  |
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Journal of the New Zealand Medical Association,
03-July-2009, Vol 122 No 1298
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Alistair Escott, Dawn E Elder, Jane M Zuccollo
Sudden infant death syndrome (SIDS) is defined as “The
sudden death of an infant under 1 year of age which remains unexplained after a
thorough case investigation, including performance of a complete autopsy,
examination of the death scene and review of the clinical
history”.1
Over recent decades SIDS rates have fallen in New Zealand
and internationally.2–6 For many infants
who die suddenly and unexpectedly and do not fulfil diagnostic criteria for
SIDS, a final diagnosis may not be possible because of failure to undertake a
complete examination of the death scene and lack of documentation of the full
clinical history. This has lead to the more general term Sudden Unexpected Death
in Infancy (SUDI) being used to describe this group of infants and to the final
cause of death increasingly being labelled as
undetermined.7
Attention has increasingly focused on the infant sleeping
environment. Asphyxia, overlaying, and strangulation have been noted as causes
of SUDI due to unsafe sleeping
environments.8–10 Reviews of coronial
SUDI death investigations have been reported internationally but not previously
in the New Zealand context.11–13 This
study aimed to review cases of SUDI, referred by the Coroner to the Wellington
Hospital Mortuary for autopsy from 1997 to 2006, to describe their demographic
characteristics and assess the sleep situation at time of death.
MethodsCases were sourced from the Wellington Hospital
Mortuary records. Infants who died suddenly and unexpectedly who were more than
28 days of age and less than 1 year old were included. Documents used to source
information about the cases were; the autopsy report, the Police 47 form (P47),
and the Coroner’s Inquest findings.
Details for hospital-based records lacking a copy of
the Coroner’s findings or an autopsy report were accessed through the
Ministry of Justice Archives. The predetermined hierarchy of evidence was the
Coroner’s Inquest, followed by the autopsy report with the least weight
placed on the P47. Some files contained copies of the complete Police file and a
clinical history taken by a paediatrician. This information was given more
weight than the P47.
The verdict from the Coroner’s findings as to
cause of death was also recorded. All subjects were included in the initial
database and cases were subsequently excluded if a specific cause of death was
determined at autopsy. Information gathered for the remaining infants included
the following; the position the infant was found in, the position the infant was
placed to sleep, the sleep surface the infant was found on, the usual sleep
surface, whether the infant shared a room, details of bedsharing, and whether
the infant was placed on a soft surface (defined as either a pillow, sheepskin
or duvet).
Hospital records were reviewed where possible. Data
collected included: age, sex, gestation, birth weight, and ethnicity.
Socioeconomic status was assessed using the home address as reported on the P47
form. Addresses were geo-coded by Statistics New Zealand into New Zealand
Deprivation Index 2001 mesh-blocks and then converted to 2001 decile
scores.14
Data were entered in an excel spreadsheet and pivot
tables used to facilitate analysis. Data for sex, birth weight, ethnicity and
gestation were compared with regional data from the Ministry of Health Report on
Maternity for 2002 and Chi-squared analysis was used to compare variables where
comparable data were available.15
Binary logistic regression was applied to determine
whether any of the variables increased the risk of bedsharing. Incomplete data
prevented a correlation between decile and bedsharing being calculated. A p
value < 0.5 was considered significant. Ethical approval for the study was
granted by the Central Regional Ethics Committee. Approval for the study was
given by the Wellington Coroner.
ResultsCase selection—There were 99
Coroner’s case records that met inclusion criteria. The most complete
information was available for infants referred to the Wellington Coroner so only
these were included in the study. Other cases had been referred to Dr Zuccollo
for forensic postmortem from Coroner’s outside of the Wellington region.
Final causes of death are listed in Table 1.
There were 64 postneonatal deaths between 1997 and 2006.
Some cases were labelled as SUDI as a final diagnosis. To ensure all possible
SUDI deaths were included in the analysis the following classifications were
also included in the SUDI group: accidental asphyxia, SIDS and undetermined with
possible and probable classifications of these diagnostic labels also included.
There were 54 (84.4%) SUDI cases, and 10 (15.6%) non-SUDI cases.
Table 1. Final cause of death for cases
referred to the Wellington Coroner 1997–2006
SUDI infants—The demographic details
of the SUDI infants are listed in Table 2. Two sets of twins were found dead
together. For SUDI infants, 75.0% (95%CI: 61.1–85.2) were full term
compared with 92.6% (95%CI: 91.8–93.2) from the local population as
reported in the 2002 National Maternity Report (Chi squared 20.91,
p<0.0001).15
The median birth weight was 2880g compared with the national
mean birth weight of 3400g documented in the 2002 National Maternity Report.
Māori (Chi squared 40.1, p<0.001) and Pacific (Chi squared 6.5, p=0.01)
infants were over-represented in the sample compared with ethnicity reported for
Wellington infants in the 2002 Maternity Report. New Zealand European infants
were under-represented (chi square 30.3, p<0.001)
Mothers of SUDI infants—Maternal age
was known for 49 cases. The median maternal age was 24 years (range 17–40
years). Body mass index (BMI) values were only available for 18 mothers so could
not be reported on. NZ Deprivation Index decile ratings are illustrated for
mother and infant pairs in Figure 1. For 30 (55.5%) infants, the decile rating
was 9 or 10 whereas only 6 (11.1%) infants were in decile 1 or 2.
Timing of deaths—The number of deaths
varied from one to nine per year with no trend to an increase or decrease in
deaths over the time period. There was a slight majority (59.3%) of deaths in
winter and autumn. Deaths occurred on the weekend (Friday, Saturday, Sunday) in
46.2% of cases. Sunday was the commonest day for death to occur (18.5% cases).
The majority of infants (57.4%) were found dead between 0601 and 1200h. There
were 11 (20.4%) infants found between 0001–0600h, 9 (16.7%) infants
between 1201–1800h, and 3 (5.6%) infants between 1801–2400
hours.
Table 2. Demographic variables for 54 SUDI
cases referred to the Wellington Coroner 1997-2006 and for some variables for
infants born in the Hutt Valley and Capital and Coast DHBs in
2002
Figure 1. NZ Deprivation index 2001
classification based on maternal address for SUDI infants referred to the
Wellington Coroner 1997–2006
 Sleep environment at death—Overall,
50% of infants for whom the information was known were placed to sleep in a
non-recommended sleep position and 38% usually slept in a non-recommended
location. Table 3 lists the position the infants were placed for the last sleep,
the position found and the usual sleep position.
Table 3. Usual sleep position, position placed
at last sleep and position found, for SUDI deaths referred to the Wellington
Coroner 1997–-2006
Table 4 lists the usual location for sleep and location
found. Although only 12 of the cases usually slept in a double bed, 24 cases
were found dead in a double bed. There were 29 (53.7%) babies bedsharing at the
time of death. For one infant no information was available about bedsharing. Of
the infants in the bedsharing group 16 (55%) were <3 months of age compared
with 9 (37.5%) of the non-bedsharing infants. Of those placed supine, 17 were
found supine and 13 of those were bedsharing.
Table 4. Normal sleep location and location
found for SUDI deaths referred to the Wellington Coroner
1997–2006
Of the 17 infants placed on the side, 8 remained on the side
but 6 of these were bedsharing. Four infants moved from side to prone, four
infants moved from side to supine (three were bedsharing) and one infant was
found entrapped. Two of the bedsharing infants were twins sharing a cot. Of the
others, 23 (79.3%) were sharing a double bed, two (6.9%) a couch and two (6.9%)
a single bed. For 12 of the infants bedsharing there was one other person in the
bed, for another 12 there were two other persons in the bed, for four there were
three others and for one there were four others.
Of the 10 infants found dead on a Sunday, 9 (90%) were
bedsharing. The time found was between midnight Saturday and lunchtime on Sunday
for eight of these infants. Breast-feeding was the main reason for 14 of the
infants sharing a bed during the last sleep (48.3%). Of those 14 infants, 13
(92.8%) of their mothers fell asleep before the infant was removed from the
breast. Insufficient information was available to comment on specific bedding
characteristics.
Risk factors for death in a bedsharing
situation—Infants found dead on Sunday were more likely to be
bedsharing than infants who died on other week days (OR 15.0, 95%CI:
1.2–185.2, p=0.04). Diagnosis at postmortem, age, sex, time of day,
ethnicity, position put down, weekend death or death in the second half of the
study were not significant risk factors for death in a bedsharing situation.
Change in diagnoses—There were 27
deaths in the first half of the decade and 27 in the second. Bedsharing was a
factor in 12 (44.4%) of the deaths in the first 5 years and in 17 (63%) of the
deaths in the second. This difference was not significant (Chi squared 1.51,
p=0.22).
For 44 (80.5%) Wellington SUDI cases the Coroner had
released a verdict as to cause of death. Of these verdicts 36 (81.8%) were in
agreement with the pathologist’s findings at postmortem. Of the eight
cases that were changed, seven received the verdict “Undetermined”
by the Coroner. A Coroner’s verdict of SIDS or possible SIDS was given for
12 (44.4%) cases in the first half of the decade but for no cases in the second
half of the decade.
The most common final Coroner’s verdict in the second
half of the decade was SUDI. Information presented at inquest lead to a
confirmed diagnosis in only one of these cases which was to confirm a case of
presumed accidental asphyxia.
Accidental asphyxia—To identify the
reasons for accidental asphyxia as a cause for SUDI, all cases of suspected or
presumed accidental asphyxia, as determined after autopsy, were grouped
together. This resulted in 12 cases of accidental asphyxia when possible and
probable cases were included (Table 5).
Table 5. Suspected mechanism for SUDI deaths
concluded to be due to presumed or possible accidental asphyxia
DiscussionThis study provides a profile of factors associated with
post-neonatal SUDI deaths referred to the Wellington-based coronial paediatric
pathology service. In this cohort of infants, 88.7% were less than 6 months old
and 24.9% were preterm.
Māori and Pacific infants were over-represented in the
cohort and just over half the infants were from a decile 9 or 10 area. Half had
been placed to sleep in a non-recommended sleep position and just over half were
found dead in a bedsharing situation. For 38% the usual place of sleep was a
non-recommended sleep location.
There was a strong association between being found dead on a
Sunday morning and bedsharing at the time of death. While the rate of SIDS has
decreased, worldwide, deaths listed as accidental asphyxia or undetermined have
increased.16 This trend is reflected in the
current study, which showed a significant decrease in the proportion of SIDS
cases and an increase in cases labelled SUDI between the first and second
five-year periods.
The prevalence of bedsharing among the SUDI cases in this
study at 53.7% is comparable to an international prevalence of between 35-50% in
cohorts based on similar criteria.3 17 18 In
this study 90% of infants who were bedsharing for the purpose of breast-feeding
were not removed from the breast before the mother fell asleep. There was a
significant association between bedsharing and being found dead on Sunday
morning.
Increased risk of death at weekends and in particular on
Sunday, was found in the New Zealand Cot Death Study and in the United Kingdom
the weekend effect was more marked in younger
infants.19 20 It was not possible to determine
from the current data why this might be so. Mothers in a bedsharing situation
appear to usually respond well to an infant’s needs during the
night.21
It is possible that the weekend, and in particular Saturday
night, might be a time when parents are more likely to socialise and the
combination of staying up later than usual and perhaps consumption of alcohol,
even in small amounts, may be enough to affect a mother’s ability to
respond to her infant’s needs while co-sleeping. The increased tiredness
may also mean that an infant-mother pair is more likely to fall asleep together
after a feed.
Māori and Pacific infants were over-represented in this
group of SUDI deaths. For Caucasian infants only 23.5% of infants were in a
double bed at the time of death but the corresponding proportion for Māori
and Pacific infants was 50% and 66.7% respectively. Of the infants found dead in
a double bed, 66.7% lived in a decile 9 or 10 area. Deprivation is a recognised
association with sudden infant death in other countries
also.22
We were unable to get accurate data on levels of maternal
smoking. This is recognised as being the main associated risk for infants in a
bedsharing situation.23,24 More recent
case-control studies have shown that infants of mothers who do not smoke are
also at increased risk when bedsharing in the first 3 months of
life.18,24,25 Half the bedsharing infants in
this study were <3 months of age.
The rate of bedsharing in this cohort (53.7%) compares with
a rate of 31.7% reported previously in 1993 from the New Zealand Cot Death study
case data and 11.7% from the New Zealand Cot Death study
controls.23 The policy conclusions drawn from
these data published in 1995 suggested that stopping bedsharing for all infants
would have a minimal effect on the rate of sudden infant
death.26 However this analysis did not take
into account age at death or maternal weight which have both been shown since to
contribute to risk.17
It is concerning so many infants were placed to sleep in a
non-recommended sleep position or location. Recent reviews of infant care
practices in New Zealand indicate that prone placement of infants for sleep is
now rare. 27, 28 Of the 17 infants placed to
sleep on their side, only 8 were found dead on their side. Of the other 9, 1 was
found entrapped, 4 were found prone, and 4 supine. This is not a recommended
sleep position because of this recognised instability.
The demographic profile of the cases in this study suggests
that some Māori and Pacific parents and parents from more deprived groups
of the community may be less likely to either be aware of or adhere to
guidelines for safe sleep practice for infants.
This study had three significant limiting factors. Firstly a
lack of information in case files and medical records meant that known
associations such as maternal smoking, gravida, parity and maternal weight could
not be analysed. Secondly the absence of controls meant that relative risks for
each risk factor could not be calculated. Thirdly, the change in use of
diagnostic categories over time had to be accounted for. This was addressed by
combining all unexplained deaths together and including those determined
originally as SIDS and accidental asphyxia in with the SUDI deaths.
A major strength of the study is that all the cases were
autopsied by one perinatal pathologist. Also the cases derived from the
Wellington Coroner’s jurisdiction contained all cases referred to the
Coroner in the region for the ten year time period and consequently form a
complete dataset for the Wellington population.
The overall rate of sudden infant death has not varied
between 1997 to 2006 but there has been a significant shift in diagnostic
categories from SIDS to SUDI. This trend probably reflects that more information
is available to the pathologist to enable cause of death to be determined. An
increasing number of deaths appear to be occurring while bedsharing.
Māori and infants from lower socioeconomic deciles
appear to be at greater risk from sudden unexpected death in infancy in the
Wellington region. Educational messages about safe sleeping practice need to be
particularly targeted towards these groups if further improvements are to be
seen in decreasing unexpected death in this age group.
Competing interests: None known.
Author information: Alistair Escott,
Medical student, School of Medicine and Health Sciences; Dawn E Elder, Senior
Lecturer, Dept of Paediatrics & Child Health, School of Medicine and Health
Sciences; Jane M Zuccollo, Senior Lecturer, Dept of Obstetrics &
Gynaecology, School of Medicine and Health Sciences;
University of Otago, Wellington
Acknowledgements: Alistair Escott was
supported by a summer studentship from the Wellington Medical Research
Foundation. We also thank the Wellington Coroner (Mr Garry Evans) and the staff
at the Coroner’s office for supporting the study and providing access to
the Coroner’s verdict for each case as well as James Stanley for
assistance with statistical methods.
Correspondence: Dr Dawn Elder, Dept of
Paediatrics, SMHS, University of Otago, Wellington, PO Box 7343, Wellington, New
Zealand. Fax: +64 (0)4 3855898; email: dawn.elder@otago.ac.nz
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||
| | Current
issue | Search journal |
Archived issues | Classifieds
| Hotline (free ads) Subscribe | Contribute | Advertise | Contact Us | Copyright | Other Journals |
|