean, a medical comorbidity, or a flagged fetal concern should not be planning a clinic delivery far from operative capability. Determining and communicating the intended level of delivery is an antenatal decision that, made late or not at all, produces the emergency transfer in labour that the whole screening programme exists to pre-empt.

Pre-eclampsia risk assessment and aspirin is the highest-yield preventive act of the schedule and belongs at booking. Stratify by history and clinical factors (prior pre-eclampsia, chronic hypertension, diabetes, renal disease, autoimmune disease, multiple pregnancy as high-risk; nulliparity, age, BMI, family history, interpregnancy interval as moderate-risk that aggregate). Women at high risk — or with enough moderate factors — should start low-dose aspirin from before 16 weeks and continue to ~36 weeks; the threshold and the prevention argument are developed in pre-eclampsia-prevention-aspirin. The dose is the live point: the ASPRE trial's effect was achieved with 150 mg at night, and the meta-analytic signal suggests doses below 100 mg underperform — yet much routine practice and many supply chains default to lower doses. Starting aspirin late (after 16 weeks) or at too low a dose are the two ways this intervention is quietly rendered ineffective.

Gestational diabetes screening is where guidelines openly disagree, so the consultant must state a defensible local position rather than recite one body's cut-offs. The two schools are universal screening (everyone gets a glucose test) versus risk-factor-based screening (test only those with risk factors — prior GDM, prior macrosomia, raised BMI, family history, ethnicity, PMOS); and one-step (a single 75 g OGTT read against the IADPSG/HAPO-derived thresholds) versus two-step (a glucose challenge, then a diagnostic OGTT only in those who screen positive). HAPO is the pivot: it showed a continuous, threshold-free relationship between maternal glucose and adverse outcome, which is the intellectual basis for the lower, one-step IADPSG criteria — but those criteria roughly double the diagnosed prevalence, and whether that extra diagnosis translates into better outcomes (rather than more intervention) is exactly the unsettled question. In SA, with a high background burden of obesity and type 2 diabetes, the pragmatic position is risk-factor-based screening with a 75 g OGTT in the late second trimester (around 24–28 weeks), accepting that purely risk-based approaches miss some cases.

Group B Streptococcus is the clearest example of a screening question with no single right answer. The two approaches are universal culture-based screening (the US model: rectovaginal culture at 36–37 weeks, intrapartum antibiotics to carriers) versus a risk-factor-based strategy (the UK/RCOG model: intrapartum antibiotic prophylaxis triggered by risk factors — previous affected infant, GBS detected in this pregnancy, preterm labour, intrapartum fever or suspected chorioamnionitis — without routine screening). The risk-based approach treats fewer well women but accepts that it will miss some carriers; SA practice has historically been risk-based, consistent with resource constraints and the RCOG position. The exam-relevant point is that universal screening is not self-evidently superior — it is a value judgement about over-treatment versus missed cases that reasonable systems decide differently.

Risk escalation — when to leave the normal track. The schedule is built for low-risk women, and the management skill is recognising the trigger to step up: a booking stratifier (the conditions listed under risk stratification), or a new finding on surveillance — rising blood pressure or proteinuria, a static or lagging SFH, reduced fetal movements, malpresentation at term, an abnormal screening result, intercurrent illness. Escalation in SA is geographic as well as clinical: from midwife-led BANC-Plus at the clinic, to the district hospital, to a regional or tertiary unit, and the consultant's task is to escalate early enough that the journey — often long and on poor roads — does not consume the safety margin. The danger is a woman flagged at the clinic but escalated too slowly to reach definitive care before she deteriorates.

Guidelines compared

The substantive divergences are three. First, visit number: WHO sets a floor of eight to reduce perinatal mortality; NICE goes higher (ten for nulliparous women) because the UK can afford intensity; SA adopted the WHO eight as the achievable evidence-based standard. Second, what is screened universally: HIV and syphilis are universal-and-repeated in SA in a way they are not in a low-prevalence NICE setting — the screening package is correctly different because the epidemiology is different. Third, GBS, where the UK/SA risk-based stance and the US universal-culture stance genuinely disagree, and neither is the demonstrably "correct" answer.

The change worth flagging is recent and directional: the move from the 2002 four-visit Focused ANC model to the WHO 2016 eight-contact model (adopted in SA from April 2017 as BANC-Plus) reversed two decades of "fewer, goal-directed visits are enough" orthodoxy, on the strength of a perinatal-mortality signal in the trial data.

The evidence & the controversy

The headline shift is that fewer is not better, and the four-visit model was a mistake that took fifteen years to correct. The 2002 Focused ANC model — championed precisely because it concentrated care into fewer, goal-directed visits — rested on trials that were not powered for perinatal mortality. When Dowswell's Cochrane review pooled them, reduced-visit packages carried a perinatal-mortality risk ratio of 1.14 (95% CI 1.00–1.31) overall, and in low- and middle-income countries specifically 1.15 (95% CI 1.01–1.32) — a small but real excess of dead babies in the very settings that had adopted the streamlined model to save resources. WHO's 2016 recommendation E.7 — eight contacts to reduce perinatal mortality — is the response. The uncomfortable lesson for a resource-limited service is that the efficiency argument for fewer visits was paid for in perinatal deaths, and that the binding constraint in SA is now not the recommended number of contacts but whether women attend them and whether each contact does its screening job.

The second live controversy is gestational diabetes: whether the HAPO-derived, one-step IADPSG criteria — which lower the diagnostic threshold and roughly double prevalence — improve outcomes or merely medicalise more pregnancies. HAPO established the continuous glucose–outcome relationship beyond dispute; what it could not settle is where to draw a treatment line on a continuum, and trials of treating milder hyperglycaemia show benefit that is real but modest. In a high-burden SA context the argument is sharpened by capacity: a screening strategy is only worth adopting if the system can deliver the dietary advice, monitoring and insulin that a positive result implies, and a doubled diagnosis rate the service cannot resource is not obviously a gain.

A third thread is the quiet failure of well-evidenced interventions delivered badly. Aspirin for pre-eclampsia prevention is a genuinely effective intervention (ASPRE) that is routinely undermined in practice by starting it after 16 weeks or at sub-therapeutic doses; iron and calcium supplementation, recommended in the WHO model partly for SA-relevant deficiency and pre-eclampsia-risk reasons, are only as good as adherence. The consultant point is that the evidence base is increasingly settled while the implementation gap widens — and in SA the implementation gap (late booking, missed retests, supply interruptions, slow escalation) is where babies are actually lost, not in the choice between guidelines.

A current topical strand worth holding lightly: the WHO model's explicit reframing of ANC around a "positive pregnancy experience" — respectful care, IPV enquiry, woman-centredness, mental-health attention — is sometimes dismissed as soft alongside the hard screening tests. It is not. Disrespectful or unsafe-feeling care is a documented driver of late booking and non-attendance, which is the upstream cause of the perinatal deaths the eight-contact model is meant to prevent; the experiential and the biomedical agendas are the same agenda seen from two ends.

Worked viva — structuring the answer

A typical stem: "A 24-year-old, gravida 2 para 1, books at a district clinic at 26 weeks. Her previous pregnancy ended in a fetal death at term. Booking BP 138/88, urine dipstick trace protein, BMI 34, HIV non-reactive. How would you manage her antenatal care?" A high-scoring answer runs:

1. Recognise she is not low-risk and leave the standard track. A previous unexplained term stillbirth, a raised BMI and a borderline booking blood pressure put her on a high-risk, more intensive pathway with specialist input, not the routine BANC-Plus schedule.
2. Repair what late booking has cost. Establish gestational age as accurately as possible — a 26-week scan dates poorly, so commit to serial growth surveillance rather than relying on a single biometry. The first-trimester aneuploidy window and the pre-16-week aspirin window have both closed, and that should be stated honestly.
3. Address pre-eclampsia risk concretely. With prior adverse outcome, raised BMI and a borderline BP she warrants pre-eclampsia surveillance; aspirin started now is later than ideal but the booking baseline BP must be recorded so subsequent readings are interpretable, and the trace proteinuria quantified with a protein:creatinine ratio rather than acted on from the strip.
4. Run the screening agenda she has missed. Booking bloods including HIV (with a planned repeat later in pregnancy despite the non-reactive result), syphilis at booking and a repeat at 32–34 weeks, blood group and antibodies, haemoglobin, and GDM screening with a 75 g OGTT given her BMI and prior loss.
5. Plan the surveillance and the place of birth. Growth scans driven by SFH and risk, fetal-movement counselling, presentation from the early third trimester, and an early decision that her delivery belongs at a hospital with operative capability — not the clinic.
6. Justify from evidence. The eight-contact model and its perinatal-mortality basis (Dowswell/WHO), aspirin from ASPRE (and why dose and timing matter), and the SA-specific universal HIV/syphilis screening rationale — framed around completing the screening agenda in a woman who books late and may attend irregularly.

Landmark trials & key evidence

A worked figure from ASPRE: preterm pre-eclampsia fell from 4.3% to 1.6%, an absolute risk reduction of 2.7% in this screen-enriched high-risk group, so the number needed to treat ≈ 1/0.027 ≈ 37 high-risk women given 150 mg aspirin to prevent one case of preterm pre-eclampsia — a strong yield from a cheap, available drug, which is why getting the dose and the timing right (150 mg, before 16 weeks) is the whole game.

The schedule models themselves rest on the WHO 2016 ANC recommendations (recommendation E.7: a minimum of eight contacts to reduce perinatal mortality) and, in the UK, NICE NG201 (2021); SA's BANC-Plus rollout from April 2017 is documented by Hlongwane et al in the Bulletin of the World Health Organization. The first-trimester dating evidence — that routine early ultrasound reduces inductions for apparent post-term pregnancy and dates to ±5–7 days by CRL — is the Whitworth Cochrane review.

Exam traps & red flags

• Treating "number of visits" as the point. The eight-contact model is a floor, not a target; the evidence (Dowswell) is about not going below it. The binding problem in SA is attendance and content, not the recommended count — a programme that books women late and skips retests fails regardless of the schedule on paper.
• Late booking treated as administrative. Late presentation forfeits accurate dating, first-trimester aneuploidy screening, and the pre-16-week window for aspirin. In SA it is the dominant failure mode; naming it as a clinical loss, not a logistics issue, is part of the answer.
• Aspirin too late or too low. Starting after 16 weeks, or at a sub-therapeutic dose, squanders an intervention that works at 150 mg from before 16 weeks (ASPRE). This is the commonest way a correct decision is rendered useless.
• Single blood pressure with no baseline. Without a booking BP, a late reading cannot be interpreted for pre-eclampsia versus chronic hypertension — dating the baseline is part of dating the pregnancy.
• Wrong dates poisoning everything downstream. An inaccurate gestational age mislabels growth, prematurity and post-maturity; reassuring a growth-restricted fetus as "small for optimistic menstrual dates" is a classic, fatal error.
• Skipping the syphilis or HIV retest. Universal booking testing is necessary but not sufficient — seroconversion and newly acquired syphilis during pregnancy are caught only by the repeat test (HIV later in pregnancy; syphilis at 32–34 weeks). A negative booking result is not a clearance for the rest of pregnancy.
• Asserting universal GBS screening as "correct". Risk-based prophylaxis (RCOG GTG 36; SA practice) is a defensible, evidence-grounded choice, not an oversight; presenting universal culture-based screening as the only right answer ignores a genuine, unresolved trade-off.
• Over-claiming the GDM thresholds. Presenting the one-step IADPSG criteria as settled, without acknowledging that they double prevalence and that benefit on hard outcomes is debated, reads as uncritical — especially in a resource-limited service.
• Forgetting postpartum rubella vaccination. A susceptible woman cannot be vaccinated antenatally (live vaccine); the action — postpartum vaccination — is easy to drop because it does not help this pregnancy.
• Escalating too slowly. A risk flag at the clinic is worthless if referral to definitive care lags the deterioration; in a geographically dispersed SA service, the timing of escalation is itself a clinical decision.

Evidence anchors

• Dowswell et al. — Alternative versus standard packages of antenatal care for low-risk pregnancy, Cochrane Database Syst Rev 2010
• HAPO Study Cooperative Research Group — Hyperglycemia and Adverse Pregnancy Outcomes, N Engl J Med 2008
• Crowther et al. (ACHOIS) — treating mild GDM cut serious perinatal complications 4%→1% (RR 0.33), N Engl J Med 2005
• Landon et al. (MFMU) — treating mild GDM reduced macrosomia, shoulder dystocia, caesarean and hypertensive disease, N Engl J Med 2009
• Rolnik et al. (ASPRE) — Aspirin versus Placebo in Pregnancies at High Risk for Preterm Preeclampsia, N Engl J Med 2017
• WHO recommendations on antenatal care for a positive pregnancy experience (2016) — the 8-contact model, recommendation E.7
• NICE NG201 — Antenatal care (2021)
• Hlongwane et al. — Implementing antenatal care recommendations (BANC-Plus), South Africa, Bull World Health Organ 2021
• RCOG Green-top Guideline No. 36 — Prevention of Early-onset Neonatal Group B Streptococcal Disease (2017)
• South Africa NDoH BANC-Plus / Guidelines for Maternity Care in South Africa — eight-contact model adopted from April 2017; universal antenatal HIV (PVT) and syphilis screening with repeat testing; respectful care and intimate-partner-violence enquiry incorporated.
• SA NDoH National Consolidated Guidelines on HIV (PVT) — universal opt-out antenatal HIV testing with repeat testing in pregnancy.