Key words
miscarriage - ectopic pregnancy - HELLP syndrome - pregnancy
Introduction
One definition of pregnancy-associated death is death from any cause during pregnancy
and up to one year post partum [10], [20], [23]. However, the WHO defines pregnancy-related death as the death of a woman which
occurs between nidation of the fertilized ovum and 42 days after termination of pregnancy
[27], [28]. But neither of the definitions require the cause of death to be anatomical changes
immediately related to pregnancy. This means that in some cases the pregnancy may
be the trigger for the fatal event (e.g., due to hormonal changes and/or increased
circulatory volume load) but death may also be caused by events which could not occur
outside of pregnancy (e.g., amniotic fluid embolism).
Immediate causes of sudden death during pregnancy and pregnancy-associated causes
of death are usually disease, intoxication, accidents [11], [43] or, rarely, suicide [37]. In addition to pulmonary thromboembolism, amniotic fluid embolism, air embolism
[3], [4], [26], [30], [31] and ruptured aneurysm [35], [36], other possible causes are:
-
acute and chronic undetected cardiovascular disease (acute myocardial infarction,
myocarditis, postpartum cardiomyopathy, cardiac valve disease, endocarditis, arrhythmias
including Long QT syndrome, etc.)
-
acute ruptured tubal or ectopic pregnancy with internal bleeding [1], [6], [45]
-
fatal intoxication, primarily for drug-addicted pregnant women [18]
-
postpartum bleeding (e.g. postpartum atonic bleeding which was not detected or only
detected too late, possibly because of inadequate monitoring) [34]
-
undetected uterine rupture with hemorrhage [13], [19]
-
fatal outcome in HELLP syndrome [48], [49]
-
Sheehanʼs syndrome [39], [41], [42]
If no macroscopic cause of death can be identified, then histopathological examination
can help identify the cause of death. This can be useful in providing evidence for
or excluding acute or chronic myocarditis or establishing amniotic fluid embolism
as the cause of death. If death was due to massive hemorrhage following uterine rupture,
the cause of death can be identified, for example by detecting granulation tissue
and scar tissue at the site of rupture in women with previous C-section. Other reported
causes of death include postpartum coronary artery thrombosis after the administration
of bromocriptine [29]. This wide range of causes of pregnancy-associated deaths has led to pregnancy-related
causes of death being defined more narrowly as death from causes which cannot occur
in non-pregnant women. In addition to deaths which clearly had natural causes, some
deaths which occur in pregnancy, especially in countries with high-quality medical
care, almost always raise the question whether medical malpractice could be the cause
of death, particularly when death involved perinatal or postpartum hemorrhagic complications.
All of the above shows that the cause of any death which occurs in connection with
pregnancy must be carefully differentiated.
Material and Methods
The autopsies listed in the Registry of Autopsies of the Institute of Forensic Medicine
of Justus-Liebig University Gießen between 1992 and 2016 (i.e. which occurred over
a period of 25 years) were analyzed. A total of 22 autopsies were identified where
it was assumed that death occurred in connection with pregnancy in accordance with
the above-mentioned definition of the WHO. One death was not included in our study,
as it was a homicide in early pregnancy and the pregnancy was merely the reason for
the killing. One mortality from amniotic fluid embolism, a very rare cause of death,
was processed by the Institute for Forensic Medicine of the University of Münster
and added to the Gießen cohort because of the rarity of the findings, meaning that
the final evaluation consisted of a total of 22 pregnancy-associated deaths. The extent
of data depended on the recorded information, with some cases recorded in their entirety
and while the records of other cases were incomplete. However, data on the cause of
death were available for evaluation in all cases. In addition to age and the time
of death during or after the pregnancy, the diagnosed medical cause of death was recorded
in all cases and the manner of death was classified as natural, unnatural or unexplained
in all cases. If an accusation of medical malpractice was raised, the available expert
opinions were reviewed.
Results
All 22 autopsies were carried out by order of the respective department of public
prosecution, sometimes following substantiated accusations of malpractice brought
by dependents of the deceased or by the department of public prosecution. The average
age of the women who died from pregnancy-associated causes was 31.8 years (22 – 45
years).
Cause of death was differentiated into natural and unnatural causes. In the majority
of cases, death occurred suddenly from natural causes. Of the 22 deaths, 9 deaths
were classified as pregnancy-associated deaths because of the time when death occurred
but were considered to be due to natural causes occurring independently of the pregnancy
status ([Table 1]).
Table 1 Pregnancy-associated deaths from natural causes (n = 9): age of the pregnant women,
duration of pregnancy and cause of death.
|
Age
|
Duration of pregnancy
|
Cause of death
|
|
GW = week of gestation
|
|
32
|
6th month of pregnancy
|
Fulminant pulmonary embolism in a woman with intact twin pregnancy
|
|
23
|
7th month of pregnancy
|
Cerebral hemorrhage affecting the right-sided cerebellum in a woman with hypertension
|
|
23
|
GW 40
|
Fulminant pulmonary embolism 27 days post partum
|
|
34
|
7th month of pregnancy
|
Cerebral hemorrhage with ventricular rupture in a woman with hypertension
|
|
28
|
GW 23
|
Aortic rupture, pericardial tamponade in a woman with intact twin pregnancy
|
|
25
|
GW 38
|
Myocarditis 2 days after C-section in a woman with a history of drug abuse
|
|
23
|
8th month of pregnancy
|
Fiedlerʼs myocarditis
|
|
39
|
6th month of pregnancy
|
Asthma attack in a woman known to have bronchial asthma
|
|
22
|
GW 15
|
Acute H1N1 infection with fulminant course
|
The cause of death in 3 other cases was classified as pregnancy-induced fatal disease,
i.e., the women presented with syndromes which cannot occur without being pregnant
([Table 2]).
Table 2 Pregnancy-induced natural causes of death (n = 3): age of the pregnant women, duration
of pregnancy and cause of death.
|
Age
|
Duration of pregnancy
|
Cause of death
|
|
GW = week of gestation
|
|
29
|
GW 9
|
Ruptured tubal pregnancy; on autopsy approx. 2300 ml blood found in the abdominal
cavity ([Fig. 1])
|
|
33
|
GW 40
|
Fatal puerperal sepsis 24 days post partum; on the day she died, the patient complained
at home of experiencing a “shivering fit” but did not consult a doctor
|
|
39
|
GW 41
|
Fatal amniotic fluid embolism requiring reanimation during C-section ([Fig. 2])
|
Fig. 1 Necrotic abortive placental villi and circular hemorrhagic lesions in the center
part of the isthmus of the fallopian tube in the vicinity of the ruptured ectopic
pregnancy (H & E × 40).
Fig. 2 Fatal amniotic fluid embolism with numerous non-nucleated keratin lamellae in the
peripheral branches of the pulmonary artery and septal capillaries of the pulmonary
tissue (H & E × 400).
Pregnancy-related complications or fatal hemorrhage were identified in 7 cases as
the cause of death ([Table 3]). Accusations of medical malpractice, i.e., that the bleeding was detected too late
or was not treated in accordance with accepted standards of care, were usually levelled
in cases of fatal hemorrhage.
Table 3 Fatal hemorrhages induced directly or indirectly by the pregnancy (n = 7): age of
the pregnant women, duration of pregnancy and cause of hemorrhage.
|
Age
|
Duration of pregnancy
|
Cause of hemorrhage
|
|
GW = week of gestation
|
|
26
|
GW 39
|
Para 2, gravida 3; bleeding from initially undetected cervical tear with secondary
puerperal sepsis (Group A streptococci) during vaginal delivery with vacuum extraction
after discontinued homebirth
|
|
35
|
GW 40
|
Para 1, gravida 2; postpartum hemorrhage from uterine atony; multiple organ failure
(MOF) from hemorrhagic shock despite hysterectomy
|
|
40
|
GW 40
|
Bleeding from high vaginal tear and hemorrhagic shock from clinically diagnosed uterine
atony
|
|
33
|
Early pregnancy
|
Iatrogenic injury of the right iliac artery during laparoscopy after curettage performed
to exclude ectopic pregnancy; cause of death was hemorrhagic shock
|
|
31
|
GW 39
|
Clinically diagnosed uterine atony with postpartum hemorrhage and death from hemorrhagic
shock (exhumation)
|
|
36
|
GW 34
|
Hemorrhagic shock in a woman with placenta previa, bicornuate uterus and twin pregnancy
|
|
38
|
9th month of pregnancy
|
Death occurred five days post partum, following initially undetected high vaginal
tear following vaginal delivery; cause of death was hemorrhagic shock
|
Deaths of autopsied women were classified either as pregnancy-associated natural deaths
from causes which can also affect non-pregnant women ([Table 1]; n = 9) or pregnancy-induced deaths from natural causes ([Table 2]; n = 3), with fatal hemorrhage induced directly or indirectly by pregnancy considered
separately ([Table 3]; n = 7). In addition to the above-mentioned causes of death, there were 3 further
deaths which were also considered separately; in 2 cases the cause of death could
not be determined (1 × sudden death 1 day after curettage in the 8th week of gestation;
1 × death 4 weeks after miscarriage), in one case, the cause of death was identified
as medical error (administration of the wrong medicine) after the patient had received
a very high dose of ropivacaine during delivery with epidural anesthesia (EDA) ([Table 4]).
Table 4 Unusual pregnancy-associated deaths (n = 3): age of the pregnant women, duration
of pregnancy and cause of death.
|
Age
|
Duration of pregnancy
|
Cause of death
|
|
EDA = epidural anesthesia; GW = week of gestation; s/p = status post
|
|
45
|
GW 8
|
s/p curettage carried out one day before death; cause of death unclear
|
|
35
|
s/p miscarriage in early pregnancy
|
Found lifeless approx. 3 – 4 weeks after miscarriage, cause of death unclear; toxicological
findings were positive for paracetamol + propyphenazone, but not in toxic doses
|
|
31
|
GW 39
|
Seizures after administration of medication while receiving EDA, then asystole; C-section
performed with reanimation; administration of wrong dose of NAROPIN® (= ropivacaine) in clearly toxic amounts, ampoules of ropivacaine 0.75% were administered
inadvertently instead of ampoules of ropivacaine 0.25%
|
Medical malpractice was found to be an important cause of death, particularly in cases
with fatal complications of bleeding where postpartum monitoring was inadequate. The
specialist obstetricians and gynecologists were of the opinion that there was a causal
relationship between 3 of the 22 deaths discussed in this study and medical malpractice.
These cases were: 1 × an excessively high dose of ropivacaine, 2 × inadequate monitoring
of the pregnant woman which resulted in fatal bleeding being detected too late.
Discussion
Pregnancy-associated deaths are extremely rare in countries with high-quality medical
care. This can mean that in some cases the threshold for reviewing the cause of death
may be very low and an initial suspicion of medical malpractice is not dismissed ad
hoc and never dismissed without first carrying out an autopsy.
In addition to a number of case histories, there are also studies with larger numbers
of maternal deaths [9], [11], [43] and reviews on the processing of, and investigation into, these deaths together
with their respective histological findings [14], [20]. Investigations are carried out in both criminal and civil proceedings [10], [12], [27], [28], [38].
Pregnancy-related hormonal effects may facilitate death from natural causes (e.g.
fulminant pulmonary embolism, myocarditis, hypertensive intracerebral hemorrhage,
aortic rupture, etc.) without it being ultimately possible, in individual cases, to
prove this causal relationship beyond further doubt. Such deaths from natural causes
were the most commonly identified deaths. Information on these causes of death is
found in the literature, particularly cases of arterial rupture or dissecting arterial
aneurysm with fatal hemorrhage during pregnancy or very shortly after delivery [5], [35], [44], [46], [50] and cases of fulminant pulmonary embolism [33]. Causes of death reported in the literature include amniotic fluid embolism [7], [13], [21], [25], [26], [30], [31], [32] and – although this is rare – pregnancy-associated air embolism [4]. In some cases, rupture of an ectopic pregnancy, particularly of a tubal pregnancy,
can result in fatal hemorrhage [1], [6], [45]. Other causes of death include pregnancy-related complications of bleeding [8], [17], [34], deaths from HELLP syndrome or (pre-)eclampsia [3], [48], [49], and fatal puerperal sepsis with Group A streptococci infection [24].
Other causes of death reported in case histories include spontaneous rupture of an
unscarred gravid uterus [19], fatal methadone intoxication of a pregnant woman [18], pregnancy outcomes after traffic accidents [22], spontaneous postpartum liver rupture [47], and acute postpartum coronary artery thrombosis in a woman receiving bromocriptine
[29]. Other rare causes of deaths during pregnancy are pregnancy-associated disease,
particularly Sheehanʼs syndrome [39], [40], [42], [43]. Based on the findings of our study, deaths can be differentiated into four groups:
-
Group I: Deaths from natural causes which could, in principle, also have occurred without
the woman being pregnant (e.g., fulminant pulmonary embolism, myocarditis, arterial
rupture, etc.).
-
Group II: Pregnancy-related deaths which were unavoidable but were due to natural causes, such
as amniotic fluid embolism, uncontrollable puerperal sepsis despite state-of-the-art
care, or late detection of a ruptured tubal pregnancy with fatal hemorrhage.
-
Group III: Deaths which occurred in the context of fatal pregnancy-related bleeding, such as
atonic postpartum bleeding and bleeding from (initially undetected) delivery-related
rupture or injury. This also includes deaths from what must be considered medical
malpractice (inadequate monitoring).
-
Group IV: Deaths which coincided with pregnancy but where the cause of death was not clear.
This includes death from myocardial disease not diagnosed prior to the time of death
(e.g. ionic channel defects or cardiomyopathies where the definitive diagnosis can
be extremely time-consuming and difficult).
Of course, every individual death must be carefully reviewed to determine whether
treatment was in accordance with accepted standards of medical care. If medical malpractice
did occur, the question of whether the cause of death was malpractice must be investigated
separately.
Deaths which coincide with pregnancy should generally be investigated by autopsy,
although this is currently not a legal requirement. The law on the indications required
for terminating pregnancy in the former GDR stipulated that all deaths occurring in
connection with a pregnancy required postmortem examination. Autopsies were mandatory
for “deceased pregnant women and women who died in childbirth as well as women who
died within 6 weeks of giving birth” (Section 8 subsection 1b of the “Decree Concerning
Medical Postmortem Examinations”) [2]. Other countries have similar legal requirements [16].
Autopsy studies in forensic medicine have shown that in the majority of cases, questions
about the cause of death are often already answered by the macroscopic findings at
autopsy, at the latest after additional histological examination. Moreover, when allegations
of medical malpractice were raised, autopsies had an exculpating effect in the majority
of cases, as the autopsy was able to show that death was from natural causes.
Perinatal and postnatal fatal bleeding are adverse events specific to pregnant and
postpartum women, as is puerperal sepsis. These causes of death habitually raise the
question whether monitoring on the labor ward and post partum was adequate. It is
important to note that complications can always occur; any allegation of malpractice
must focus on the question whether the bleeding complication or the puerperal sepsis
could have been detected and controlled in good time.
Forensic medicine is useful for establishing the status quo at the time of death and
for interpreting the autopsy findings but it reaches its limits when events which
occurred over a period of time need to be evaluated and it is necessary to determine
whether these events could have been diagnosed and treated. The latter two questions
require the expert opinion of obstetricians and gynecologists based on the autopsy
findings. This approach (i.e., a forensic autopsy report including histological and
toxicological examinations and the opinion of an obstetrician or gynecologist on whether
medical malpractice was involved) which takes the expertise of the respective specialist
into account [15] is standard practice and is accepted in both criminal and civil proceedings.
Particularly in the two cases of fatal complications of bleeding reported in this
study, if criminal proceedings are launched, it is the specialistʼs job to state whether
the patient would have almost certainly survived if she had been monitored properly
and the bleeding had been detected in time. If this is not possible, then in criminal
proceedings the principle of “in dubio pro reo” must apply. But irrespective of any
criminal proceedings, it is nevertheless possible to bring a separate civil action
for compensation, as this requires a lower standard of proof.
Conclusion
Pregnancy-induced and pregnancy-associated deaths are very rare in Germany and usually
require an autopsy to determine the cause of death and, in some cases, additional
histological, microbiological, virological and toxicological examinations. The majority
of deaths are from natural causes but, particularly if there were complications of
bleeding which raised the suspicion of medical malpractice, an expert opinion to identify
both the cause and the manner of death is required. The expert opinion must take account
of the forensic autopsy report and assess the clinical processes in its obstetrical
and gynecological evaluation. The most important aspect is to determine whether the
bleeding could have been detected earlier and whether the measures taken to control
the bleeding complications complied with the requisite medical standards of care;
in the context of criminal proceedings this should serve to answer the question whether
the patient would have almost certainly survived if the bleeding had been detected
in good time. Irrespective of any criminal proceedings, it is nevertheless possible
to bring a civil action and claim for damages, as the standard of proof with regard
to determining causality between malpractice and death is lower in civil actions.
