Lessons for care of specific
infectious morbidities
Derek Tuffnell: London 9th December 2014
Alison Rodger: Edinburgh 12th December 2014
Summary of Deaths from
Sepsis
• 83 women died from sepsis between 2009
and 2012
– 20 deaths from genital tract sepsis
– 63 deaths from other infections
• 36 due to influenza
• 27 due to other infections
• 12 of the deaths were late deaths (>42 days)
Background: Risk Factors for Maternal Sepsis
Incidence of severe sepsis
Septic Shock
= 4.7/10,000 (ci 4.2-5.2)
= 0.91/10,000 (ci 0.71-1.15)
Risk Factors
•
•
•
•
•
•
•
Ethnic minority groups (Black & other ethnic minority groups)
Low socio-economic status
Primiparous
Diabetes
Hypertension
Mode of delivery
Febrile illness and antibiotic treatment in the 2 weeks pre-admission
Acosta CD, Kurinczuk JJ, Lucas DN, Tuffnell DJ, et al. (2014) Severe Maternal Sepsis in the UK, 2011–2012: A National Case-Control
Study. PLoS Med 11(7): e1001672. doi:10.1371/journal.pmed.1001672
Absolute risk (95% CI) of all sepsis and severe
sepsis/septic shock as a function of the number of
a priori risk factors.
Acosta CD, Knight M, Lee HC, Kurinczuk JJ, et al. (2013) The Continuum of Maternal Sepsis Severity: Incidence and Risk Factors in
a Population-Based Cohort Study. PLoS ONE 8(7): e67175. doi:10.1371/journal.pone.0067175
Mortality Rates from Genital Tract Sepsis: 1985 - 2011
1.80
1.60
Rate per 100,000 maternities
1.40
1.20
1.00
0.80
0.60
0.40
0.20
0.00
1985-87
1988-90
1991-93
1994-96
1997-99
Triennium
2000-02
2003-05
2006-08
Significant decrease in mortality rate from genital tract sepsis
from 1.13 maternal deaths per 100,000 maternities in 2006-8
to 0.50 per 100,000 in 2010-12
2009-11
Causative organism according
to septic shock diagnosis
50
45
Proportion of women (%)
40
35
30
25
20
15
10
5
0
E.coli
Group A
strep
Group B
strep
Other strep
Septic shock
Staph
Mixed
Other
Unknown
No shock
Acosta CD, Kurinczuk JJ, Lucas DN, Tuffnell DJ, et al. (2014) Severe Maternal Sepsis in the UK, 2011–2012: A
National Case-Control Study. PLoS Med 11(7): e1001672. doi:10.1371/journal.pmed.1001672
No Lab
confirmed
infection
Group A Streptococcus (GAS)
• GAS is most common cause of post
partum maternal death worldwide
• Postpartum women have 20x
increased incidence of GAS disease
compared to non-pregnant women
• Puerperal infections present rapidly
and can be nonspecific, delaying
treatment.
• Primary symptoms include myalgia,
fever, confusion, euphoria,
dizziness, and abdominal pain.
• Once GAS is diagnosed, infection is
often advanced.
• Associated with STSS
Potential routes of GAS infection during
pregnancy and postpartum
Am J Reprod Immunol, 2012 February ; 67(2): 91–100
Genital Tract Sepsis
• A total of 20 women died from genital tract sepsis
• Group A Streptococcus (n=12) associated with
genital tract infections in early pregnancy (n=4)
and peripartum (6 SVD, 2 LSCS)
• Three women with GAS had no clear genital focus
• Coliforms were associated with second trimester
ascending infections (n=6)
Genital Tract Sepsis: Delays in Management
• Delay in identification of the source of infection
• When recognised as genital tract – not fully
investigated or monitored
• Over-reliance on antibiotics to control the infection at
source
• Poor recourse to imaging & repeated imaging – MRI /
CT scan
• Reluctance to take surgical measures - appropriate
drainage of collections or surgical excision of infected
tissue
“A woman presented in labour and was noted to have genital
tract sepsis. She was delivered by caesarean section. The
operation was complicated by a lateral tear and atony with
an estimated blood loss of 1500mls. She was given
intravenous antibiotics for 48 hours after delivery but her
sepsis did not improve. There were several changes of
antibiotics.
On day 5 postpartum a CT scan was performed and found
what was thought to be an area of sepsis in the wound
communicating with the uterine cavity.
On day 6 she worsened and a laparotomy was performed.
The uterus was necrotic and the abdomen contained a great
deal of pus. She had a sub-total hysterectomy and wound
debridement. Despite intensive care she developed acute
respiratory distress syndrome, deteriorated and died”
“A woman admitted in second trimester with vomiting
and preterm pre-labour rupture of the membranes.
Septic - chorioamnionitis caused intrauterine death of the
fetus.
The sepsis resuscitation bundle was promptly applied
and following blood cultures and discussion with a
consultant microbiologist, antibiotics were commenced
within one hour of diagnosis. Despite resuscitation she
failed to improve. The team then proceeded to
hysterotomy to remove the source of the sepsis. After
two days of supportive care on the intensive care unit
she made a full and complete recovery.
Her treatment was prompt and effective with rapid
source control when she failed to respond to
conservative treatments. The time from admission to
control of the sepsis was 18 hours.”
Recommendation
Genital Tract Sepsis
When sepsis is present the source should
actively be sought with appropriate imaging
and consideration given to whether surgical
or radiologically-guided drainage is required
(RCOG Green-top guideline 64b, 2012)
Influenza
• Influenza is a highly infectious acute viral
infection of the respiratory tract caused
mainly by influenza A and influenza B
infection
• In April 2009 a novel strain of influenza A
virus subtype H1N1 (‘swine flu’) was
identified in Mexico and rapidly spread
globally
Pandemic H1N1 in the UK
• First UK cases of H1N1 reported April 2009
• Two waves of pandemic activity seen in the UK
• Complications of A/H1N1 in the general population
appeared similar to seasonal influenza
Incidence of pandemic A/H1N1 cases and confirmed deaths England (source: HPA).
Pandemic H1N1 in the UK
• However as with seasonal flu those
with asthma, diabetes, heart disease or who
were pregnant were at higher risk of
complications
• In the 2009 H1N1 pandemic, there was a 4x
higher rate of hospital admission in pregnant
women compared to the general population
and a 7x higher risk of admission to an ITU
Refs: Lapinsky 2010; McLean and Pebody 2010; Siston, Rasmussen et al. 2010; Pierce, Kurinczuk et al. 2011
Deaths from Influenza
• Between 2009 to 2012, 36 pregnant women died from
influenza (comprising 43% all deaths from infection)
• Of these,
– 27 had confirmed H1N1
– 5 had probable H1N1
– 1 had Influenza A (not confirmed if H1N1 variant)
– 3 had Influenza B
• Median age was 28 years, and the majority (63%) had
been born in the UK.
• 18% had a BMI of > 30 and 28% were current smokers
• Main presenting symptoms were documented fever
(97%), cough (66%) and dyspnoea (38%)
Influenza: Delays in diagnosis and treatment
• Presentation: Route of presentation into health care
services: 33% to Primary care, 36% to Obstetric services,
28% to A+E/UCC
• Diagnostic delay: Influenza was not even considered as a
possible diagnosis in the vast majority (n=34, 94%) of
women at initial presentation with respiratory illness even
at the height of the H1N1 pandemic. This led to delays in
appropriate referral, testing and treatment which likely
affected outcome
• Treatment delay: In 75% of women who died from
influenza, antiviral treatment was not commenced until
after a positive H1N1 result was received, even though this
was thought the most likely underlying aetiology in a
clinically deteriorating patient
“A woman presented to her GP during the second wave
of the H1N1 pandemic at term with symptoms suggestive
of a chest infection and was treated with amoxicillin.
Shortly afterwards, she was admitted in labour and had a
delivery by caesarean section. She was febrile postoperatively with worsening dyspnoea, desaturation and a
productive cough.
Over the following two days her respiratory symptoms
deteriorated so she was started on antibiotics. She was
transferred to ITU where H1N1 testing was finally done,
but anti-viral treatment was started only when a positive
H1N1 result was obtained. She continued to deteriorate
and died”
Influenza Vaccination
• Vaccination is the main Public Health
response to influenza
• Influenza vaccination in pregnancy
– Reduces maternal morbidity and
mortality
– improves fetal outcomes including
reduced likelihood of perinatal death,
prematurity and low birth weight,
– prevents influenza in the infant up to
6 months of age through transfer of
maternal antibodies and potentially
improves long-term adult outcomes
for the infant
Influenza Vaccination
• Vaccination coverage rates
during pregnancy in the UK
are generally low at around
25% if the only indication for
immunisation is pregnancy
(Public Health England 2014).
• This is despite the 2009
pandemic resulting in
increased attention to the
importance of vaccination
during pregnancy.
National H1N1 vaccine uptake
by priority group 2009/10
Influenza Vaccination
• None of the women who died in this review had been
documented to have received seasonal flu vaccine or
H1N1 vaccination in those who presented after 21st
October 2009 (when the UK A/H1N1 vaccination
programme commenced)
• Of the deaths from H1N1, 38% occurred prior to
vaccine availability in the UK and 62% after a vaccine
was available
• Of those who could have been vaccinated against
H1N1, 3 women refused
• It was not clear how many of the others had been
actively offered vaccination through primary care or
obstetric services
“A pregnant woman who smoked refused
vaccination offered by her GP for seasonal
influenza or H1N1. She presented in the
third trimester with flu-like symptoms
including fever, cough and dyspnoea. Rapid
deterioration led to her requiring ventilation
in ITU, but she died within 24 hours of
admission”
Use of neuraminidase inhibitors
• A Cochrane Review (2014) on the efficacy of
neuraminidase inhibitors for influenza found little
evidence of benefit
• However most patients in the treatments studies
included in the review were not at high risk of severe
complications and no pregnant women were included
• Observational data in pregnancy show significant
benefit from early use
• DH/RCOG guidelines recommend that antiviral
treatment should be commenced as early as possible
in pregnant women with signs of flu, particularly within
the first 48 hours of onset of symptoms
• Recommended agent is Relenza (zanamivir) though
Tamiflu if other airway problems or severe H1N1
Refs: Cochrane Database Syst Rev 2014; 4: CD008965, Siston, Rasmussen et al. 2010; Pierce,
Kurinczuk et al. 2011; Muthuri, Venkatesan et al. 2014; Nguyen-Van-Tam, Openshaw et al. 2014
“A pregnant woman presented to A&E and was admitted
overnight with a cough, fever and dyspnoea. All members
of her family had influenza-like illness. She was discharged
home the next day with no relevant investigations
performed and no antibiotic or antiviral treatment initiated.
Her condition deteriorated and when reviewed by her GP
the next day, emergency readmission was arranged.
Testing for H1N1 was conducted which was positive and
antiviral treatment commenced, however the woman
rapidly deteriorated and died 3 days later”
Ventilation and extra-corporeal membrane
oxygenation (ECMO)
• Several women had ECMO but in a number of
cases it does not seem to have been discussed
• In women with severe respiratory failure referral
to a respiratory centre has been shown to be of
benefit
• Protective lung ventilation is beneficial in this
group
– tidal volumes 6ml/kg and plateau pressures
<30cm H2O
Key recommendations
- Influenza
1. Department of Health/RCOG Guideline on the
investigation and management of pregnant women with
seasonal or pandemic flu should be followed (Department of Health
and the Royal College of Obstetricians and Gynaecologists 2009)
2. The benefits of influenza vaccination to pregnant women
should be strongly promoted and pregnant women at any
stage of pregnancy should be offered vaccination against
seasonal and pandemic influenza with inactivated
vaccine. Immunisation against Infectious Disease – “The Green Book” (Public Health England
2014)
3. Early neuraminidase inhibitor treatment should be
instigated for pregnant women with symptoms consistent
with influenza, in line with national UK guidance. (Department of
Health and the Royal College of Obstetricians and Gynaecologists 2009)
Streptococcus pneumoniae
• Streptococcus pneumoniae is the commonest cause of
community acquired pneumonia and can also lead to
invasive bacteraemia with complications such as
meningitis (mortality rate of 20-30%).
• Invasive pneumococcal infection occurs in both
immunocompetent and immunosuppressed individuals,
but remains unclear whether pregnant women per se are
a higher risk group.
• Pneumococcal vaccine recommended in the UK in risk
groups, but not pregnant women with no additional risk
factors
Ref: Bartlett and Mundy 1995; Public Health England 2014
Pneumococcal disease
• Nine women died from pneumococcal
disease
– Five from pneumococcal meningitis
– One from a pneumococcal brain abscess.
– One from pneumococcal pneumonia and
invasive pneumococcal infection
– Two from pneumonia of unclear cause
• None of these deaths were influenza
associated.
“A woman in the second trimester collapsed at work with
headache and loss of balance. The patient spoke to NHS 24
twice and was advised to take paracetamol. The patient also
saw her GP and was given the same advice.
Her symptoms deteriorated over the next 48 hours and she
was taken again to see the GP by her family with a
temperature, severe headaches and poor balance. The GP’s
only advice was to increase analgesia.
Her conscious level deteriorated over the next few hours and
the family called an ambulance. On arrival at A&E she was
intubated and a CT scan demonstrated meningoencephalitis.
She died 24 hours later. S.Pneumonia was cultured from a
right ear swab and brain tissue.”
“A woman had a three month history of left frontal
headache during the second and third trimester. She saw
her GP multiple times about this complaint but she was not
given any antibiotic therapy.
Four weeks before her death she was also reviewed by a
doctor in the antenatal unit who did not refer for further
investigations despite localising features and a prolonged
history of pain.
She presented again three days before her eventual death
when her CNS infection (untreated frontal sinusitis with
subsequent development of osteomyelitis of the frontal
bone and a cerebral abscess ) was finally diagnosed”
Bacterial Meningitis
• Limited published data on bacterial meningitis in pregnancy.
• In one review 42 women with bacterial meningitis during
pregnancy: 60% due to Streptococcus pneumoniae (n=25),
17% Listeria monocytogenes (n=7)1
• A Dutch review of 15 deaths from meningitis during
pregnancy: 10 due to S. pneumoniae with ENT site of origin
in 7. Limited evidence of higher incidence pneumococcal
meningitis in pregnancy compared to background population
• In 2006-2008 review, three women died of central nervous
system infection of whom two had pneumococcal
meningitis3
Refs: 1. Adriani et al. 2012. 2 Schaap et al. 2012. 3. Lewis et al. 2011
Bacterial Meningitis
• In this review 7 women died of bacterial meningitis
– Five from pneumococcal meningitis
– One from pneumococcal brain abscess
– Two from other streptococcal meningitis
• Two women had direct extension of S. pneumoniae from
untreated infection of the middle ear and untreated
frontal sinusitis
• Multiple attendances at primary care with symptoms
suggestive of early CNS infection including severe
headache
Key recommendation
Pneumococcus and Meningitis
Repeated presentation to the general
practitioner or community midwife or
alternatively repeated self-referral to the
obstetric triage or day assessment unit
should be considered a ‘red flag’ and
warrant a thorough assessment of the
woman to investigate for signs of sepsis.
Other deaths
Eighteen deaths due to other causes
– 3 fulminating Group A Streptococcus infections in the
third trimester, with no obvious evidence of a genital
tract focus of infection.
– 3 women from overwhelming sepsis (two of uncertain
origin)
– 2 from appendicitis.
– A urinary tract infection, a complex sinus infection, a
quinsy, fulminant herpes simplex (2 deaths), a breast
abscess, miliary tuberculosis and a pelvic abscess.
Women who survived
• Enquiry considered 34 women with sepsis
who survived from UKOSS study
– 24 genital tract sepsis
• 6 Group A Streptococcus infection.
• 3 women had septic shock in association with
genital tract sepsis following second trimester
membrane rupture and one following a second
trimester cervical rescue suture.
– 10 other causes,
• 2 non genital Group A Streptococcus
Comparison between women who
died and women who survived
Classification of
care received
Percentage of
women who died
Percentage of
women who
survived
Good care
23%
26%
Improvements to care
which would have
made no difference to
outcome
Improvements to care
which may have made
a difference to
outcome
14%
53%
63%
21%
Comparison between women who
died and women who survived
Delay in
review
Percentage of
women who
died
70%
Percentage of
women who
survived
71%
Outside hospital
46%
33%
Inside hospital
83%
75%
Any location
Comparison between women who
died and women who survived
Review
undertaken
Internal review
Root cause
analysis
Percentage of
women who
died
53%
Percentage of
women who
survived
35%
39%
33%
Morbidity v Mortality
• Still many cases where standard of care
could be improved
• Delays are in both primary and secondary
care
• Review and analysis of deaths and severe
morbidity cases should be increased
Conclusions
• Sepsis is a leading cause of death –
almost a quarter of deaths
–THINK SEPSIS
• Women with sepsis require
–Early diagnosis
–Rapid antibiotics
–Review by senior doctors and
midwives
Conclusions
• Genital tract sepsis deaths have reduced
• 1 in 11 of all the maternal deaths are from
influenza
– Vaccinate
• Repeated presentation is a ‘red flag’
• Early involvement of infectious diseases or
microbiology teams
• Adequate provision of critical care