Manage the prolonged second stage of labour

In one line

A prolonged second stage is a time-bound diagnosis of an underlying problem (malposition, inefficient uterine action, or true cephalo-pelvic disproportion) — and the correct response is never "wait longer" or "rush to caesarean", but a structured re-assessment that decides between continued pushing, assisted vaginal birth, and second-stage caesarean on the basis of fetal station, position and condition.

This chapter assumes the normal second-stage physiology, the partogram, and the routine conduct of spontaneous vaginal birth managed daily from normal labour & the partogram and malpresentation & malposition; it spends its words on the consultant-level decision — which mechanism is failing, which operative route, and when the clock stops being your friend.

Assessment

The clock starts at full dilatation, but the decision turns on what is happening to the head, not the elapsed minutes alone. Re-confirm the diagnosis before acting:

• Confirm full dilatation and the second stage is genuinely active. Distinguish the passive phase (full dilatation, no urge/involuntary pushing — descent occurring) from the active phase (involuntary bearing-down or directed pushing). Premature pushing on an un-descended head wastes maternal effort and inflates the "pushing time".
• Abdominal palpation first, then vaginal. Quantify head descent in fifths palpable above the brim — the most reliable bedside measure and the one SA guidelines anchor referral on. Excessive caput and moulding make a vaginal exam read deceptively low; a head ≥2/5 palpable abdominally is not deliverable vaginally regardless of the station you feel.
• Position and attitude. Persistent occipito-posterior (OP) and occipito-transverse (OT) malpositions are the commonest correctable cause. Deflexion, asynclitism and a large caput are the clinical signs. Intrapartum transabdominal/transperineal ultrasound to confirm position (occiput, spine, head–perineum distance) is increasingly recommended where available and resolves the clinical disagreement that precedes a failed instrument.
• Powers. Are contractions adequate (strength, ≥40 s duration, frequency)? In a nullipara with an epidural and a soft uterus, inefficient action — not disproportion — is often the limiting factor.
• Fetal condition. Continuous CTG interpretation (or intermittent auscultation per setting); meconium; rising baseline/decelerations push you toward expediting rather than extending.
• Maternal condition. Exhaustion, dehydration, ketosis, a full bladder (catheterise), analgesia adequacy, and any co-morbidity (cardiac disease, pre-eclampsia with severe features) that mandates a shortened second stage.

The synthesis: is the head low, well-flexed, OA and descending with good powers (favours continued pushing or a straightforward outlet/low assisted birth) or high, malpositioned, arrested with caput (favours second-stage caesarean)? The normal second-stage physiology and conduct are assumed; the judgement sits at this decision point. Whatever the route, a long second stage and operative birth are themselves risk factors for primary postpartum haemorrhage — plan active third-stage management and have uterotonics ready.

Classify the failure — the three mechanisms and how each changes the plan

The phrase "prolonged second stage" is a symptom, not a diagnosis. Name which of the three classic failures (the "3 Ps" — passenger, passage, powers) is operating, because each one redirects management down a different path. They co-exist, but one usually dominates, and getting the dominant one wrong leads to augmenting an obstructed multipara or sectioning a correctable malposition.

• Inefficient powers (the correctable, benign one). Typically a nullipara with a dense epidural, a soft uterus and adequate pelvis, a well-flexed OA head that simply will not descend because the expulsive drive is weak. The signs: contractions <40 s or spaced, no caput/moulding worth noting, a head that is low and central but static. This is the one subtype where oxytocin augmentation buys a spontaneous birth and time is relatively safe. The mechanism explains why: weak powers do not damage the fetus the way obstruction does, so a stable CTG licenses patience here in a way it never does in obstruction.
• Malposition (the commonest, and the one most often mismanaged as "CPD"). Persistent OP and OT, deflexion, asynclitism. The presenting diameter is larger than it needs to be not because the pelvis is small but because the head is presenting wrong. Mechanism→consequence: the deflexed OP presents the occipito-frontal (~11.5 cm) instead of the sub-occipito-bregmatic (~9.5 cm) diameter, so the same pelvis that would deliver a flexed OA head obstructs a deflexed OP. The corollary is the whole reason manual rotation exists — correct the attitude/position and the disproportion evaporates. Reaching for a caesarean here treats a positional problem as an anatomical one.
• True cephalo-pelvic disproportion / obstruction (the dangerous one). The head genuinely will not pass this pelvis at this attitude — a contracted pelvis, a macrosomic or hydrocephalic fetus, or a fixed brow/face-mentoposterior. Signs that distinguish it from the other two: a head ≥2/5 palpable abdominally that does not descend with good contractions, gross moulding (overlapping sutures, "+++"), a large caput succedaneum, a Bandl's ring in a multipara, and haematuria/an oedematous anterior cervical lip. Mechanism→consequence: continued strong contractions against an obstruction thin the lower segment toward rupture in a parous uterus and produce fetal acidosis — so this subtype forbids both augmentation and a trial of instrument, and goes to theatre.

The discriminator that separates malposition from CPD at the bedside is the response to a correctly performed manual rotation or a short trial of pushing with good powers: a malpositioned head rotates and descends; an obstructed head does neither. That single test, done deliberately, is worth more than another hour of clock-watching.

Severity and the atypical presentations that trip candidates

• The "deceptive descent" of caput and moulding. Severe caput can make the leading bony point feel at the spines (+0) when the biparietal diameter is still above the ischial spines — the classic set-up for a failed mid-cavity instrument. Always reconcile the vaginal station against the fifths palpable abdominally; if they disagree, believe the abdomen.
• The epidural-masked second stage. A dense block abolishes the bearing-down reflex, so "no urge to push" does not mean no descent. Allowing passive descent (within limits) and confirming station before declaring arrest avoids labelling a normally progressing labour as prolonged.
• The previously-sectioned uterus in second stage. A woman with a prior caesarean reaching a prolonged second stage needs the obstruction question answered fast — scar dehiscence/rupture presents as the same fetal-heart and pain picture and is the reason a multiparous arrest is never "give more syntocinon".
• The compound presentation / occult cord. A hand alongside the head, or a cord felt at the head, reframes the whole plan and may itself be the cause of arrest.

Management

Structure management immediate → ongoing → definitive.

Immediate (the re-assessment bundle): empty the bladder, optimise position, ensure hydration and effective analgesia, rupture intact membranes, and exclude obstruction (CPD, malposition) and fetal compromise. Get a senior present before committing to an instrument or theatre.

Ongoing (when extension is safe): if the fetal and maternal condition are reassuring and there is documented progress in descent/rotation, a defined further period of pushing is legitimate. Push when fully dilated — the OPTIMAL RCT (Cahill 2018) showed routine "labouring down" does not raise spontaneous vaginal birth (85.9% vs 86.5%) and adds postpartum haemorrhage and chorioamnionitis, so reserve delay for the high, well-tolerated, malpositioned head you are actively rotating, not as a default. For a persistent OP/OT, prophylactic manual rotation in the early second stage is a defensible first manoeuvre: PROPOP (2021) cut operative delivery from 41.2% to 29.4% (P = .047) without excess harm — correct the malposition before reaching for an instrument. Oxytocin augmentation is appropriate for inefficient uterine action in a nullipara (SA NDoH guidance: "consider oxytocin infusion for nulliparous women only"); it is not a substitute for delivery in a multipara, where suspect obstruction and uterine rupture risk dominate. NICE NG235 advises against routine oxytocin in the second stage for women with regional analgesia.

Definitive (assisted vaginal birth vs caesarean): decide on fifths-palpable and station.

SA-specific time triggers (NDoH Guidelines for Maternity Care in South Africa, 4th ed): the second stage is prolonged if the head has not descended to the pelvic floor after 2 hours of full dilatation, or delivery has not occurred after 45 min of pushing (nullipara) / 30 min (multipara). These are tighter than international thresholds and are calibrated for a referral system where a delayed transfer costs the baby.

Assisted vaginal birth technique points (SAMF/NDoH + RCOG): confirm prerequisites (full dilatation, ruptured membranes, defined position, adequate analgesia, empty bladder, contractions). Apply traction only with contractions. Abandon ventouse after no descent on traction, 3 pulls with functioning equipment, or 2 cup detachments; target negative pressure −0.6 to −0.8 Bar (never exceed 0.8 Bar/80 kPa/600 mmHg). Forceps only by experienced operators in hospital. Sequential use of instruments (vacuum then forceps) increases neonatal trauma and is a deliberate, senior, documented decision — not a reflex.

Subtype-specific management — match the manoeuvre to the mechanism

The route is chosen by which mechanism is diagnosed, not by a flowchart of minutes.

For malposition — rotate before you pull. When a persistent OP/OT is the dominant problem and the head is at or below the spines with a fully dilated cervix, restoring an OA position converts a difficult delivery into an easy one. Three named techniques, and how they differ:

• Digital/manual rotation — the operator's fingers (or whole hand) flex and rotate the head to OA between contractions, the woman pushes to hold it, then delivery proceeds spontaneously or with a non-rotational (direct-traction) instrument. Lowest equipment and morbidity threshold; the PROPOP and the AJOG meta-analysis (Phipps/Verhaeghe-type pooled data) show it shortens the second stage and raises spontaneous vaginal birth (pooled ~64.9% vs 59.5%) without excess harm. First line in a district/regional setting because any competent obstetrician can do it with no special instrument.
• Rotational ventouse — the cup is placed over the flexion point and rotation occurs as traction is applied (the head auto-rotates as it descends). Avoids the blind application of a rotational forceps; failure simply means abandoning, not trauma from a misapplied blade. Preferred rotational instrument where forceps expertise is scarce — which describes most SA units.
• Kielland's rotational forceps — the sliding lock and minimal pelvic curve allow rotation in the pelvis. Highest skill ceiling and the highest training/medicolegal threshold; cohort data show comparable composite morbidity to rotational ventouse and manual-rotation-plus-direct-forceps, and a higher vaginal-birth success rate in expert hands, but there are no RCTs and the complication signal (when misapplied) is real. A consultant-only manoeuvre, and in many SA units a skill that has been allowed to atrophy — naming it is fine, but the honest answer is "rotational ventouse or manual rotation, or theatre, unless a Kielland-competent consultant is present."

For inefficient powers — augment, with the safety rails. A nulliparous OA head, low and central, with weak contractions and a reassuring CTG: oxytocin titrated to adequate contractions, continuous fetal monitoring, a defined re-assessment window, and a clear ceiling. The judgement call is the parity gate: augmentation in a multipara's arrested second stage is a documented near-miss generator because it treats presumed obstruction as if it were weak powers.

For obstruction/CPD — second-stage caesarean, done well. No augmentation, no trial of instrument. The technical challenge is the deeply impacted head: anticipate it, decant before the contraction, consider a second-stage caesarean technique (reverse breech extraction / "push" by an assistant from below vs "pull" from above) and a fetal pillow if available, and brief the team on the higher risk of extension tears, haemorrhage and bladder injury. That decision and its mechanics are their own skill set — see caesarean-at-full-dilatation.

Procedural detail

• Perineal protection at operative birth. A mediolateral episiotomy cut at the correct angle is protective against obstetric anal sphincter injury (OASIS) at assisted vaginal birth — the protection is angle-dependent. Cutting at ~60° from the midline at crowning is needed because perineal distension flattens the angle; the post-delivery sutured angle should sit in the ~40–60° "safe zone", and the risk of OASIS falls steeply as the incision moves away from the midline. Many operators systematically under-cut the angle with straight scissors, which is the mechanistic reason a "mediolateral" episiotomy so often fails to protect. (Routine episiotomy for spontaneous birth is not recommended — this is a targeted manoeuvre for the instrumental/at-risk perineum.)
• The intrapartum-ultrasound numbers, and where they mislead. Transperineal angle of progression (AoP) and head–perineum distance (HPD) quantify station objectively. As guides: an AoP above roughly 120° in a prolonged-second-stage nullipara predicts a high probability of successful vacuum birth, and an HPD around ≤35–40 mm favours success while larger distances predict difficulty/failure. Use them to support the clinical call and to mediate operator disagreement — not to replace the abdominal fifths-palpable assessment, because a reassuring AoP cannot overrule a head you can feel ≥2/5 above the brim, and these cut-offs carry only modest specificity (they rule success in better than they rule it out).
• The trial of instrumental delivery in theatre. When success is uncertain (mid-cavity, malposition, borderline station), the safe structure is a trial in theatre — attempt the assisted birth set up and prepped for immediate caesarean, with the threshold to abandon defined before you start. This converts the dangerous corridor failure into a controlled move to caesarean and is the SA-appropriate answer when an instrument might fail.

When delivery is by caesarean at full dilatation, anticipate a deeply impacted head and the morbidity that follows; that decision is its own skill set — see caesarean-at-full-dilatation.

The evidence & the controversy

The central tension is how long is too long. Definitions diverge: ACOG (2024) sets prolonged second stage at >3 h pushing (nulliparous) and >2 h (multiparous), explicitly allowing individualised extension while descent is documented and status reassuring, and advising assessment for assisted vaginal birth before defaulting to caesarean. NICE NG235 (2023) expects birth within 3 h (nulliparous) and 2 h (parous) of the active second stage without an epidural, then obstetric review. SA NDoH (2016) is far tighter (45/30 min pushing). The reason the thresholds differ: they reflect the safety of the surrounding system, not a biological constant.

What does extending actually buy? Gimovsky & Berghella's RCT (AJOG 2016) randomised 78 nulliparous women meeting ACOG criteria to ≥1 extra hour vs expedited delivery; caesarean fell from 43.2% to 19.5% (RR 0.45, 95% CI 0.22–0.93) with no significant difference in maternal/neonatal morbidity — but it was small and underpowered for harm. Against that, Infante-Torres et al.'s meta-analysis (Int J Environ Res Public Health 2020; 12 studies, 266 479 women) found a prolonged second stage associated with higher 5-min Apgar <7 (nulliparous OR 1.65, 95% CI 1.20–2.27), NICU admission (OR 1.63, 1.44–1.84) and neonatal sepsis (OR 1.57, 1.07–2.29), with steeper effects in multiparae. The honest reconciliation: extending lowers caesarean rates but trades into operative-vaginal-birth, third/fourth-degree tear, shoulder dystocia and neonatal-acidosis risk — so extension is conditional on continuous reassurance and is most defensible in nulliparae with epidurals, least defensible in multiparae where arrest signals obstruction.

The newest strand is the assisted-vaginal-birth deficit. FIGO's 2025 good-practice recommendations (Ubom AE et al., Int J Gynaecol Obstet 2025) flag falling AVB rates — <1% in some African countries — as a driver of the caesarean epidemic, and argue that every skilled birth attendant should be competent in vacuum birth, with a safety checklist, sequential-instrument rules and clear "when to abandon" criteria. For SA, where a high primary-caesarean rate compounds the risk of the next pregnancy's uterine rupture and placenta accreta spectrum, reviving competent assisted vaginal birth is a public-health argument, not a stylistic one.

The sequential-instrument and instrument-vs-caesarean trade-off, quantified

The decision "vacuum, forceps, or theatre" is best argued from the neonatal-injury data rather than operator preference. The landmark dataset is Towner et al. (NEJM 1999), a cohort of ~583 000 nulliparous singleton births: rates of intracranial haemorrhage rose with instrumentation, and the worst outcome was the sequential use of vacuum then forceps, carrying roughly a 1-in-256 risk of neonatal intracranial injury — higher than either instrument alone and comparable to (not safer than) caesarean in labour. The mechanism is additive shear on a head already stressed by the first instrument and by the obstruction that made it fail. The consultant corollaries:

• A failed vacuum is information, not a cue to grab forceps. Each instrument failure raises the prior probability of obstruction; switching instruments without re-examining and without a senior decision multiplies the very injury you are trying to avoid. If the first appropriate instrument fails on a correctly selected case, the default next step is caesarean, not the second instrument.
• Caesarean in the second stage is not the risk-free alternative. Towner showed instrument-in-labour and caesarean-in-labour share a similar intracranial-injury profile, because the common factor is the abnormal labour, not the instrument — which is exactly why "just section everyone" is a false safety and why competent AVB matters.
• Protracted vacuum is its own hazard. Prolonged cup application and excessive pulls drive subgaleal haemorrhage and intracranial injury, the basis for the hard "3 pulls / 2 detachments / no descent → abandon" rule. The rule is not bureaucratic; it is the inflection point on the injury curve.

The South African system layer

SA's public-sector caesarean rate has climbed (≈28–29% of births in recent national data) while assisted-vaginal-birth skills have thinned — the two trends are linked. Every primary caesarean done for a correctable second-stage problem mortgages the next pregnancy against uterine rupture, placenta praevia and placenta accreta spectrum, complications that are far deadlier in a district setting than a well-conducted assisted birth would have been. The Saving Mothers / Saving Babies framing is therefore not "avoid caesarean to save money" but "an unnecessary first caesarean is a future maternal-death risk transferred to a system with thinner resources." This is the public-health spine of the manual-rotation-and-vacuum revival argument, and the reason the NDoH thresholds are deliberately action-forcing rather than permissive.

Landmark trials & key evidence

The observational data show morbidity rising with duration (so a ceiling exists), while the randomised data show that a defined extension — and active correction of malposition — safely lowers the caesarean rate. Reconciling the two is the central judgement.

Worked viva — how to structure the answer

Examiners give a stem like "a 24-year-old nullipara, epidural in situ, fully dilated for 2 hours, pushing 45 min, CTG normal, head 1/5 palpable abdominally, direct OP at the spines with moderate caput." A high-scoring answer runs:

1. Frame it — "This is a prolonged second stage in a nullipara; the dominant mechanism here is malposition — a persistent direct OP at the spines — with a reassuring fetus, so I have time to correct the cause rather than default to caesarean."
2. Re-assess deliberately — empty the bladder, confirm full dilatation and ruptured membranes, reconcile the vaginal station against the 1/5 abdominal palpation, assess powers and analgesia, and (if available) confirm position on transperineal ultrasound. Call a senior.
3. Match manoeuvre to mechanism — "Because this is malposition with a reassuring CTG and a low head, I would attempt manual rotation to OA and let her push; if it rotates and descends I anticipate spontaneous or low assisted birth, and I would conduct any vacuum with a correctly angled mediolateral episiotomy to protect the sphincter."
4. Define the ceiling before acting — "If rotation fails or the head will not descend with good powers, that reclassifies the problem toward obstruction; I would then do a trial of instrumental delivery in theatre, prepped for immediate caesarean, with abandon criteria set in advance — and I would not switch from a failed vacuum to forceps."
5. Justify from evidence — PROPOP and the manual-rotation meta-analysis for rotating first; OPTIMAL for pushing when fully dilated; Towner for why sequential instruments and reflex caesarean are both wrong defaults; the NDoH thresholds for the SA clock.
6. Anticipate complications and close the loop — PPH (long second stage + operative birth → active third stage, uterotonics ready), OASIS (examine the sphincter, mediolateral angle), neonatal review after any instrument, and debrief plus documentation of the decision sequence.

State explicitly which mechanism is operating and what would change your mind — that conditional reasoning is the core of the consultant decision.

Exam traps & red flags

• Calling a passive phase a prolonged second stage. Pushing time and full-dilatation time are different clocks; conflating them invents a problem and triggers a needless caesarean.
• Trusting the vaginal station over the abdomen. Caput/moulding fake descent. A head ≥2/5 palpable abdominally must not have an instrument applied — that is the classic failed/traumatic AVB on the spot exam.
• Augmenting a multipara's arrested second stage. Suspect obstruction and uterine rupture; oxytocin here is dangerous. Reassess for CPD, do not "give it more syntocinon".
• Persisting with a failing instrument. Beyond 3 pulls / 2 detachments / no descent → stop. Switching vacuum-to-forceps without senior decision multiplies neonatal trauma (Towner: ≈1 in 256 intracranial injury).
• Treating malposition as CPD. A deflexed OP presents a larger diameter, not a smaller pelvis — rotate first; sectioning a correctable malposition mortgages the next pregnancy.
• Treating caesarean as the safe default. Caesarean in labour carries a neonatal-injury profile similar to instrument-in-labour — the abnormal labour is the risk, not the route. Competent AVB is not the riskier choice.
• Missing the obligate-shortened second stage. cardiac-disease-in-pregnancy and severe pre-eclampsia-early-onset-severe need an elective assisted second stage — do not let these labour on to an exhaustion-driven emergency.
• Forgetting the deflexed OP / brow. Failure to descend with a large caput and a deflexed head is malposition until proven otherwise; rotational delivery or caesarean, not more pushing. A fixed brow or mentoposterior face is true obstruction — caesarean.
• Under-angling the mediolateral episiotomy. A "mediolateral" cut too close to the midline does not protect the sphincter — cut ~60° at crowning.
• Bladder and analgesia overlooked. A full bladder and inadequate analgesia are reversible causes of "arrest" — fix them inside the immediate bundle.

Evidence anchors

• NICE NG235 — Intrapartum care, Recommendations (2023)
• ACOG Clinical Practice Guideline — First and Second Stage Labor Management (2024)
• Guidelines for Maternity Care in South Africa, 4th ed (NDoH, 2016) — prolonged second stage, vacuum extraction conditions, fifths-palpable referral thresholds (no stable per-page URL)
• Rouse DJ, et al. (NICHD MFMU Network). Second-stage labor duration in nulliparous women: relationship to maternal and perinatal outcomes. Am J Obstet Gynecol 2009;201(4):357.e1–7
• Gimovsky AC, Berghella V. RCT of prolonged second stage: extending the time limit vs usual guidelines. Am J Obstet Gynecol 2016;214(3):361.e1–6
• Cahill AG, et al. Effect of immediate vs delayed pushing on rates of spontaneous vaginal delivery among nulliparous women receiving neuraxial analgesia (OPTIMAL): a randomized clinical trial. JAMA 2018;320(14):1444–1454
• Infante-Torres N, et al. Prolonged second stage of labor and short-term neonatal morbidity: systematic review and meta-analysis. Int J Environ Res Public Health 2020;17(21):7762
• Blanc J, et al. Prophylactic manual rotation of occiput posterior and transverse positions to decrease operative delivery: the PROPOP randomized clinical trial. Am J Obstet Gynecol 2021
• Towner D, et al. Effect of mode of delivery in nulliparous women on neonatal intracranial injury. N Engl J Med 1999;341(23):1709–1714
• Phipps H, et al. (Bertholdt/Verhaeghe-type pooled analysis). Manual rotation of occiput posterior or transverse positions: a systematic review and meta-analysis of randomized controlled trials. Am J Obstet Gynecol 2021
• Ubom AE, et al. FIGO good practice recommendations: Assisted vaginal birth and the second stage of labor. Int J Gynaecol Obstet 2025;171(3):970–982
• RCOG Green-top Guideline No. 26 — Assisted Vaginal Birth (Murphy et al., 2020)