Elective Induction at 39 Weeks: What ACOG’s Update Actually Changed
A single trial moved the window of clinical attention from 41 weeks to 39 — and the implementation data have not followed as cleanly as the guidance reads.
One-Minute Takeaway
- Elective induction at 39 weeks rests largely on one trial. ARRIVE reduced cesarean birth from 22.2% to 18.6% (RR 0.84, 95% CI 0.76–0.93) among 6,106 low-risk nulliparous women — an NNT of 28.
- ARRIVE missed its primary endpoint. The composite of perinatal death or severe neonatal complication was 4.3% versus 5.4% (RR 0.80, 95% CI 0.64–1.00).
- ACOG’s Clinical Practice Update replaced Practice Bulletin 146 and moved the guidance window from late-term and postterm pregnancy to 39 0/7–41 6/7 weeks.
- Implementation data conflict. An analysis of 2,860,942 births found cesarean rates fell; a 13-hospital time series found no change (RR 0.96, 95% CI 0.89–1.04).
- Generalisability is the open question. Of 50,581 women screened and 22,533 found eligible, only 6,106 (27%) consented to randomisation.
Why Elective Induction Matters
Cesarean birth is the most common inpatient surgery in the United States, and the first one sets the pattern for every pregnancy after it. ACOG puts the national nulliparous, term, singleton, vertex rate near 25.6%, against a Healthy People 2030 objective of 23.6%.
The question is not abstract for anyone covering a prenatal clinic. It arrives at every 39-week visit with a low-risk first-time patient: offer elective induction, or wait for labour to start. Few decisions in obstetrics are made as often, on evidence this contested.
Before 2018 the reflex was to wait. The concern was that elective induction raised the risk of cesarean birth — a belief built largely on observational studies and small underpowered trials.
Those studies compared induction against spontaneous labour. That is not the choice in front of the patient. At 39 weeks she can accept elective induction or accept expectant management, and expectant management already includes everyone who later needs an indicated induction.
Framing the comparison correctly is what made a randomised trial worth running, and what makes its result difficult to dismiss. It is also why the guidance changed before the practice settled.
Two framings of elective induction have competed since. One treats elective induction as an intervention that must justify itself against nature. The other treats 39 weeks as simply another point at which the risk of continuing and the risk of delivering can be compared.
The second framing won the guideline argument. Whether it wins on a given labour ward depends on numbers that ward has to generate for itself.
What the Guideline Recommends
ACOG’s Clinical Practice Update is a focused revision rather than a full rewrite. It integrates the ARRIVE trial and subsequent related studies into existing guidance for pregnant individuals at 39 0/7–41 6/7 weeks of gestation without a medical indication for delivery. That window is where elective induction now lives.
Two documents were folded into it. The update revises Practice Bulletin 146, and it replaces the Practice Advisory that ACOG issued in August 2018, within days of ARRIVE appearing in print.
The counselling duty is stated plainly in ACOG’s companion committee statement on safe reduction of primary cesarean birth. Obstetric and midwifery clinicians should collaboratively discuss and counsel patients on the risks and benefits of elective induction compared with expectant management at term. That conversation should include the patient’s birth preferences.
Clinical Pearl
ACOG’s own summary of the trial evidence is more guarded than the headline. The committee statement describes a large multicentre trial showing similar or decreased cesarean birth rates alongside similar or higher rates of perinatal morbidity such as hemorrhage and infection. The offer is framed as reasonable, not as expected practice.
That statement also names the constraint residents feel first. Resource availability, staffing, and unit throughput during prolonged inductions should all be weighed before non-medically indicated inductions are implemented in any setting.
Evidence Reviewed
The trials below share one axis: a policy of induction against a policy of expectant management, measured on cesarean birth and perinatal outcomes. The two implementation studies test whether the trial result held once units acted on it.
| Study / Source | Design | N | Population | Comparison | Key result |
|---|---|---|---|---|---|
| Middleton 2020 (Cochrane) | Systematic review and meta-analysis, 34 RCTs | 21,030 women (cesarean); 18,795 infants (mortality) | At or beyond 37 weeks, mostly low risk | Induction policy vs expectant management | Perinatal death RR 0.31 (95% CI 0.15–0.64), high certainty; cesarean RR 0.90 (95% CI 0.85–0.95), moderate certainty; NNTB 544 (95% CI 441–1,042) |
| Hong 2023 | Systematic review and meta-analysis, 14 studies (1 RCT, 12 retrospective cohorts, 1 cross-sectional) | >1.6 million (86,555 in the induction arms) | Elective induction at 39 weeks, no medical indication | Elective induction at 39 weeks vs expectant management thereafter | Associated with reduced likelihood of perineal injury, macrosomia and low 5-minute Apgar score; among nulliparous women only, associated with increased likelihood of shoulder dystocia |
| ARRIVE (Grobman 2018) | Multicentre RCT, 41 US centres | 6,106 | Low-risk nulliparous, randomised at 38 0/7–38 6/7 weeks | Induction at 39 0/7–39 4/7 weeks vs expectant management | Cesarean 18.6% vs 22.2% (RR 0.84, 95% CI 0.76–0.93); composite perinatal 4.3% vs 5.4% (RR 0.80, 95% CI 0.64–1.00) — primary endpoint not met |
| SWEPIS (Wennerholm 2019) | Multicentre RCT, superiority, stopped early | 2,760 | Low-risk singleton at 41 weeks, 14 Swedish hospitals | Induction at 41 weeks vs expectant management and induction at 42 weeks | Composite perinatal 2.4% vs 2.2%, no difference; 0 vs 6 perinatal deaths — stopped early for excess mortality in the expectant arm |
| INDEX (Keulen 2019) | Multicentre RCT, non-inferiority | 1,801 | Low-risk singleton at 41 weeks, Netherlands | Induction at 41 weeks vs expectant management until 42 weeks | Composite adverse perinatal 1.7% vs 3.1%, absolute risk difference −1.4% (95% CI −2.9% to 0.0%); cesarean 10.8% in both arms; non-inferiority not shown |
| 35/39 (Walker 2016) | RCT, UK NHS | 619 | Primigravid women aged 35 years or older | Induction at 39 0/7–39 6/7 weeks vs expectant management | Cesarean 32% (98/304) vs 33% (103/314), RR 0.99 (95% CI 0.87–1.14) — no significant effect; no adverse short-term maternal or neonatal effects; underpowered for stillbirth |
| Wood 2023 | Cross-sectional interrupted time series | 2,860,942 births | Low-risk nulliparous at 39 weeks or beyond, US | Before vs after ARRIVE publication | 39-week induction rate 15.0% against 13.8% projected from pre-trial trend; induction rose and cesarean birth fell |
| Nethery 2023 | Interrupted time series, 13 US hospitals | 28,256 births | Term nulliparous | Before vs after ARRIVE publication | Elective induction 3.6% to 10.8% (RR 1.42, 95% CI 1.18–1.71); cesarean birth unchanged (RR 0.96, 95% CI 0.89–1.04) |
Eight sources, three of them observational or predominantly so. Every figure above has been confirmed against its primary source.
Key Findings on Elective Induction
Cesarean birth: the outcome that moved
ARRIVE was randomised, so causal language is warranted for its comparison. Induction in the 39th week resulted in fewer cesarean births than expectant management, and the trial’s authors put the number needed to treat at 28.
The pooled picture points the same way but smaller. Cochrane’s review found that a policy of induction at or beyond 37 weeks probably reduced cesarean birth, rated moderate certainty across 31 trials.
The only other randomised trial of 39-week induction in nulliparous patients did not reproduce the effect. The 35/39 trial randomised 619 primigravid women aged 35 or older, and cesarean rates were 32% with induction against 33% with expectant management (RR 0.99, 95% CI 0.87–1.14).
Perinatal outcomes: where the trial fell short
ARRIVE did not meet its primary endpoint. The composite of perinatal death or severe neonatal complication was not significantly lower with elective induction, and the upper bound of the confidence interval reached 1.00.
The distinction matters for consent. The cesarean reduction was the principal secondary outcome, while the neonatal benefit patients most often ask about was the outcome the trial was designed around and did not show.
The pooled data are stronger on mortality, though the absolute effect is small. Cochrane rated the reduction in all-cause perinatal death as high-certainty evidence, with absolute rates of roughly 0.4 against 3 deaths per 1,000.
The two 41-week trials sharpen the timing question rather than the 39-week one. SWEPIS was stopped early after six perinatal deaths occurred in the expectant arm and none in the induction arm, even though its composite primary outcome did not differ between groups.
INDEX could not show that expectant management to 42 weeks was non-inferior to induction at 41 weeks. Its composite adverse perinatal outcome favoured induction by 1.4 percentage points, and cesarean rates were identical in both arms.
Beyond cesarean: what else moves
Cesarean birth is not the only outcome elective induction touches. ARRIVE reported less hypertensive disease of pregnancy in the induction arm, and less need for neonatal respiratory support in the first days of life.
A meta-analysis aimed specifically at the 39-week question found a broader pattern across more than 1.6 million participants. Elective induction at 39 weeks was associated with a reduced likelihood of perineal injury, of macrosomia, and of a low 5-minute Apgar score.
Its authors did not present that as a clean win, and the exception lands squarely on this article’s population. Among nulliparous women only, elective induction was associated with an increased likelihood of shoulder dystocia.
That is the trade to name out loud in a first-time mother’s counselling. The authors’ own conclusion is that elective induction at 39 weeks may be safe and beneficial for some women, and that the risks should be discussed with nulliparous women specifically.
That pooled analysis rests mostly on retrospective cohorts, so its estimates are associations rather than effects. It is a useful counterweight to reading ARRIVE on its own, not a substitute for it.
Whether the effect survived contact with practice
Practice changed quickly after publication. Both of the analyses below are observational, so their findings are associations rather than demonstrated effects.
The larger examined 2,860,942 US births. It found that 39-week induction rose above the rate projected from pre-trial trends, and that cesarean birth fell.
A 13-hospital time series reached a different conclusion. Elective induction roughly tripled after ARRIVE, yet cesarean birth did not change, and neither did hypertensive disorders of pregnancy.
A third analysis found both things at once. In the year after ARRIVE there were more inductions, more deliveries at 39 weeks, and fewer cesarean births. Alongside that came a small but statistically significant increase in some adverse maternal and neonatal outcomes: maternal blood transfusion, maternal intensive care admission, and assisted ventilation for neonates.
Its authors did not claim elective induction caused those harms. They argued the signal should be examined against the concurrent changes in obstetric practice, which is a narrower and more careful claim.
Both cannot be the last word. That disagreement is arguably the most clinically useful finding in the implementation literature. It suggests the cesarean benefit may depend on how an elective induction is run, not only on when it is offered.
Read alongside ACOG’s own guarded phrasing, the implementation record is less triumphant than the trial. The guidance moved on one randomised result; the observational follow-up has not settled what that result is worth outside the trial’s walls.
Key Changes and Their Basis
The title carries the change. Practice Bulletin 146 was called Management of Late-Term and Postterm Pregnancies; the document that supersedes it is called Management of Full-Term Nulliparous Individuals Without a Medical Indication for Delivery.
That is a shift in where clinical attention begins. The old bulletin addressed pregnancies that had already gone past their due date. The update addresses a window opening at 39 0/7 weeks, before most patients would consider themselves overdue at all.
The stated basis is ARRIVE and the work that followed it. The update describes itself as integrating data from that trial and subsequent related studies into existing ACOG guidance.
The document class changed as well. The 2018 Practice Advisory was ACOG’s rapid-response format for guidance that could not wait; a Clinical Practice Update is the durable mechanism for revising a section of an existing document.
Clinical Pearl
One practical consequence for anyone studying from an older reading list: searching for Practice Bulletin 146 will surface a document that no longer carries current guidance on this question. The same applies to the August 2018 ARRIVE Practice Advisory, which the update explicitly replaces.
Quality and Consistency of the Evidence
- High — perinatal death with a policy of induction at or beyond termCochrane rated this high certainty across 22 trials and 18,795 infants. The direction is consistent; the absolute difference is small.
- Moderate — reduction in cesarean birthCochrane rated the pooled estimate moderate certainty. ARRIVE is the single largest contributor at 39 weeks, and the only other randomised trial in nulliparous patients was null.
- Low — reproducibility outside academic centresTwo observational analyses of the same natural experiment disagree on whether cesarean rates actually fell.
- Insufficient — childhood neurodevelopmentCochrane recorded that existing trials have not reported it.
The consistency problem is not the direction of effect but its size and its setting. ARRIVE’s expectant-management arm had a cesarean rate well below the overall US rate, so the comparator was already outperforming the national average.
Hospital-level variation compounds this. Across 99,441 induced nulliparous, term, singleton, vertex births in California, the median cesarean rate after induction was 32.2%, with a range of 18.5% to 84.6% between hospitals.
That median deserves a second look. It sits well above ARRIVE’s induction arm, and almost exactly where the 35/39 trial’s cesarean rates landed — and risk adjustment did not narrow the spread.
A trial effect of ARRIVE’s magnitude cannot be assumed to survive a range of that width. Where a unit sits within it is a local empirical question, not a guideline one.
What the Evidence Does Not Show
- Whether the result applies to the patients who declined. ARRIVE screened 50,581 women and found 22,533 eligible; 6,106 consented, or 27%. The characteristics and outcomes of those who declined were not reported.
- Whether it applies outside the trial’s settings. ARRIVE ran across 41 hospitals in a research network with 24-hour in-house obstetrician coverage, and applied a standardised definition of failed induction in the latent phase.
- Whether it applies to multiparous patients. ARRIVE enrolled nulliparous women only, and the updated guidance on elective induction is titled accordingly.
- What happens to the children. Cochrane records that trials in this area have not yet reported childhood neurodevelopment, and names it a priority.
- Whether the finding replicates in another health system. FRENCH-ARRIVE is randomising 4,200 low-risk nulliparous women on the ARRIVE model, with completion estimated for 2027.
None of these gaps argues against offering elective induction. They argue against presenting the trial’s numbers as though they were the patient’s numbers.
Practical Implications
The evidence supports an offer of elective induction, not a default. ACOG’s language is that clinicians should counsel on risks and benefits and incorporate the patient’s birth preferences — not that delivery at 39 weeks is the expected course.
Two findings belong in that conversation, and they point in different directions. The cesarean benefit is real but modest; the neonatal benefit patients tend to ask about was not shown.
Local data are more informative than the trial’s. With a real-world median of 32.2% after induction in nulliparous patients, most units are not starting where ARRIVE started, and the trial’s arithmetic does not transfer unexamined.
Capacity is a clinical variable here rather than an administrative one. ACOG names staffing and unit throughput during prolonged inductions among the factors to weigh before offering non-medically indicated induction.
The 41-week boundary has not loosened. SWEPIS was stopped early for excess perinatal deaths under expectant management beyond 41 weeks, and INDEX could not show non-inferiority for waiting until 42.
Counsel the harms alongside the benefit. The pooled data point to less perineal injury and less macrosomia with elective induction, and to more shoulder dystocia in nulliparous women — which is to say, in exactly the patients this guidance covers.
Finally, treat the offer of elective induction as a conversation with a timeline. ACOG frames the discussion as prenatal rather than intrapartum, which gives a patient the room to change her mind before she is on the ward.
Evidence Grade and Bottom Line
- Strong evidence supports a reduction in perinatal death with a policy of induction at or beyond term, though the absolute benefit is small and the NNTB is 544.
- Evidence suggests elective induction at 39 weeks reduces cesarean birth in low-risk nulliparous patients, resting on one large trial and a moderate-certainty pooled estimate.
- Insufficient evidence to determine whether that cesarean benefit reproduces outside academic centres, where two observational analyses reach opposite conclusions.
- The update changed the window of clinical attention, not the strength of the underlying data.
For a resident, the durable point is narrower than the headline. What changed is which weeks the guidance covers and which documents are current, and the trial behind that change answered a secondary question convincingly and its primary one not at all.
Article Information and References
- Middleton P, Shepherd E, Morris J, Crowther CA, Gomersall JC. Induction of labour at or beyond 37 weeks’ gestation. Cochrane Database Syst Rev. 2020;7(7):CD004945. DOI: https://doi.org/10.1002/14651858.CD004945.pub5
- Hong J, Atkinson J, Roddy Mitchell A, Tong S, Walker SP, Middleton A, Lindquist A, Hastie R. Comparison of maternal labor-related complications and neonatal outcomes following elective induction of labor at 39 weeks of gestation vs expectant management: a systematic review and meta-analysis. JAMA Netw Open. 2023;6(5):e2313162. DOI: https://doi.org/10.1001/jamanetworkopen.2023.13162
- Grobman WA, Rice MM, Reddy UM, et al. Labor induction versus expectant management in low-risk nulliparous women. N Engl J Med. 2018;379(6):513–523. DOI: https://doi.org/10.1056/NEJMoa1800566
- Wennerholm UB, Saltvedt S, Wessberg A, et al. Induction of labour at 41 weeks versus expectant management and induction of labour at 42 weeks (SWEdish Post-term Induction Study, SWEPIS): multicentre, open label, randomised, superiority trial. BMJ. 2019;367:l6131. DOI: https://doi.org/10.1136/bmj.l6131
- Keulen JK, Bruinsma A, Kortekaas JC, et al. Induction of labour at 41 weeks versus expectant management until 42 weeks (INDEX): multicentre, randomised non-inferiority trial. BMJ. 2019;364:l344. DOI: https://doi.org/10.1136/bmj.l344
- Walker KF, Bugg GJ, Macpherson M, et al. Randomized trial of labor induction in women 35 years of age or older. N Engl J Med. 2016;374(9):813–822. DOI: https://doi.org/10.1056/NEJMoa1509117
- Management of full-term nulliparous individuals without a medical indication for delivery: ACOG Clinical Practice Update. American College of Obstetricians and Gynecologists. Obstet Gynecol. 2025;145(1):e45–e50. DOI: https://doi.org/10.1097/AOG.0000000000005783
- Quality-improvement strategies for safe reduction of primary cesarean birth: ACOG Committee Statement No. 17. American College of Obstetricians and Gynecologists. Obstet Gynecol. 2025;145(5):542–552. DOI: https://doi.org/10.1097/AOG.0000000000005888
- Wood R, Freret TS, Clapp M, Little S. Rates of induction of labor at 39 weeks and cesarean delivery following publication of the ARRIVE trial. JAMA Netw Open. 2023;6(8):e2328274. DOI: https://doi.org/10.1001/jamanetworkopen.2023.28274
- Nethery E, Levy B, McLean K, Sitcov K, Souter VL. Effects of the ARRIVE (A Randomized Trial of Induction Versus Expectant Management) trial on elective induction and obstetric outcomes in term nulliparous patients. Obstet Gynecol. 2023;142(2):242–250. DOI: https://doi.org/10.1097/AOG.0000000000005217
- Gilroy LC, Al-Kouatly HB, Minkoff HL, McLaren RA Jr. Changes in obstetrical practices and pregnancy outcomes following the ARRIVE trial. Am J Obstet Gynecol. 2022;226(5):716.e1–716.e12. DOI: https://doi.org/10.1016/j.ajog.2022.02.003
- Main EK, Chang S, Cheng YW, Rosenstein MG, Lagrew DC. Hospital-level variation in the frequency of cesarean delivery among nulliparous women who undergo labor induction. Obstet Gynecol. 2020;136(6):1179–1189. DOI: https://doi.org/10.1097/AOG.0000000000004139
- Sentilhes L, et al. Study protocol FRENCH-ARRIVE: labor induction in low-risk nulliparous women at 39 weeks of gestation to reduce cesarean births: a randomized trial of induction versus expectant management in France. BMC Pregnancy Childbirth. 2026;26(1):693. DOI: https://doi.org/10.1186/s12884-026-09193-z