You are on page 1of 8

Incidence and Risk Factors for Amniotic-Fluid

Embolism
Marian Knight, MBChB, DPhil, Derek Tuffnell, MBChB, Peter Brocklehurst, MBChB, MSc, Patsy Spark, BSc,
and Jennifer J. Kurinczuk, MBChB, MD, on behalf of the UK Obstetric Surveillance System

OBJECTIVE: To estimate the incidence of amniotic-fluid newborns of women with antenatal amniotic-fluid em-
embolism and to describe risk factors, management, and bolism died (perinatal mortality 135 per 1,000 total births,
outcomes. 95% CI 45–288). Women who died were significantly
METHODS: Through a population-based cohort study more likely to be from ethnic-minority groups (adjusted
and nested case-control analysis, using the UK Obstetric OR 11.8, 95% CI 1.40 –99.5).
Surveillance System, we identified 60 women in the CONCLUSION: High-quality supportive care can result
United Kingdom who had an amniotic-fluid embolism in good maternal outcomes after amniotic-fluid embo-
between February 2005 and February 2009 and 1,227 lism. Clinicians should consider both the risks and ben-
women for the control group. We investigated the po- efits of induction and cesarean delivery because more
tential factors underlying amniotic-fluid embolism using restricted use may result in a decrease in the number of
an exploratory logistic regression analysis to estimate women suffering a potentially fatal amniotic-fluid embo-
odds ratios (ORs) and 95% confidence intervals (CIs). lism. The observed increased risk of fatality in ethnic-
RESULTS: Sixty cases of amniotic-fluid embolism were minority women may be associated with differences in
reported, an estimated incidence of 2.0 per 100,000 underlying medical conditions or access to care, and
deliveries (95% CI 1.5–2.5). Amniotic-fluid embolism oc- clinicians should that ensure appropriate services are
currence was significantly associated with induction of provided to minimize this risk.
labor (adjusted OR 3.86, 95% CI 2.04 –7.31) and multiple (Obstet Gynecol 2010;115:910–7)
pregnancy (adjusted OR 10.9, 95% CI 2.81– 42.7); an LEVEL OF EVIDENCE: II
increased risk also was noted in older, ethnic-minority
women (adjusted OR 9.85, 95% CI 3.57–27.2). Cesarean
delivery was associated with postnatal amniotic-fluid
embolism (adjusted OR 8.84, 95% CI 3.70 –21.1). Twelve
women died (case fatality 20%, 95% CI 11–32%); 5 of 37
A mniotic-fluid embolism is a leading cause of
maternal mortality, with recent rates appearing to
have increased in some countries1,2 although not in
others.3,4 There were no changes in diagnostic criteria
From the National Perinatal Epidemiology Unit, University of Oxford, Oxford, or methods of case ascertainment to account for the
United Kingdom; and the Bradford Hospitals NHS Trust, Bradford, United
Kingdom. observed rise in the number of deaths in the United
Dr. Knight is funded by the National Coordinating Centre for Research Capacity
Kingdom and Australia, but the rarity of the disorder
Development of the Department of Health. Dr. Kurinczuk was partially funded makes it particularly difficult to investigate whether
by a National Public Health Career Scientist Award from the Department of the larger number of deaths reflects an increase in the
Health and NHS R&D (PHCS022).
incidence of the condition, an increase in the case
The authors thank the UK Obstetric Surveillance System reporting clinicians
who reported cases and completed the data-collection forms.
fatality, or a chance finding. Two recent retrospective
analyses of incidence using large national hospital
This paper reports on an independent study which is funded by the Policy
Research Programme in the Department of Health. The views expressed are not databases in the United States and Canada,3,4 did not
necessarily those of the Department. find a temporal trend in the occurrence of fatal or
Corresponding author: Marian Knight, National Perinatal Epidemiology Unit, nonfatal cases.
University of Oxford, Old Road Campus, Oxford, UK; e-mail: marian. Prospective surveillance of amniotic-fluid embo-
knight@npeu.ox.ac.uk.
lism has been undertaken in the United Kingdom
Financial Disclosure
The authors did not report any potential conflicts of interest.
through two routes. In a passive system, between
1997 and 2004, cases were reported on a voluntary
© 2010 by The American College of Obstetricians and Gynecologists. Published
by Lippincott Williams & Wilkins. basis to the United Kingdom Amniotic Fluid Embo-
ISSN: 0029-7844/10 lism Register5; since 2005, cases have been reported

910 VOL. 115, NO. 5, MAY 2010 OBSTETRICS & GYNECOLOGY


actively on a national basis through the routine to return cards indicating a “nil report” so that we
monthly mailing of the UK Obstetric Surveillance could monitor card-return rates and confirm the
System.6 The aim of this study was to estimate the denominator to calculate the incidence. Of note, any
current incidence of amniotic-fluid embolism in the women with amniotic-fluid embolism delivering in
United Kingdom and to investigate whether there is other settings in the United Kingdom, including either
any evidence to suggest that disease incidence is in a midwife-led unit or at home, would be transferred
changing. In addition, we sought to describe risk to a consultant-led unit and therefore would be
factors, current management, case fatality, and other captured by the UK Obstetric Surveillance System.
outcomes. When a clinician reported a case, they were sent a
data-collection form asking for details of presentation,
management, and outcomes. All data were collected
MATERIALS AND METHODS anonymously.
We carried out a prospective, population-based co- Data were double entered into a customized data-
hort and nested case-control study. We identified base. Two of the authors (D.T., M.K.) reviewed all cases
cases through the monthly mailing of the UK Obstet- in which postpartum hemorrhage was reported as the
ric Surveillance System6,7 between February 2005 and first symptom of amniotic-fluid embolism against the
February 2009. Clinicians were asked to report any diagnostic criteria (Box 1) to determine independently
woman diagnosed with amniotic-fluid embolism with whether amniotic-fluid embolism was the most likely
symptoms and signs consistent with amniotic-fluid diagnosis.
embolism or any woman in whom amniotic-fluid The Centre for Maternal and Child Enquiries,
embolism was confirmed by biopsy or postmortem which has collected information on all maternal
examination (Box 1). deaths in the United Kingdom for more than 50 years,
provided information (year of birth and date of
Box 1. Diagnostic Criteria for Amniotic-Fluid
diagnosis) about any maternal deaths from amniotic-
Embolism
fluid embolism occurring during the study period.
IN THE ABSENCE OF ANY OTHER CLEAR CAUSE These were compared with maternal deaths reported
Either through the UK Obstetric Surveillance System; no
Acute maternal collapse with one or more of the follow- additional cases were identified.
ing features: All analyses were carried out using STATA 10
• Acute fetal compromise software (StataCorp LP, College Station, TX). Inci-
• Cardiac arrest dence with 95% confidence intervals (CIs) was calcu-
• Cardiac rhythm problems
• Coagulopathy lated using the most recently available birth data
• Hypotension (2005 to 2007) as a proxy for February 2005 to
• Maternal hemorrhage February 20098 –10; the total estimated number of
• Premonitory symptoms, eg, restlessness, numbness, maternities (women delivering) was 3,049,100.
agitation, tingling We investigated the potential factors underlying
• Seizure
• Shortness of breath amniotic-fluid embolism using an exploratory logistic
Excluding: women with maternal hemorrhage as the first regression analysis to estimate odds ratios (ORs) and
presenting feature in whom there was no evidence of 95% CIs. We used national data where available to
early coagulopathy or cardio-respiratory compromise conduct a univariable analysis and information from
1,227 women in the nested control group to conduct
Or
an adjusted analysis because national data do not
Women in whom the diagnosis was made at postmor-
tem examination with the finding of fetal squames or have sufficient information on potential confounders.
hair in the lungs The women in the control group were identified by
the UK Obstetric Surveillance System reporters as the
The UK Obstetric Surveillance System method- two women delivering in the same hospital immedi-
ology has been described in detail elsewhere.6 In ately before other UK Obstetric Surveillance System
brief, every month, the UK Obstetric Surveillance cases.11 A full regression model was developed by
System case-notification cards were sent to nominated including both explanatory and potential confound-
reporting clinicians in each hospital in the United ing factors in a core model if there was a preexisting
Kingdom with a consultant-led maternity unit, with a hypothesis or evidence to suggest they were causally
tick box to indicate whether they had seen a woman related to amniotic-fluid embolism, for example, in-
with amniotic-fluid embolism. They also were asked duction of labor. Continuous variables were tested for

VOL. 115, NO. 5, MAY 2010 Knight et al Amniotic-Fluid Embolism in the UK 911
departure from linearity by the addition of quadratic CI 1.5–2.5). There was no significant change in
terms, and potential interactions were tested by the incidence over the 4 years of the study, although the
addition of interaction terms between all variables in study has low power to detect any difference owing to
the model and subsequent likelihood-ratio testing on small case numbers (data not shown). The nine
removal. P⬍.05 was considered evidence of a signif- women whose cases were excluded all had primary
icant interaction or departure from linearity. The postpartum hemorrhage with no evidence of addi-
association of amniotic-fluid embolism with cesarean tional features, such as cardiac or respiratory compro-
delivery was examined in a regression model includ- mise or early coagulopathy, to suggest amniotic-fluid
ing only those women who had amniotic-fluid embo- embolism.
lism after delivery to exclude cases in which cesarean Characteristics of women with amniotic-fluid em-
delivery was likely to be a consequence rather than a bolism are shown in Table 1. Unadjusted analysis
cause of the amniotic-fluid embolism. The analysis suggested a linear increase in risk of amniotic-fluid
had 80% power at the 5% level of statistical signifi- embolism with increasing age (OR 1.12 for every one
cance to detect an OR for induction of labor of 2.6 or year increase in age, 95% CI 1.07–1.18). For ease of
greater. presentation we have shown the data in binary groups
To estimate the proportion of cases attributable to (OR 2.59, 95% CI 1.53– 4.41 in women aged 35 and
specific causes, we calculated population-proportional over compared with women aged under 35). There
attributable risks using adjusted ORs and the propor- was a nonsignificant association with maternal ethnic-
tion of cases exposed.12 The UK Obstetric Surveil- ity (OR 1.18, 95% CI 0.61–2.27 in ethnic minority
lance System general methodology (04/MRE02/45) women compared with white women). The adjusted
and this study (04/MRE02/46) were approved by the analysis revealed a significant interaction between
London Multicentre Research Ethics Committee. ethnicity and age (P⬍.001), suggesting that the asso-
ciation with age varies between different ethnic
RESULTS groups. We therefore have presented stratified results
All 229 eligible U.K. hospitals contributed data to the that show the highest odds in ethnic-minority women
UK Obstetric Surveillance System during the study older than 35 years (OR 9.85, 95% CI 3.57–27.2)
period (100% participation). Data collection was com- (Table 2). In the U.K. population, the proportions of
plete for 97% of cases (Fig. 1). different ethnicities in the ethnic-minority group were
There were 60 confirmed cases in an estimated Asian 47%, black 31%, mixed race 6%, Chinese 5%,
3,049,100 maternities,8 –10 representing an estimated and other ethnic-minority groups 10%. We were
incidence of 2.0 cases per 100,000 maternities (95% unable to identify any significantly raised risk associ-
ated with a particular minority ethnicity, but note that
this may be the result of small group sizes and limited
Cases notified
N=93
study power. After adjustment, no other sociodemo-
graphic factors significantly affected the risk of amni-
Excluded: n=21; 23% otic-fluid embolism.
Duplicates: 4
The occurrence of amniotic-fluid embolism was
Subsequently
reported by significantly associated with induction of labor and
clinicians as not multiple pregnancy (Table 2). Corresponding popu-
being cases: 17
lation proportional attributable risks are 35% for
No data received induction of labor, 13% for ethnic-minority women 35
n=3; 3% years or older and 7% for multiple pregnancy.
Cases identified Twenty-six women had amniotic-fluid embolism
Data collection
from additional
forms received after delivery; 19 (73%) of these occurred after cesar-
reporters
n=69
n=0 ean delivery. The adjusted OR for delivery by cesar-
Did not meet ean in women who had amniotic-fluid embolism after
case definition
n=9 delivery was 8.84 (95% CI 3.70 –21.1). Ten of the 19
women who had amniotic-fluid embolism after cesar-
Amniotic fluid
embolism ean delivery were not in labor at the time of the
n=60 cesarean. Five of these had elective cesarean opera-
Fig. 1. Case reporting and completeness of data collection. tions (one at maternal request and four after previous
Knight. Amniotic-Fluid Embolism in the UK. Obstet Gynecol cesarean deliveries), and four had emergency cesar-
2010. ean deliveries (three for fetal distress and one for

912 Knight et al Amniotic-Fluid Embolism in the UK OBSTETRICS & GYNECOLOGY


Table 1. Characteristics of Women With Amniotic-Fluid Embolism
Fatal Cases* Non-Fatal Cases* Total Cases*
Characteristic (nⴝ12) (nⴝ48) (nⴝ60)

Sociodemographic characteristics
Age 35 y or older 8 (67) 17 (35) 25 (42)
Ethnic minority group 5 (45) 7 (15) 12 (20)
Professional or managerial occupation 4 (33) 19 (45) 23 (38)
BMI 30 or higher 4 (40) 11 (26) 15 (25)
Current smoking 1 (8) 13 (27) 14 (23)
History of atopic disease 1 (8) 15 (31) 16 (27)
Coexisting medical problems 3 (25) 18 (38) 21 (35)
Pregnancy characteristics
Primiparity 3 (25) 19 (40) 22 (37)
Twin pregnancy 2 (17) 3 (6) 5 (8)
Induction of labor using any method 6 (50) 22 (46) 28 (47)
Induction of labor with vaginal prostaglandin 5 (42) 19 (40) 24 (40)
Oxytocin in labor† 1 (11) 12 (38) 13 (22)
Hyperstimulation 0 (0) 5 (16) 5 (8)
Cesarean delivery 9 (75) 39 (81) 48 (80)
Preterm delivery 4 (33) 5 (11) 9 (15)
AFE presenting before or at delivery 7 (58) 26 (54) 33 (55)
Management
Obstetrician present at time of event or arrived 11 (92) 42 (89) 53 (90)
within 15 min
Anesthetist present at time of event or arrived 10 (83) 38 (83) 48 (83)
within 15 min
Neither obstetrician nor anesthetist present at 7 (58) 19 (41) 26 (43)
time of event
BMI, body mass index; AFE, amniotic-fluid embolism.
Data are n (%).
* Percentages of those with data.

Defined as any use of oxytocin to enhance uterine contractions, whether as part of an induction process or to augment spontaneous
labor.

suspected placenta previa); for one woman, the The majority of women (55, 92%) had ruptured
reason for cesarean delivery is unknown. The pop- membranes at or before the time of presentation
ulation-proportional attributable risk associated with amniotic-fluid embolism. In those women who
with cesarean delivery (before amniotic-fluid em- had amniotic-fluid embolism before delivery, the
bolism) was 62%. amniotic-fluid embolism presented a median of 45
Information on time of event was missing for one minutes after membrane rupture (range 0 minutes
woman. Thirty-three women (56%) had symptoms or to 6 days after); in four women the amniotic-fluid
signs of embolism at or before delivery; the remain- embolism presented at the time of membrane rup-
der had symptoms or signs after delivery. The amni- ture, and a further three women presented within 10
otic-fluid embolism occurred at a median gestation of minutes of membrane rupture. An obstetrician was
39 weeks (range 28 – 42) (Fig. 2) and presented clini- present at the time of presentation in 28 cases (47%)
cally within a 6-hour range around delivery (range 2 and an anesthetist in 27 (45%); neither was present in
hours, 18 minutes before delivery to 4 hours after 26 cases (43%) (Table 1).
delivery). Symptoms and signs at presentation are Twenty-eight women (85%) who presented with
shown in Table 3. All the women had at least one, and amniotic-fluid embolism at or before delivery were
16 (27%) women had at least four of the five cardinal delivered by cesarean. A range of management strat-
features of shortness of breath, hypotension, maternal egies were used after amniotic-fluid embolism in
hemorrhage, coagulopathy, and premonitory symp- addition to supportive therapies (Table 4). Supportive
toms. Fetal distress occurred before maternal collapse therapies alone, defined as fluids and blood products
in 19 cases (32%); in six (32%) of these cases, the only, were used in 35 women (58%); 15 women (25%)
woman was delivered urgently as a consequence of had hysterectomy to control hemorrhage, and 15
fetal distress and collapsed after delivery. women (25%), including one who also had a hyster-

VOL. 115, NO. 5, MAY 2010 Knight et al Amniotic-Fluid Embolism in the UK 913
Table 2. Maternal Risk Factors for Amniotic-Fluid Embolism
Number of UKOSS
Women* UKOSS National Case-Control
Cases* Controls* (National Case-Control Comparison Data [Adjusted
Risk Factor (nⴝ60) (nⴝ1,227) Data) Data Data OR (95% CI)]†

Sociodemographic factors
Ethnicity
White
Younger than 35 y 29 (51) 751 (63) N/A 1‡ N/A 1‡
35 y or older 16 (28) 223 (19) N/A 1.86 (0.99–3.48) N/A 1.54 (0.72–3.28)
Nonwhite
Younger than 35 y 5 (9) 182 (15) N/A 0.71 (0.27–1.86) N/A 0.60 (0.17–2.13)
35 y or older 7 (12) 38 (3) N/A 4.77 (1.96–11.6) N/A 9.85 (3.57–27.2)
Socioeconomic group
Managerial and 31 (57) 334 (30) N/A 1.70 (0.98–2.96) N/A 1.55 (0.79–3.04)
professional
occupations
Other occupations 23 (43) 766 (70) N/A 1‡ N/A 1‡
Smoking status
Never/ex smoker 45 (76) 940 (78) N/A 1‡ N/A 1‡
Current smoker 14 (24) 258 (22) N/A 1.13 (0.61–2.10) N/A 1.28 (0.57–2.90)
Booking body mass index
Lower than 30 37 (71) 886 (83) N/A 1‡ N/A 1‡
30 or higher 15 (29) 186 (17) N/A 1.93 (1.04–3.59) N/A 1.65 (0.82–3.33)
Factors related to pregnancy
and reproduction
Parity
0 22 (37) 525 (43) N/A 1‡ N/A 1‡
1 or more 38 (63) 696 (57) N/A 1.30 (0.76–2.23) N/A 1.03 (0.53–2.00)
Multiple pregnancy
No 55 (92) 1,211 (99) 652,915 (98)§ 1‡ 1‡ 1‡
Yes 5 (8) 13 (1) 10,137 (2)§ 8.47 (2.92–24.6) 5.86 (1.83–14.5) 10.9 (2.81–42.7)
Induction of labor (medical
and surgical
methods)
No 32 (53) 953 (78) 378,434 (80)㛳 1‡ 1‡ 1‡
Yes 28 (47) 269 (22) 96,389 (20)㛳 3.10 (1.83–5.24) 3.44 (1.99–5.89) 3.86 (2.04–7.31)
Delivery by cesarean
No 12 (20) 925 (76) 452,709 (76)㛳 1‡ 1‡ Not included¶
Yes 48 (80) 299 (24) 145,322 (24)㛳 12.4 (6.49–23.6) 12.5 (6.52–25.8) Not included¶
Placenta praevia
No 58 (98) 1,215 (100) N/A 1‡ N/A Not included
Yes 1 (2) 2 (0.2) N/A 10.5 (0.94–117.2) N/A Not included
UKOSS, UK Obstetric Surveillance System; OR, odds ratio; CI, confidence interval; N/A, national data not available.
Data are n (%) or unadjusted odds ratio (95% CI) unless otherwise specified.
* Percentages of those with complete data.

Adjusted for all the other factors indicated in the table.

Baseline comparison group.
§
Data for maternities in England and Wales 2006, Office for National Birth Statistics 2006.

Data for maternities in England 2006 –7, source Hospital Episode Statistics 2006 –2007.

See text for adjusted OR in women who had amniotic-fluid embolism after delivery.

ectomy, were treated with factor VIIa. In total, five hour, 40 minutes after the acute event (range 0
women (8%) had a combination of different treatment minutes to 23 hours, 18 minutes). Women who died
strategies in addition to supportive therapies. Women were significantly more likely to be from an ethnic-
were transfused a median of 6.5 units of blood (range minority group than were those who survived (OR
0 – 62 units). 4.64, 95% CI 1.11–19.5); this association persisted
Twelve women died (case fatality rate 20%, 95% after adjustment for differences in age, socioeconomic
CI 11–32%) (Table 1). All the women died within 1 status, body mass index, and parity (adjusted OR
day of the amniotic-fluid embolism at a median of 1 11.8, 95% CI 1.40 –99.5). There were no other signif-

914 Knight et al Amniotic-Fluid Embolism in the UK OBSTETRICS & GYNECOLOGY


16 Table 4. Management Techniques Used
Event after delivery
14 Survivors Deaths
Event before or at delivery
Technique Used (nⴝ48)* (nⴝ12)* P†
12
Hysterectomy 12 (25) 3 (25) 1.00
10 Exchange transfusion 3 (6) 0 (0) 1.00
Women (n)

Plasma exchange 4 (8) 0 (0) .57


8
Factor VIIa 14 (29) 1 (8) .26
6 Supportive therapies 26 (54) 9 (75) .33
only
4
Data are n (%).
* Some women had multiple techniques, therefore totals are
2
greater than 100%.

Fisher’s exact test.
0
27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42
Gestation at event/delivery (weeks)
There was one stillbirth and four neonatal deaths
Fig. 2. Gestational age at amniotic fluid embolism (antena-
tal cases) or delivery (postnatal cases).
(all due to asphyxia before birth) among the 37
Knight. Amniotic-Fluid Embolism in the UK. Obstet Gynecol
neonates for whom outcomes are known and who
2010. were born to mothers who had amniotic-fluid embo-
lism at or before delivery, giving a perinatal mortality
icant differences between fatal and nonfatal cases, rate of 135 per 1,000 total births (95% CI 45–288).
although this analysis has limited power to detect Neither mother nor neonate survived in three cases.
differences. Eight women who survived had other None of the neonates born to mothers who had
severe morbidities: four had cerebral neurologic in- amniotic-fluid embolism after delivery died.
jury, two suffered a thrombotic event, one had septi-
DISCUSSION
cemia, and one had renal failure. Forty-five of the 48
women who survived (94%) were admitted to inten- The incidence of amniotic-fluid embolism as esti-
sive care. Only 3 of the 12 women who died (25%) mated by this 4-year prospective national study is 2.0
were admitted to intensive care; the remainder died per 100,000 maternities (95% CI 1.5–2.5), which is
before admission. The median length of stay in inten- significantly lower than the rate documented in retro-
sive care was 3 days (range 1–39 days). spective reviews of population-based hospital dis-
charge databases in Canada (6.1 cases per 100,000
deliveries, 95% CI 5.3–7.14) and the United States (7.7
Table 3. Features of Amniotic-Fluid Embolism at
cases per 100,000 births, 95% CI 6.7– 8.73). As we
Presentation
demonstrated, amniotic-fluid embolism is difficult to
Women Women Exhibiting diagnose; of 86 nonduplicate cases reported, further
Exhibiting Feature as the First examination of the records revealed a more likely
Feature*† Symptom or Sign
(nⴝ60) of AFE*† (nⴝ60) diagnosis in 26 (30%). Hospital discharge databases
do not allow for examination of sufficiently detailed
Maternal hemorrhage 39 (65) 1 (2) clinical information to either confirm or refute the
Hypotension 38 (63) 5 (8) diagnosis of amniotic-fluid embolism and, therefore,
Shortness of breath 37 (62) 12 (20)
Coagulopathy 37 (62) 0 (0) are likely to be subject to a degree of overreporting.
Premonitory symptoms 28 (47) 18 (30) Neither of the previous studies undertook validation
(eg, restlessness, of the diagnosis by examining medical charts. The
agitation, numbness, authors of the Canadian report cite their observed
tingling) associations of amniotic-fluid embolism with compli-
Acute fetal compromise 26 (43) 12 (20)
Cardiac arrest 24 (40) 5 (8) cations including eclampsia, placenta previa, and
Cardiac rhythm 16 (27) 3 (5) abruption as evidence of overreporting; similar asso-
problems ciations were observed in the U.S. study.
Seizures 9 (15) 4 (7) Although a possibility, we believe that our lower
AFE, amniotic-fluid embolism. incidence estimate is unlikely to be due to underre-
Data are n (%). porting of cases because the UK Obstetric Surveil-
* Percentages of those with complete data.

Some women had multiple features, therefore totals are greater lance System is an active, prospective surveillance
than 100%. system in which we require negative reports, ie,

VOL. 115, NO. 5, MAY 2010 Knight et al Amniotic-Fluid Embolism in the UK 915
participating hospitals return a report card every ciated with cesarean delivery in the group who had
month, regardless of whether there are cases to report. amniotic-fluid embolism after delivery, suggesting
We thus can be confident that reports of “no cases” that there indeed may be a causal relationship. The
from an individual hospital are true negative reports. clinical message from this study thus echoes other
Furthermore, all of the fatal cases reported to the studies in noting that cesarean delivery is not risk-
Centre for Maternal and Child Enquiries in the study free7 and that all risks and benefits should be assessed
period had been reported to the UK Obstetric Sur- before a decision for cesarean delivery is made.
veillance System. If the true incidence in the U.K. Older maternal age was also an important factor
population were the same as that estimated in Canada in all the studies, although, in our population, having
(6.1 per 100,000), there would have been 186 cases in adjusted for confounders, the increase in risk with age
total, implying that we had missed 126 nonfatal cases was limited to ethnic-minority women. The small
and that the case fatality rate was only 6% (12 of 186); number of older ethnic-minority women with amni-
levels of both missing cases and fatalities that seem otic-fluid embolism means that we do not have suffi-
highly improbable. cient power to investigate this association further, but
A third explanation for the threefold difference in continued surveillance through the UK Obstetric
incidence estimates is that this reflects a true differ- Surveillance System may allow us to address this in
ence in incidence in the United Kingdom compared the future. In the meantime, clinicians should be
with North America. Given that the population-pro- aware of the increased risk when caring for older
portional attributable risk associated with cesarean pregnant women who are ethnic-minorities.
delivery is 62% and that the cesarean-delivery rate in All the women in our study experienced at least
the United States is 50% higher than in the United one of the cardinal features of shortness of breath,
Kingdom, one might indeed anticipate a higher rate hypotension, hemorrhage, coagulopathy, and pre-
of amniotic-fluid embolism in the United States than monitory symptoms at the time of amniotic-fluid
in the United Kingdom. embolism, and all cases were consistent with the
Induction of labor was associated with a popula- reporting criteria used in a U.S. amniotic-fluid embo-
tion-attributable risk of 35% in our study, suggesting lism registry.13 Premonitory symptoms such as numb-
that, assuming causality, if induction of labor were no ness, tingling, and agitation were the most common
longer performed, 35% of cases of amniotic-fluid initial presenting feature, although acute fetal com-
embolism could be prevented. The practice of induc- promise was the first feature noted in 36% of antenatal
ing labor, with its many potential benefits, clearly will cases. Fetal compromise was noted first as a fetal heart
continue, and amniotic-fluid embolism remains a rate abnormality in the majority of cases; clinicians
very rare complication; nevertheless, clinicians should consider amniotic-fluid embolism as part of
should be aware of both the risks and benefits of the differential diagnosis of a sudden, unexplained
induction because more restricted use may result in a deterioration in the fetal heart rate even in the ab-
decrease in the number of women suffering a poten- sence of maternal symptoms or signs.
tially fatal amniotic-fluid embolism. The mainstay of the management of amniotic-
Both Canadian and U.S. studies show an in- fluid embolism remains supportive therapy, and this
creased rate of cesarean delivery among women with study shows that high-quality supportive care can
amniotic-fluid embolism, which the authors of the result in a good maternal outcome in the majority of
U.S. study interpret as causal. However, as the inves- cases. The results also highlight the use of a number of
tigators discuss, one of the limitations of administra- therapies for amniotic-fluid embolism previously de-
tive databases is that they do not indicate whether the scribed only in case reports. Coagulopathy is a com-
amniotic-fluid embolism occurred before, during, or mon feature of amniotic-fluid embolism (occurring in
after delivery. Our prospectively collected data in- more than 60% of our cases), and, therefore, use of
clude timing of all the events, and we found that 56% recombinant factor VIIa, a procoagulant leading to
of women had amniotic-fluid embolism at or before increased thrombin formation, has been proposed,
delivery, with 44% after delivery. More than 80% of with individual reports of its success.14 –16 In our study,
women who had amniotic-fluid embolism at or before 14 women were treated with factor VIIa for coagu-
delivery subsequently were delivered by cesarean; lopathy; 13 survived.
clearly, in these cases, the cesarean delivery was a Seven women were treated with exchange trans-
consequence and not a cause of amniotic-fluid embo- fusion or plasma exchange, a therapy first described
lism. However, we also found more than an eightfold more than 20 years ago.17 Authors have suggested
increase in the odds of amniotic-fluid embolism asso- variously that clinical improvement is related to the

916 Knight et al Amniotic-Fluid Embolism in the UK OBSTETRICS & GYNECOLOGY


removal of red cell debris and hemoglobin from the based study on 3 million births in the United States. Am J
Obstet Gynecol 2008;199:49.e1– 8.
circulation17 or that the process removes amniotic
fluid and cytokines from the circulation and corrects 4. Kramer MS, Rouleau J, Baskett TF, Joseph KS. Amniotic-fluid
embolism and medical induction of labour: a retrospective,
metabolic acidosis.18 Thus, these therapies should be population-based cohort study. Lancet 2006;368:1444 – 8.
regarded as an extension of supportive care and not as 5. Tuffnell DJ. United Kingdom Amniotic Fluid Embolism Reg-
a substitute. All seven women treated with these ister. BJOG 2005;112:1625–9.
therapies survived, although the small numbers in- 6. Knight M, Kurinczuk JJ, Tuffnell D, Brocklehurst P. The UK
volved means that we should not assume that this Obstetric Surveillance System for rare disorders of pregnancy.
form of therapy is more effective than any other; BJOG 2005;112:263–5.
survival to the point where these techniques are 7. Knight M, Kurinczuk JJ, Spark P, Brocklehurst P. Cesarean
delivery and peripartum hysterectomy. Obstet Gynecol 2008;
possible may put the women in a more favorable 111:97–105.
prognosis group. 8. Office for National Statistics. Key population and vital statistics
One fifth of women with amniotic-fluid embolism (2005). Newport (UK): Office for National Statistics; 2007.
and 1 in 10 of their newborns died. Maternal fatality 9. Office for National Statistics. Key population and vital statistics
was comparable with that in the Canadian and U.S. (2006). Newport (UK): Office for National Statistics; 2008.
studies (13% [95% CI 9 –19%] and 22% [95% CI 10. Office for National Statistics. Key Population and Vital Statis-
16 –28%], respectively). This confirms the view that tics (2007). Newport (UK): Office for National Statistics; 2009.
maternal fatality in large, unselected populations is 11. Knight M, Kurinczuk JJ, Spark P, Brocklehurst P, UKOSS.
Inequalities in maternal health: national cohort study of ethnic
lower than previously suggested. We found that variation in severe maternal morbidities. BMJ 2009;338:b542.
women who died were significantly more likely to
12. Rockhill B, Newman B, Weinberg C. Use and misuse of
come from ethnic-minority groups than were survi- population attributable fractions. Am J Public Health 1998;88:
vors, an association that could not be explained by 15–9.
differences in age, socioeconomic status, body mass 13. Clark SL, Hankins GD, Dudley DA, Dildy GA, Porter TF.
index, or parity. In previous UK Obstetric Surveil- Amniotic fluid embolism: analysis of the national registry.
Am J Obstet Gynecol 1995;172:1158 – 67.
lance System studies,11 we found that the incidence of
severe maternal morbidity is higher among ethnic- 14. Kahyaoglu I, Kahyaoglu S, Mollamahmutoglu L. Factor VIIa
treatment of DIC as a clinical manifestation of amniotic fluid
minority women, and we postulate that this may be embolism in a patient with fetal demise. Arch Gynecol Obstet
due to differences in underlying medical conditions or 2009;280:127–9.
access to care. It is possible that these factors also may 15. Lim Y, Loo CC, Chia V, Fun W. Recombinant factor VIIa
play a role in the increased risk of fatality from after amniotic fluid embolism and disseminated intravascular
coagulopathy. Int J Gynaecol Obstet 2004;87:178 –9.
amniotic-fluid embolism for these women.
16. Prosper SC, Goudge CS, Lupo VR. Recombinant factor VIIa
to successfully manage disseminated intravascular coagulation
REFERENCES from amniotic fluid embolism. Obstet Gynecol 2007;109:
1. Lewis GE, editor. The Confidential Enquiry into Maternal and 524 –5.
Child Health (CEMACH). Saving mothers lives: reviewing 17. Dodgson J, Martin J, Boswell J, Goodall HB, Smith R. Proba-
maternal deaths to make motherhood safer 2003–2005. Lon- ble amniotic fluid embolism precipitated by amniocentesis and
don (UK): CEMACH; 2007. treated by exchange transfusion. Br Med J (Clin Res Ed)
2. Sullivan E, Hall B, King J. Maternal deaths in Australia 1987;294:1322–3.
2003–2005. Canberra (Australia): Australian Institute of 18. Kaneko Y, Ogihara T, Tajima H, Mochimaru F. Continuous
Health and Welfare; 2008. hemodiafiltration for disseminated intravascular coagulation
3. Abenhaim HA, Azoulay L, Kramer MS, Leduc L. Incidence and shock due to amniotic fluid embolism: report of a dramatic
and risk factors of amniotic fluid embolisms: a population- response. Intern Med 2001;40:945–7.

VOL. 115, NO. 5, MAY 2010 Knight et al Amniotic-Fluid Embolism in the UK 917

You might also like