Defend the management of breech presentation, including external cephalic version and complications

In one line

Term breech is a decision problem, not a delivery technique: offer external cephalic version (ECV) first to convert the problem away, then counsel honestly that planned caesarean lowers the small absolute neonatal risk while planned vaginal breech — safe only with strict selection and an experienced accoucheur — spares the mother a caesarean and the next pregnancy a scar.

This chapter assumes the mechanics and manoeuvres covered in breech basics; it spends its words on defending a plan from the trial data — the appraisal, the thresholds, and the judgement calls a consultant owns, rather than a recital of how to perform Løvset's.

Why this matters in South Africa

The breech consultation in a South African public hospital is rarely a clean elective choice between two equivalent routes — it is a systems decision constrained by skill, theatre access and referral geography. Three SA-specific realities shape the plan. First, the skills-loss spiral: as caesarean became the default after the Term Breech Trial, registrars stopped accruing supervised vaginal-breech numbers, so fewer clinicians can safely offer or supervise it — which sections even good candidates, which loses more skill. Second, caesarean is not free in this setting: it carries real anaesthetic and sepsis morbidity in a high-HIV, often under-resourced theatre environment, and — because grand-multiparity and limited access to safe future operative delivery are common — every first scar genuinely constrains the next pregnancy's options (uterine-rupture, multiple-pregnancy for the compounding cases). Third, the woman often meets the system late and undiagnosed: the breech is first found in advanced labour at a district hospital with no on-site experienced accoucheur and no immediate theatre. The defensible SA stance flows from this: maximise ECV (the one high-value, low-risk intervention that reduces caesarean without the neonatal trade-off), refer breech antenatally from district to regional level before labour rather than improvising in second stage, and concentrate planned vaginal breech in the few units that maintain the skill.

Aetiology / pathophysiology — the advanced layer

The Intermediate groundwork (the mechanics of breech, the four types) is assumed. What matters at this level is why the subtype matters mechanically and what each one predicts, because the type is not a label — it changes both the probability of a safe vaginal birth and the specific catastrophe you are guarding against.

• Persistent breech as a marker, not a coincidence. Most fetuses are breech in mid-pregnancy and turn spontaneously; persistence to term means something has prevented version. The clinically useful corollary is that a term breech is a screening signal: actively exclude the conditions that hold a fetus breech because several of them independently contraindicate vaginal birth or even ECV — fetal anomaly (hydrocephalus, anencephaly, neuromuscular conditions that abolish the fetal tone needed to flex and turn), uterine factors (bicornuate/septate uterus, a low or fundo-cornual fibroid, a short or contracted cavity), placental factors (praevia/low-lying placenta restricting the lower pole), and liquor extremes (oligohydramnios fixes the fetus; polyhydramnios allows it to revert after version). A term breech without a deliberate ultrasound look for these has been under-assessed.

• Frank (extended-leg) breech — the safest vaginal candidate. Flexed hips with extended knees splint the legs against the trunk, so the breech is bulky and well-applied to the cervix. Mechanistically this gives the best fit and the lowest cord-prolapse risk because there is no irregular presenting part for the cord to slip past — the reason frank breech is the type that may be selected for vaginal birth. It is also the type most likely to revert after ECV (the splinted attitude resists the turning forces).

• Complete (flexed) breech. Flexed hips and knees, so feet sit alongside the buttocks. The presenting part is more irregular and less perfectly applied, raising cord-prolapse risk above frank breech — the clinical consequence is closer intrapartum surveillance and a lower threshold to abandon.

• Footling / kneeling (incomplete) breech — the dangerous mechanism. One or both feet (or knees) present below a non-engaged breech. The presenting part is small and irregular, the cervix is not snugly filled, and the feet can deliver through an incompletely dilated cervix — so this is the type with the highest cord-prolapse risk and the highest risk of head entrapment behind an undilated cervix. This mechanism, not dogma, is why footling presentation contraindicates planned vaginal birth: the body delivers through a cervix that has not opened enough for the aftercoming head.

• The two attitudes that kill regardless of type. A hyperextended ("stargazing") fetal neck converts the aftercoming head from a flexed, smallest-diameter passage into an extended one that can hinge on the symphysis and cause cervical-spinal-cord injury — an absolute contraindication to vaginal birth and a finding you must specifically seek on the version-day or pre-labour scan. A nuchal arm (an arm trailing up beside or behind the head) enlarges the diameter that must pass and is a leading cause of birth trauma at vaginal breech — its mechanism (traction-induced extension of the arm during a hurried or mistimed delivery) is exactly why the "hands-off until the scapulae" discipline exists.

The mechanism→consequence links that drive the route decision: footling → cord prolapse + head entrapment → abdominal delivery; hyperextended neck → cord injury → caesarean; traction before the scapulae → nuchal arm / extended head → trauma; large or growth-restricted fetus → disproportion or fragility → caesarean.

Assessment

The diagnosis is the easy part; stratifying the candidate for vaginal birth is the consultant task. The basic clinical picture is assumed from Intermediate; the selection judgement builds on it.

• Confirm presentation and type by ultrasound — clinical examination is unreliable. A hard ballottable mass at the fundus and FHR above the umbilicus are suggestive at best; ultrasound is mandatory before any plan. The scan is not just "is it breech" — it must define type, fetal neck attitude, estimated fetal weight (EFW), placental site, liquor volume and any anomaly, because each shifts the route decision.

• Severity-stratify the vaginal-breech candidate. The defensible "safe to offer vaginal birth" phenotype is narrow: frank or complete breech, EFW roughly 2.0–3.8 kg with growth on the centile, a flexed or neutral neck, no fetal compromise, normal liquor, and a skilled accoucheur available (RCOG GTG 20b). Each of these failing pushes toward caesarean. The subtle calls are at the edges — an EFW of 3.7 kg in a multipara with a proven pelvis is different from 3.7 kg in a primigravida; a marginally small-for-gestational-age fetus is more fragile and tolerates the mechanical stress of breech birth less well, so growth restriction shifts you toward caesarean even within the "acceptable weight" band.

• Pelvimetry does not predict outcome and is not recommended. Routine clinical or radiological pelvimetry does not improve selection; the labour itself, with normal progress, is the functional test of the pelvis. Quoting pelvimetry as a selection tool is a dated answer.

• Screen ECV eligibility deliberately. There is no absolute contraindication to ECV in: recent antepartum haemorrhage, an abnormal CTG, major uterine anomaly, ruptured membranes, any independent indication for caesarean, or multiple pregnancy (except for version of a second twin). Check rhesus status — ECV can provoke feto-maternal haemorrhage, so Rh-negative women need anti-D (and a Kleihauer to size the bleed). Relative factors that lower success — but do not forbid the attempt — are nulliparity, an anterior placenta, low liquor, an engaged breech and a tense abdominal wall.

• Context — the SA system question. At a district hospital, the operative question is not "which route is statistically better" but "do ECV and experienced vaginal-breech support exist here?" If neither does, the safe assessment conclusion is antenatal referral to a regional unit for ECV, with planned caesarean as the fallback — not waiting to meet the breech in second stage.

Investigations at depth

The investigation in breech is ultrasound, read like a staging scan, plus the rhesus work-up. Where each measurement misleads matters most.

• EFW — know its error bars. Third-trimester sonographic EFW carries a wide margin (commonly quoted around ±10–15%), so a single estimate near a threshold is a soft number, not a hard cut-off. A 3.9 kg estimate could be a 3.4 kg or a 4.4 kg baby; use the trend and the clinical picture, not one printout, and counsel accordingly rather than mechanically sectioning on a borderline figure.
• Neck attitude — the measurement most often skipped. Deliberately image the fetal neck for hyperextension; it is the single finding most often omitted and the one that converts an apparently fine vaginal candidate into an absolute caesarean.
• Liquor and Doppler. Oligohydramnios suggests a fixed, possibly growth-restricted or anomalous fetus; combine with umbilical-artery Doppler and growth if FGR is suspected, because a compromised placenta tolerates labour poorly and shifts the route.
• Anomaly scan if not already done. A persistent breech with no prior anatomy survey warrants one before committing to a vaginal plan — undiagnosed hydrocephalus is the classic catastrophe.
• Kleihauer–Betke after ECV in Rh-negative women to size the feto-maternal haemorrhage and confirm the standard anti-D dose is adequate; a large bleed needs a top-up.

Differential diagnosis and mimics that change management

"Breech" on palpation is sometimes wrong, and the alternative changes the whole plan:

• Transverse or oblique lie / shoulder presentation — there is no presenting pole; ECV (or in labour, a decision about internal podalic version of a second twin only) and the contraindication to vaginal birth are different. A persistent transverse lie at term is a caesarean and mandates exclusion of placenta praevia and a pelvic mass.
• Face or brow presentation mistaken for a soft breech on a difficult vaginal examination — the management diverges entirely.
• The undiagnosed footling masquerading as "advanced cephalic labour" when a foot is felt and misread as a hand or cord — the do-not-miss error, because it changes an apparent normal labour into a cord-prolapse-and-entrapment emergency.
• A pelvic mass (fibroid, ovarian) holding the fetus breech or obstructing descent — found on the same scan that confirms the breech, and itself an indication for caesarean.

Management

Structure the plan immediate (this antenatal contact) → ongoing (labour) → long-term (next pregnancy) — and subtype-specifically, because the route and the disqualifiers differ by type. This builds on the manoeuvre basics in breech basics; the additions here are the subtype-specific conduct, the named regimens and the rescue for each complication.

Immediate — try to make the problem disappear (ECV)

Offer ECV to every eligible woman with a term breech; it is the highest-value, lowest-risk move and the only one that lowers caesarean without the neonatal trade-off. Per RCOG GTG 20a, success is ~50%, and the predictors are mechanical: higher in multiparae (an independent association — adjusted OR around 2 for multiparity vs nulliparity in pooled series) and lower with nulliparity, anterior placenta, low liquor, an engaged breech and a tense abdomen.

• Timing. Offer from 37+0 weeks (nulliparae may be offered from 36+0). Doing it earlier does not reduce caesarean — the Early ECV 2 Trial turned more fetuses cephalic (RR 0.84) but did not lower the caesarean rate and trended toward more preterm birth, so the term timing stands.
• Tocolysis is the evidence-based adjunct. Give terbutaline 250 µg subcutaneously (a parenteral beta-stimulant) before the attempt — the Cochrane synthesis shows beta-stimulants raise cephalic presentation in labour (RR 1.68) and lower caesarean (RR 0.77). This is the single intervention with the best evidence for raising success and should be routine, not optional.
• Regional anaesthesia — the escalation for a failed or painful attempt. Where ECV fails or is limited by maternal discomfort/guarding, neuraxial blockade significantly raises success: pooled RCT data show success rising from roughly the high-30s/45% to roughly 58–60%, with a relative risk for success of about 1.6 and a reduction in caesarean (RR ~0.65), at the cost of more maternal hypotension (RR ~8 for a transient, treatable drop). A defensible figure is that of the order of 6–8 women need neuraxial anaesthesia for one extra successful version. The SA caveat is resource: a regional block for ECV consumes anaesthetic and theatre-adjacent capacity, so it is a regional-unit adjunct for selected failed attempts, not a district-level routine.
• Procedure and safety wrap. Perform where caesarean can be delivered promptly; pre-procedure ultrasound and CTG, re-check CTG afterwards, give anti-D to Rh-negative women with a Kleihauer. Counsel that reversion after success is uncommon, and that the risk of a complication forcing emergency caesarean is low (of the order of 0.5%).

Ongoing — if ECV fails or is declined, choose and conduct the route

Subtype-specific conduct of planned vaginal breech. Select frank (best) or complete type only. Allow spontaneous descent to the perineum before active pushing; practise "hands-off the breech" until the scapulae are visible — premature traction is what causes nuchal arms and an extended head. Deliver the arms by Løvset's manoeuvre and the after-coming head by Mauriceau–Smellie–Veit, or apply forceps to the after-coming head for controlled flexion. Maternal position (dorsal/lithotomy versus an upright/all-fours "physiological" breech approach) is an accoucheur-skill-and-unit-protocol decision; what is non-negotiable is continuous fetal monitoring and an immediate route to theatre.

Managing the complications when they occur — each has a specific rescue, not just a manoeuvre:

• Nuchal arm (~0–5% of vaginal breeches, higher in breech extraction, with neonatal trauma in around a quarter of cases): rotate the fetal trunk toward the side of the offending hand so the face turns to the maternal symphysis, reducing the tension splinting the arm against the occiput, then sweep it down. Prevention is the hands-off-until-scapulae discipline.
• Extended / deflexed after-coming head: maintain flexion with Mauriceau–Smellie–Veit (suprapubic pressure from an assistant to flex, jaw–shoulder traction), or forceps to the after-coming head; do not apply axial traction to an extended head.
• Head entrapment behind an incompletely dilated cervix (highest in footling and preterm breech, where the body is smaller than the head): this is the feared catastrophe. Standard rescue is Dührssen's cervical incisions at the cervix to release the head when full dilatation has been wrongly assumed — a manoeuvre that exists precisely because footling and preterm breech bodies deliver through an unready cervix. The deeper lesson is prevention: never let a footling or a preterm breech body deliver vaginally through a cervix you have not confirmed is fully dilated.
• Cord prolapse: breech carries roughly a ~1% cord-prolapse risk against a background of ~0.1–0.6%, and the risk is higher in complete and footling than in frank breech because of the irregular presenting part. With the fetus high and not imminently deliverable, this is an emergency caesarean with the standard temporising measures (relieve cord compression, elevate the presenting part, fill the bladder, knee-chest/Trendelenburg) en route.

The cardinal abandonment rule. A poorly progressing breech is delivered abdominally, not forced. Abandon to caesarean for delay in descent in the active second stage, any difficulty, a non-reassuring CTG, or cord prolapse with a high fetus. Traction on the breech before the scapulae are visible is never a rescue — it causes the disaster.

The undiagnosed breech in labour — the common real-world SA scenario — is decided on the spot by type, progress and who is in the building. An advanced, multiparous frank breech with an experienced accoucheur and a normally progressing labour may deliver vaginally; an early-labour footling, a primigravida, a non-reassuring trace, or no skilled help means caesarean.

Preterm breech is a different problem. Routine planned vaginal birth is not recommended: the preterm body is smaller relative to the head, so head entrapment behind an incompletely dilated cervix is the dominant risk. Below roughly 32 weeks the balance favours caesarean, though it remains individualised at the margins of viability and against the mother's wishes and local neonatal outcomes.

Long-term

Frame the caesarean within the woman's reproductive plans, because a first caesarean for breech commits future pregnancies to the VBAC-versus-repeat decisions and scar-rupture risk — this is part of informed consent, not an afterthought, and weighs especially heavily in SA where access to safe operative delivery in a future labour is not guaranteed. Counsel on the recurrence of breech in a future pregnancy and the renewed offer of ECV next time. Document the route discussion: both RCOG and ACOG require recorded counselling that planned vaginal breech carries a higher short-term neonatal risk than planned caesarean.

The evidence & the controversy

The defensible position rests on holding two findings in tension. The Term Breech Trial (2000) randomised 2083 women and found planned caesarean cut the composite of perinatal/neonatal death or serious morbidity from 5.0% to 1.6% (RR 0.33) — a result that, almost overnight, made planned caesarean the global default. The arithmetic matters: an absolute risk reduction of 5.0% − 1.6% = 3.4%, giving a number-needed-to-treat of 1 / 0.034 ≈ 29 caesareans to avoid one composite adverse neonatal outcome — a real but modest effect that is easy to overstate by quoting only the relative risk. The trial was promptly criticised: heterogeneous sites (including settings without continuous monitoring or skilled breech accoucheurs), some protocol violations, and inclusion of intrapartum events not plausibly attributable to route. Critically, the 2-year follow-up found no difference in death or neurodevelopmental delay between groups (3.1% vs 2.8%; RR 1.09) — the headline neonatal benefit did not translate into a durable handicap difference.

Into that gap came PREMODA (2006): 8105 women, strict protocols, experienced clinicians, and no significant difference in serious neonatal outcome between planned vaginal and planned caesarean breech (1.60% vs 1.59%; OR 1.10). The synthesis: with rigorous selection and skilled accoucheurs, planned vaginal breech is reasonable; the absolute excess neonatal risk is small but real and rises sharply when those conditions are not met. A large 2022 meta-analysis of observational data (94 285 births) still found planned vaginal breech carried higher perinatal mortality (RR 5.48) and birth trauma (RR 4.12), while planned caesarean carried higher severe maternal morbidity (the comparison favouring vaginal birth on maternal morbidity, RR 0.30) — quantifying the trade-off you are asking the woman to weigh.

On ECV, the evidence is unambiguous and points the other way — it is the intervention to maximise. The Cochrane review shows ECV at term lowers non-cephalic presentation at birth (RR 0.42) and caesarean (RR 0.57); the interventions review shows beta-stimulant tocolysis raises success (RR 1.68 for cephalic in labour) and lowers caesarean (RR 0.77); and neuraxial blockade adds a further, dose-of-anaesthesia-limited gain in success (RR ~1.6) and a lower caesarean rate (RR ~0.65). The reason ECV is the strategic centre of the SA answer is precisely that it captures benefit without the neonatal trade-off the route debate is stuck on.

The genuinely South African controversy is the skills-loss spiral described above: caesarean-as-default erodes vaginal-breech competence, which sections even good candidates, in a system where caesarean itself carries real morbidity and constrains future care. The defensible SA stance: maximise ECV, concentrate planned vaginal breech in units that maintain the skill, and refer antenatally from district level rather than meeting the breech first in second stage.

Landmark trials & key evidence

Screening — who, when, how

There is no laboratory screen for breech; the "screen" is presentation detection and risk-factor exclusion.

• Detection. Clinical palpation at term has limited sensitivity, so a persistent suspicion of malpresentation, an unstable lie, or an unengaged "head" at term should prompt confirmatory ultrasound — undiagnosed term breech entering labour is a preventable system failure.
• When. The actionable window opens at 37 weeks, when ECV is offered; earlier detection allows planning but not earlier intervention (the Early ECV 2 Trial settled timing).
• Risk-factor screen. Because persistent breech flags an underlying cause, the term breech scan doubles as a screen for fetal anomaly, uterine anomaly, placenta praevia/low-lying placenta, and liquor extremes — each of which independently changes the route or contraindicates ECV.

Pitfalls / exam traps / red flags

• Quoting the Term Breech Trial as the last word. Reciting RR 0.33 (or even ARR 3.4%/NNT ≈ 29) without the flat 2-year follow-up and PREMODA misses the controversy — the pairing is the answer.
• Offering ECV without the prerequisites. No ultrasound, no tocolysis, no anti-D for Rh-negative women, or doing it where caesarean is not promptly available — each is a fail. Offering it before term to "buy time" contradicts the Early ECV 2 Trial.
• Treating footling like any other breech. Footling/kneeling is the highest-risk subtype for cord prolapse and head entrapment and is not a planned-vaginal candidate — mislabelling it "vaginal" is the cases that injure babies.
• Forcing a vaginal breech. Continuing to push through delayed descent, traction on the breech before the scapulae (→ nuchal arm, extended head), or a hyperextended-neck fetus labelled "vaginal" — these end in trauma and litigation.
• Letting a footling or preterm body deliver through an unready cervix. Head entrapment behind an incompletely dilated cervix is the catastrophe; confirm full dilatation before the body delivers, and know that Dührssen's incisions are the in-extremis rescue.
• The undiagnosed footling in second stage with no experienced clinician. Do not improvise a difficult vaginal breech — this is an emergency caesarean (or a call for senior help), not a teaching opportunity.
• Cord prolapse with a high breech — the classic do-not-miss emergency; footling/complete breech carry the highest risk.
• Sectioning a borderline EFW mechanically. A 3.9 kg estimate carries ±10–15% error — counsel on the trend and the whole picture, not one printout.
• Consent shortcuts. Both ACOG and RCOG require documented counselling that planned vaginal breech carries a higher short-term neonatal risk than planned caesarean. An undocumented decision is indefensible.

Worked viva — how to structure the answer

A stem like "a 31-year-old G3P2, 38 weeks, frank breech confirmed on scan, EFW 3.3 kg, presents in early labour at a regional hospital; she wants to avoid a caesarean." A high-scoring answer runs:

1. Frame it — "This is a term frank breech in a multipara with a favourable EFW and prior vaginal births; the decision is route of delivery, and ECV is no longer relevant in established labour."
2. Stratify the candidate — confirm type (frank, the best), neck attitude (exclude hyperextension), EFW on centile, liquor, no fetal compromise, and — critically — is an experienced accoucheur and immediate theatre available? If yes, she is a reasonable planned-vaginal candidate; if not, recommend caesarean.
3. Conduct it safely — continuous CTG, hands-off until the scapulae, Løvset's for the arms, Mauriceau–Smellie–Veit (or forceps) for the head, theatre on standby; name the abandonment triggers (delay in descent, non-reassuring trace, any difficulty → caesarean).
4. Justify from evidence — pair the Term Breech Trial (planned CS lowers the small absolute neonatal risk, ARR ~3.4%, NNT ≈ 29) against PREMODA and the flat 2-year follow-up (with selection and skill, vaginal breech is reasonable); note the maternal-morbidity cost of caesarean.
5. Anticipate complications — cord prolapse, nuchal arm, extended/entrapped head, and the rescue for each.
6. Close the loop — document the counselling, frame any caesarean within her future reproductive plans, and offer ECV in the next pregnancy if the fetus is breech again.

For an antenatal stem at 37 weeks, the answer instead leads with ECV (eligibility check, tocolysis, anti-D if Rh-negative, where to do it), then the route counselling if ECV fails or is declined.

Exam traps & red flags

• Confusing the antenatal and intrapartum stems. At 37 weeks the answer is ECV-first; in established labour ECV is off the table and the answer is route-selection-and-conduct. Leading with the wrong one signals you haven't anchored the gestation/timing.
• Forgetting anti-D and the Kleihauer in a Rh-negative woman having ECV.
• Offering vaginal breech without naming the prerequisite — an experienced accoucheur present and immediate theatre are non-negotiable; "I'd do a vaginal breech" without them is a fail.
• Naming the manoeuvres but not the rescues — knowing Løvset's and Mauriceau–Smellie–Veit is Intermediate; the consultant names the management of the nuchal arm, the entrapped head (Dührssen's), and the prolapsed cord.
• Ignoring the SA system constraint — the right academic route is worthless if the unit can't deliver it; the defensible plan accounts for skill, theatre and referral.

Evidence anchors

• Term Breech Trial — Hannah et al, Lancet 2000
• TBT 2-year child outcomes — Whyte et al, AJOG 2004
• TBT 2-year maternal outcomes — Hannah et al, AJOG 2004
• PREMODA — Goffinet et al, AJOG 2006
• Planned vaginal vs caesarean breech meta-analysis — JOGH 2022
• ECV for breech at term — Cochrane 2015
• Interventions to help ECV — Cochrane 2015
• Early ECV 2 Trial — Hutton et al, BJOG 2011
• Neuraxial anaesthesia and ECV success — GRADE meta-analysis, Syst Rev 2024
• RCOG GTG 20a — ECV, 2017
• RCOG GTG 20b — Management of Breech Presentation, 2017
• ACOG Committee Opinion 745 — Mode of Term Singleton Breech Delivery, 2018
• South African NDoH Integrated Maternal and Perinatal Care Guideline (2024); SA Family Practice district-hospital breech series — see source notes.