Article Text
Abstract
Background Accidental suffocation during sleep, leading to death, has been described as due to overlay or wedging of infants, particularly in a bed-sharing situation. Bed sharing is a risk factor for sudden infant death syndrome but the mechanism of death is not clearly defined. Accidental suffocation may be one such mechanism.
Objective To describe accidental suffocation deaths during sleep in New Zealand between 2002 and 2009.
Design The New Zealand mortality database, which holds data collected by the Child Youth Mortality Review Committee and the Perinatal and Maternal Mortality Review Committee, was searched for potential deaths by accidental suffocation in infants less than 1 year of age. Deaths underwent a detailed analysis by demographic data and qualitative report.
Results There were 48 deaths due to accidental suffocation between 2002 and 2009 in New Zealand, equating to a rate of 0.10 deaths per 1000 live births. The most common age at death was 1 month or under (n=11, 23%). Deaths were due to overlay (n=30, 63%) or wedging (n=18, 37%) and two-thirds (n=34, 71%) were in a bed-sharing situation. A quarter of deaths (n=12, 25%) occurred in makeshift bedding arrangements, some of which were away from home.
Conclusions Accidental suffocation in bed was responsible for 48 preventable deaths. Prevention of these accidental deaths needs to focus on supporting changes in family behaviour with safety messages that are consistent, persistent and disseminated widely.
Statistics from Altmetric.com
What is already known on this topic?
Bed sharing increases the risk of death in infants. Those under 1 month of age are at the greatest risk of death while bed sharing.
Suffocation in place of sleep can occur by overlay from an adult/another child, or wedging between surfaces, like a mattress and the wall.
Unsafe sleep surfaces include cushions, pillows, sofas and soft bedding that can lead to wedging, entrapment or airway compromise.
What this study adds?
Accidental suffocation contributes to infant deaths in a bed-sharing situation and should be avoided, especially in the youngest infants.
Faulty or inadequately constructed cots can cause entrapment and death.
Infant deaths often occur due to a change in the regular sleep routine, when an infant is placed in an unsafe sleeping place.
Introduction
Accidental suffocation in bed has been reported previously1 ,2 and falls under the broader classification of Sudden Unexpected Death in Infancy (SUDI). However some infant deaths due to accidental suffocation may be classified as Sudden Infant Death Syndrome (SIDS) due to the common risk factor of bed sharing. The reporting of SUDI deaths allows the inclusion of a number of other categories of infant death which may share features of SIDS, such as accidental suffocation, thus avoiding classification bias3–5 and provides an opportunity for a clearer understanding of the causes of sudden death in infancy. SUDI is an important cause of death in infants aged between 28 days and 1 year, accounting for up to 40% of postnatal deaths in New Zealand.6 SIDS, which contributes to SUDI deaths and still accounts for 14.6% of postnatal deaths in England and Wales,7 has declined internationally over the last few decades to 0.1–0.8 deaths per 1000.8–11 The reduction in SIDS deaths has been largely due to the identification and subsequent public education of the major SIDS risk factors of prone sleep position, tobacco exposure, bed sharing and lack of breast feeding.10 ,12 ,13 Despite these dramatic improvements, avoidable deaths are still occurring in the infant population. In New Zealand the number of deaths reduced from more than 200 deaths per annum in the late 1980s to 50 deaths per annum14 in recent years. However, if all infants were placed to sleep supine in their own safe sleep space free of tobacco exposure the mortality rate could be as low as 0.08 deaths per 1000 live births. This equates to six to seven deaths per annum in New Zealand.12
We reviewed all SUDI deaths in New Zealand from 2002 to 2009 to describe the situations where suffocation in place of sleep occurred in order to identify preventable circumstances to help further reduce rates of SIDS/SUDI.
Method
The Child and Youth Mortality Review Committee (CYMRC) and the Perinatal and Maternal Mortality Review Committee (PMMRC) store data regarding deaths between 0 days and 25 years in the New Zealand Mortality Review Database (NZMRD). The NZMRD is a comprehensive collection of data from multiple sources. Sources include police reports (police attend all unexpected infant deaths and complete a detailed death scene examination with a narrative of events leading to the discovery of the infant) postmortem report, coroner reports, death certificate, demographic and health utilisation data (which includes records of immunisations, medical prescriptions that have been filled, general practitioner visits and hospital admissions). In addition local mortality review groups add data after a multidisciplinary review, which includes representatives from health, education and social service providers. During this review, services share any relevant information that might not ordinarily be available to each service, such as child protection, family violence, health and education issues.
Table 1 shows the ICD-10-AM ‘cause of death’ codes that were used to detect possible cases. Detailed information from the NZMRD was then used to assign inclusion and exclusion criteria.
Inclusion criteria were: that the death was clearly unintentional, had strong evidence of external compression of the neck leading to strangulation or external compression of the chest leading to suffocation, or oronasal compression or blockage of the airway with an object other than vomit and that the death occurred during sleep.
Exclusion criteria were: that it was unclear if the death was unintentional, such as deaths with any indication of non-accidental injury, if postmortem findings were consistent with a cause other than suffocation, such as sepsis, or if there was insufficient information to permit further classification.
Final cases for inclusion were decided after independent review and consensus agreement by three paediatrician investigators (RMH, SRD and NJdCB).
Demographic data, risk factors for SIDS/SUDI, and a descriptive account surrounding the death, including the mechanism were collected. Cases were assigned descriptive terms to group them as either wedging deaths or overlay deaths. Wedging was defined to occur when an infant was found with face, neck or chest trapped in a manner that would impede breathing and/or compromise the airway in their sleeping space. Overlay was defined to occur when an infant was found with the body part/body of a co-sleeping partner causing external compression of the neck or chest, or oronasal compression or blockage of the airway. Emphasis was placed on the narrative account obtained by police at the death scene investigation.
Data are presented using descriptive statistics. Rates are calculated using all New Zealand live births for the study time frame as the denominator.
Results
Between 2002 and 2009, the New Zealand CYMRC and PMMRC reported 492 SUDI deaths in infants less than 1 year of age, equating to a rate of 1.03 deaths per 1000 live births. An additional five deaths were identified by the broader search of ICD-10-AM codes from the NZMRD. Following detailed review of all NZMRD information 48 of the deaths were classified as suffocation in place of sleep (0.10 per 1000 live births). Of the 48 deaths, 36 (75%) had been classified with an ICD-10-AM code of Accidental suffocation and strangulation in bed (W75), 8 (17%) classified as SIDS (R95), 3 (6%) classified as other ill-defined and unspecified cause of mortality (R99) and 1 (2%) as other specified threat to breathing (W83).
Detailed and narrative information came from four main sources, in addition to the demographic data. The majority had a coroner report, 44 (92%). Thirty-nine (81%) had a police report, 35 (73%) had a postmortem report and 12 (25%) had a local mortality group review.
Twenty-four (50%) deaths occurred in male infants. Death was more prevalent in younger infants with 11 (23%) of the deaths occurring in those aged 1 month or less (figure 1). Maori and Pacific infants had a higher mortality rate compared with European infants (Maori rate ratio 6.8; Pacific rate ratio 3.9) (table 2). Thirty (63%) deaths were classified as being due to overlay, with 18 (38%) deaths classified as due to wedging (table 3).
Overlay deaths
All 30 deaths due to being overlaid were in a co-sleeping situation. Of the 48 suffocation deaths 8 (17%) infants were overlaid by their mother while breast feeding, 4 (8%) infants were overlaid by their sibling in a co-sleeping situation and 17 (35%) infants were overlaid by either their mother or father in a co-sleeping situation (the individual who overlaid the infant was undetermined in one case).
The eight deaths while breast feeding occurred when the mother fell asleep while feeding and infants were found entrapped against the mother's arm, side or armpit. In 9 of the 30 deaths infants were taken from their own cot to be fed in their parents’ bed. In 2 of the 30 deaths infants were taken to a couch or sofa to be fed by either parent. In both situations death occurred when the infant became entrapped when the parent fell asleep. In the four deaths where infants were overlaid by their sibling, infants were sharing a sleeping surface with adults and older siblings, older siblings alone, or the infant was placed nearby and was moved (either by older sibling or fell) into a shared sleeping surface.
Wedging deaths
There were 18 wedging deaths. Two (4%) infants were wedged between a sleeping surface (ie, mattress) and bedding, 10 (21%) infants were wedged between a sleeping surface and a wall or broken cot and 6 (13%) infants were wedged between a couch or sofa and cushions. Seven of the deaths due to wedging occurred in cots, six of which were faulty directly causing entrapment. For the seventh infant there was insufficient information available on the condition of their cot. Faults included incomplete assembly of the cot, inappropriately sized mattresses that allowed wedging between the mattress and the base of the cot, and faulty sides to cots allowing infants’ heads to slip through. One cot was recorded as secondhand, while another was recorded as a portable cot.
Some of the wedging deaths were due to makeshift sleeping arrangements, created with cushions, mattresses and bedding. In some situations the infant was moved from a safe sleeping place to an unsafe place to be settled or fed.
Location and circumstance of last sleep
Information regarding where the infants were put to sleep was available for all of the 48 deaths. Seven (15%) infants were placed to sleep in a cot, 32 (67%) infants in a bed and 9 (19%) infants on a couch or chair. Only 13 (27%) deaths occurred in what was recorded as the infant's routine sleeping situation. Twelve (25%) deaths occurred away from home, all in makeshift bedding arrangements, with the remainder of deaths occurring due to breaks from routine with infants moved for feeding, to help them settle, because of a social gathering, home renovations, household overcrowding or low environmental temperature.
Bed sharing
Information regarding co-sleeping was available in 46 (96%) of the 48 deaths. Overall, 34 (71%) of the deaths occurred in a co-sleeping situation (30 of 30 overlay deaths; 4 of 18 wedging deaths).
SUDI/SIDS risk factors
Unfortunately, the presence, and particularly the absence of risk factors for SUDI/SIDS were not routinely recorded in the NZMRD. However, two (4%) infants were placed to sleep prone, nine (19%) infants were exposed to household tobacco smoke ex utero, eight (17%) infants were noted to have an illness at the time of death (most commonly an upper respiratory tract infection), three (6%) infants were premature and in four (8%) cases alcohol use by caregivers was noted to be a important contributing factor.
Discussion
In this review we highlight 48 deaths that were identified as suffocation in place of sleep in New Zealand infants between 2002 and 2009. Consistent with previous research, accidental suffocation has been described in the parental bed, when a co-sleeping partner obstructs the infant's airway, or pillows and bedding cover the infants face.2 ,15 Occasionally a sibling is unintentionally responsible for causing suffocation in a shared sleeping space.15 ,16
In our cohort, the youngest infants were at highest risk.2 ,17 This could be due to many reasons; they are less able to alert caregivers or extract themselves from compromising situations; they have softer, more compressible airways; carers are more likely to be sleep deprived; safety concerns may not always be immediately obvious.
Undoubtedly placing an infant on a sleep surface such as a chair, couch or pillow or adult bed to sleep leads to risk of wedging.1 ,18 But a less obviously risky situation arises where an infant is picked up from a safe sleep surface and is settled or fed on the couch and the caregiver falls asleep. Additionally, makeshift bedding arrangements have the potential to create spaces that the infant can become trapped in.19 Makeshift arrangements often demonstrate that caregivers have a focus on softness, for example, use of pillows or beanbags, rather than a firm but safe space.
Two-thirds of deaths were in a bed-sharing situation, despite evidence that bed sharing is not a common practice. Only 8% of a cohort of 6184 New Zealand children were reported to share the parental bed.20 Of these almost two-thirds were in a defined space, such as a pepi-pod or wahakura (local devices that separate the infant from their co-sleeping partner/s). Those sharing the parental bed were more likely to be Asian or Pacific peoples, with over 10% sleeping in the parental bed at 6 weeks of age.20 However, in a smaller survey of Maori mothers, more than half of whom smoked during pregnancy, 21% reported sharing a bed with their infant/s.21 Despite the low rate of bed sharing in the general population (8%), Hutchison reported on a SUDI cohort of 221 infants, 141 (64%) of who were bed sharing at the time of death. Alarmingly this was more common in young infants (92%).22 In addition, 83% of the deaths were Maori or Pacific infants.
Maori and Pacific peoples are over-represented in our study also. This may be due to the increased bed sharing with young infants in addition to domicile in overcrowded houses, with less access to safe cots and safe alternative sleeping arrangements.
Unfortunately heeding basic safe sleep messages was not enough for some infants, due to faulty cots. Cot entrapments due to failure of the structure or hardware of the cot have been previously reported.15 ,16 We have recommended elsewhere that secondhand cot sales should have a warning and checklist for cot safety.23
There are limitations of this study, namely the lack of control cases to compare risk behaviours with. In addition the detailed information for each case was variable, with some having few risk factors recorded, briefer narratives and occasionally no coronial review. Our inability to analyse the data with respect to risk factors will lead to an underestimation of the impact of known risk factors, but also previously reported risk factors for overlay such as alcohol and obesity.17 ,18 We feel confident that each case we report on was definitely caused by accidental suffocation. The detailed police-recorded narrative at the death scene investigation left no room for conjecture, with deaths described clearly as overlay or wedging. However, we are aware that there are more cases that were highly likely to be due to accidental suffocation, but were excluded as the death scene description did allow for conjecture. For example, there was one excluded case, where, presumably the mother rolled over the infant, as she described going to bed on one side of the infant but woke on the other side of the bed, finding the infant deceased.
Despite these limitations, we feel that three key messages can be taken from these data. The first is that accidental suffocation can occur during bed sharing, and that sharing a sleep surface with an infant, particularly an infant under 3 months of age should be avoided. With the knowledge provided from the recent individual level analysis of five major SIDS case-control studies,12 bed sharing has been shown to confer at least a fivefold increase in death in an infant less than 3 months of age with no other risk factors. We speculate that one mechanism for this could be accidental suffocation.
Ensuring that all cots are safe and correctly assembled is the second key message. New Zealand has a compulsory standard for cots (AS/NZS 2172:2003), however it is difficult to regulate resale or to assess reused cots with wear and tear. Standards for other types of sleep spaces such as portable cots and bassinets are not mandatory in New Zealand, so not enforced. In a recent review of portable cots for sale in New Zealand four out of six failed to meet the safety Australian/New Zealand standard (AS/NZS 2195:99)24 yet they are still available for sale, as this is not compulsory. At the minimum Well Child home checks by community child health nurses (equivalent to health visitors in the UK) should include an assessment of the cot or other spaces the infant might sleep in.
Finally, we feel it is important to highlight that a quarter of accidental suffocation deaths occurred during a break in routine or under makeshift arrangements. A key safety message for caregivers is for a safe sleep every time, including when away from home, with visitors, moving house or renovating and when settling with a feed. Additional care and careful advanced planning is needed when the caregiver is very tired, may be impaired by alcohol or drugs, when the infant is unsettled or sick and during feeding. The principles of safe sleep should be considered for every sleep and include checking for potential spaces that could cause entrapment.
Acknowledgments
Data provided by Child and Youth Mortality Review Committee (CYMRC) and Perinatal and Maternal Mortality Review Committee (PMMRC), with statistical analysis from the NZ Mortality Review Data Group, University of Otago. The authors thank Dr John Holmes, acting Clinical Leader, NZ Mortality Review Data Group for additional data analysis.
References
Footnotes
Correction notice This paper has been amended since it was published Online First. There was an error in the data in the first sentence of the results section of the abstract. The “rate of 0.41 deaths per 100 000 live births” has been corrected to “a rate of 0.10 deaths per 100 000.”
Contributors RMHayman was a substantial contributor to the design of the work, acquisition, analysis and interpretation of the data. She drafted the work and revised the manuscript critically, gave final approval for the version published and agrees to be accountable for all aspects of the work. GMD was a substantial contributor to the analysis and interpretation of data, responsible for critical revision of the manuscript, gave final approval for the version published and agrees to be accountable for all aspects of the work. NJdCB was a substantial contributor to the design of the work, acquisition, analysis and interpretation of the data. He was responsible for critical revision of the manuscript, gave final approval for the version published and agrees to be accountable for all aspects of the work. EAMwas a substantial contributor to the design of the work, analysis and interpretation of the data. He was responsible for critical revision of the manuscript, gave final approval for the version published and agrees to be accountable for all aspects of the work. SRD was a substantial contributor to the conception and design of the work, acquisition, analysis and interpretation of the data. He was responsible for critical revision of the manuscript, gave final approval for the version published and agrees to be accountable for all aspects of the work.
Funding NZ Child and Youth Mortality Review Committee. EAM is supported by Cure Kids.
Competing interests None.
Ethics approval Northern Regional X Ethics Committee, New Zealand.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement Parts of this data have been published previously as a report for the NZ Ministry of Health, Child and Youth Mortality Review. Special Report: Unintentional suffocation, foreign body inhalation and strangulation, 2013; https://www.hqsc.govt.nz/assets/CYMRC/Publications/CMYRC-special-report-March-2013.pdf (accessed 4 Apr 2014).