Amniotic Fluid EmbolismSMFM 2026 · Interactive checklist

Does uterine hyperstimulation cause amniotic fluid embolism?

A question that matters for this checklist, because the answer determines whether a clinician standing at the bedside of a woman with sudden hypertonus and fetal bradycardia should reach for a tocolytic. They should not.

This page is LiveEvidence analysis, not SMFM text The 2026 statement does not address uterine activity, prevention, or tocolysis. Everything below is an independent reading of the published literature, presented so that it can be checked and disagreed with. The references are numbered as on the Evidence page.

Short answer

The association appears throughout the textbook literature and largely disappears in the controlled data. The specific claim that excessive uterine activity causes AFE does not survive scrutiny, and the causal direction is probably backwards.

Where “tetanic contraction” as a risk factor comes from

It is a legacy claim. Reviews still list strong, frequent or tetanic uterine contractions among the risk factors, but the citations lead back to case series and autopsy reports from the pre-1990 era. No controlled study of AFE has ever used tachysystole, Montevideo units or hypertonus as an exposure variable. The claim has never been tested. It has only been repeated.

Note what is absent from the risk-factor list in the most rigorous evidence-based review of AFE (ref 24): no tumultuous labour, no hypertonus, no tachysystole. What it lists instead is advanced maternal age, placental abnormalities, operative deliveries, eclampsia, polyhydramnios, cervical laceration and uterine rupture — a list of portals of entry.

What the controlled data show

ExposureKramer, Canada, 3M births (ref 22)Abenhaim, US, 3M births (ref 23)Fitzpatrick, UKOSS (ref 25)Fitzpatrick, INOSS 5 countries (ref 11)
Induction, any methodaOR 1.8 (1.3–2.7)aOR 1.5 (0.9–2.3), NSraisedraised
Prostaglandin inductionaOR 6.46 (3.47–12.01)aOR 2.46 (1.17–5.15)
Oxytocin in labouraOR 2.80 (1.48–5.29)aOR 1.31 (0.83–2.07), NS
Prostaglandin + oxytocinaOR 2.43
DystociaaOR 0.6 (0.4–0.8)aOR 0.4 (0.2–0.7)

Oxytocin is not associated with AFE. In the largest and methodologically cleanest study — pooled prospective population-based surveillance from five countries with agreed case definitions — the odds of AFE were significantly raised with prostaglandin induction but not with oxytocin use in labour (ref 11). Oxytocin is the agent that produces tachysystole, it is titratable, and exposure is dose-dependent and prolonged. If contraction force were the causal pathway, oxytocin is where the signal would be. It is not there.

The gradient runs the wrong way. In UKOSS, prostaglandin without oxytocin carried the highest odds, oxytocin alone lower, and both together lower still (ref 25). Adding the more potent uterotonic lowered the odds. Under a hyperstimulation model, prostaglandin plus oxytocin should be the worst cell in the table. It is the best of the three.

Dystocia is protective, consistently. Both three-million-birth cohorts found dystocia significantly protective (refs 22, 23, tabulated in ref 24). Dystocia is the clinical antonym of tumultuous labour. A hypothesis that says forceful labour causes AFE has to explain why obstructed, inefficient labour is protective in six million births.

The original registry found nothing. Clark's 1995 national registry — the very case series usually cited for the tumultuous-labour claim — reported no correlation with either prolonged labour or oxytocin use (ref 21).

The causal direction is probably backwards

SMFM has said so since 2016: abnormalities of uterine tone, hypo- or hypertonic, described commonly in cases of AFE may be the consequence of uterine hypoperfusion secondary to profound maternal shock and hypoxia with massive catecholamine release, rather than the cause (ref 2). The same guideline describes the antepartum presentation as decelerations, loss of variability and terminal bradycardia as oxygenated blood is shunted away from the uterus, with catecholamine-induced uterine hypertonus causing a further decline in uterine perfusion.

So the hypertonic contraction observed at the onset of an AFE is a sign of the event in progress, not its trigger. It has the same status as the fetal bradycardia beside it on the strip.

There is a further physiological problem with the mechanical model, flagged here as reasoning rather than as a sourced claim: during a tetanic contraction, intrauterine pressure exceeds uterine venous pressure and uteroplacental venous drainage stops. A pressure gradient that closes the exit route is a poor candidate for a mechanism that forces material into the venous circulation.

The pathophysiology has in any case moved away from mechanics entirely. Amniotic fluid is soluble in blood and cannot obstruct; the quantities of fetal debris involved are too small to occlude enough pulmonary vasculature to explain the haemodynamics (ref 4); squamous cells are not the obstructing agent (ref 26). Once the syndrome is immune rather than embolic, contraction force loses its role. What remains explanatory is the portal of entry — which is exactly what the surviving risk factors describe.

Should the checklist include preventing hyperstimulation?

No — and adding it would make the document worse in three ways.

It is the wrong document. The SMFM statement is explicitly scoped to initial management on the labour and delivery unit after the event has declared itself. Prevention belongs in a labour-management guideline. A cognitive aid used during a resuscitation should contain only actions that can be taken in the next four minutes.

The evidence does not support it. Recommending an intervention to prevent AFE requires believing that uterine activity is on the causal path. The oxytocin data argue against it.

It would be actively dangerous. A clinician who believes hypertonus causes AFE, standing at the bedside of a woman with sudden hypertonus and fetal bradycardia, will reach for a tocolytic. Published case reports describe exactly this sequence — terbutaline given for fetal bradycardia during cervical ripening, followed within minutes by maternal deterioration and emergency caesarean. Tocolysis in a patient whose actual problem is acute pulmonary hypertension, right ventricular failure and impending cardiovascular collapse treats the epiphenomenon and worsens the disease.

What the checklist should say instead — the two lines this tool adds

1. Uterine hypertonus and fetal bradycardia at onset are signs of the AFE, not causes of it. Do not give tocolytics — terbutaline, nitroglycerin or magnesium — in suspected AFE. They worsen hypotension and right ventricular failure.

2. Stop an oxytocin infusion already running at the moment of collapse, then restart it after delivery for atony prophylaxis — a sequencing point the current checklist leaves ambiguous, since it says to give oxytocin prophylaxis without addressing an infusion in progress.

Both appear on the live checklist, clearly marked as LiveEvidence annotations rather than SMFM text.

Uncertainty, stated

Absence of evidence is not evidence of absence. Tachysystole has never been measured as an exposure, so the honest statement is unsupported, not refuted. The induction association is real, consistent across five countries and unexplained; confounding by indication is severe, since prostaglandin ripening selects for post-dates, intrauterine fetal death and unfavourable cervices. Prostaglandins also cause tachysystole, so a contraction-mediated pathway is not formally excluded — the oxytocin null result is the strongest single argument against it, not a proof. Ref 28, a methodological case study of how the uterine-stimulant association was constructed, is listed as a pointer only; its full text has not been read for this page and the direction of its argument is not characterised here.

References for this page
2
GUIDELINE
Pacheco LD, Saade G, Hankins GD, Clark SL; Society for Maternal-Fetal Medicine. Amniotic fluid embolism: diagnosis and management. Am J Obstet Gynecol. 2016;215(2):B16-B24.
doi:10.1016/j.ajog.2016.03.012
SMFM Clinical Guideline No. 9. Diagnostic framing, the >80% DIC figure, and the statement that uterine tone abnormalities may be a consequence of shock rather than the cause.
4
NARRATIVE REVIEW
Clark SL. Amniotic fluid embolism. Obstet Gynecol. 2014;123(2 Pt 1):337-348.
doi:10.1097/AOG.0000000000000107
Clinical Expert Series. Incidence of 2-6 per 100,000, case fatality range, DIC as the confirming finding, and the immune rather than embolic model of the syndrome.
11
COHORT
Fitzpatrick KE, van den Akker T, Bloemenkamp KWM, Deneux-Tharaux C, Kristufkova A, Li Z, et al. Risk factors, management, and outcomes of amniotic fluid embolism: a multicountry, population-based cohort and nested case-control study. PLoS Med. 2019;16(11):e1002962.
doi:10.1371/journal.pmed.1002962
INOSS pooled surveillance from five countries. Cited by the 2026 statement for tranexamic acid; also the source of the finding that oxytocin in labour is not significantly associated with AFE while prostaglandin induction is.
21
REGISTRY
Clark SL, Hankins GD, Dudley DA, Dildy GA, Porter TF. Amniotic fluid embolism: analysis of the national registry. Am J Obstet Gynecol. 1995;172(4 Pt 1):1158-1169.
doi:10.1016/0002-9378(95)91474-9
The original US registry. Reported no correlation between AFE and either prolonged labour or oxytocin use, and proposed the anaphylactoid rather than embolic model.
22
COHORT
Kramer MS, Rouleau J, Baskett TF, Joseph KS. Amniotic-fluid embolism and medical induction of labour: a retrospective, population-based cohort study. Lancet. 2006;368(9545):1444-1448.
doi:10.1016/S0140-6736(06)69607-4
Three million Canadian deliveries. Medical induction aOR 1.8 (1.3-2.7); dystocia protective. The paper that put induction on the AFE risk-factor list.
23
COHORT
Abenhaim HA, Azoulay L, Kramer MS, Leduc L. Incidence and risk factors of amniotic fluid embolisms: a population-based study on 3 million births in the United States. Am J Obstet Gynecol. 2008;199(1):49.e1-49.e8.
doi:10.1016/j.ajog.2007.11.061
Three million US births. Induction aOR 1.5 (0.9-2.3), not significant; dystocia again protective (aOR 0.4).
24
REVIEW
Conde-Agudelo A, Romero R. Amniotic fluid embolism: an evidence-based review. Am J Obstet Gynecol. 2009;201(5):445.e1-445.e13.
doi:10.1016/j.ajog.2009.04.052
Side-by-side tabulation of the two large cohorts. Its risk-factor list contains no entry for tumultuous labour, uterine hypertonus or tachysystole.
25
COHORT
Fitzpatrick KE, Tuffnell D, Kurinczuk JJ, Knight M. Incidence, risk factors, management and outcomes of amniotic-fluid embolism: a population-based cohort and nested case-control study. BJOG. 2016;123(1):100-109.
doi:10.1111/1471-0528.13300
UKOSS. Prostaglandin without oxytocin aOR 6.46; oxytocin without prostaglandin aOR 2.80; both together aOR 2.43 — a gradient that runs the wrong way for a contraction-force mechanism.
26
LETTER
Funk M, Damron A, Bandi V, Aagaard K, Szigeti R, Clark S. Pulmonary vascular obstruction by squamous cells is not involved in amniotic fluid embolism. Am J Obstet Gynecol. 2018;218(4):460-461.
doi:10.1016/j.ajog.2017.12.225
Cited by the 2026 statement for the pathophysiology update. Removes mechanical obstruction from the causal account.
27
COMMENTARY
Lisonkova S, Kramer MS. Amniotic fluid embolism: a puzzling and dangerous obstetric problem. PLoS Med. 2019;16(11):e1002976.
doi:10.1371/journal.pmed.1002976
Perspective accompanying ref 11, by the author of the 2006 induction cohort.
28
METHODOLOGICAL
Wagner M. From caution to certainty: hazards in the formation of evidence-based practice — a case study on evidence for an association between the use of uterine stimulant drugs and amniotic fluid embolism. Paediatr Perinat Epidemiol. 2005;19(2):173-176.
doi:10.1111/j.1365-3016.2005.00630.x
A documented case study of how the uterine-stimulant/AFE association was constructed. Listed as a pointer only — the full text has not been read for this site and the direction of its argument is not characterised here.