Article Text

Use of child death review to inform sudden unexplained infant deaths occurring in a large urban setting
  1. Suzanne N Brixey1,
  2. Brianna C Kopp2,
  3. Amy E Schlotthauer3,
  4. Abigael Collier4,
  5. Timothy E Corden5
  1. 1Department of Pediatrics, Policy Core, Injury Research Center, Medical College of Wisconsin, Downtown Health Center Pediatric Clinic, Milwaukee, Wisconsin, USA
  2. 2Wisconsin Department of Health Services, Madison, Wisconsin, USA
  3. 3Injury Research Center, Medical College of Wisconsin, Milwaukee, Wisconsin, USA
  4. 4Children's Health Alliance of Wisconsin, Milwaukee, Wisconsin, USA
  5. 5Department of Pediatrics, Division of Critical Care Medicine, Policy Core, Injury Research Center, Medical College of Wisconsin, Milwaukee Wisconsin, USA
  1. Correspondence to Dr Timothy E Corden, Department of Pediatrics, Children's Hospital of Wisconsin, 9000 W. Wisconsin Ave, PO Box 1997, MS 681, Milwaukee, WI 53201, USA; tcorden{at}


Objective To illustrate the benefits and utility of the child death review (CDR) reporting system when examining risk factors associated with infant death occurring within two subgroups of sudden unexpected infant deaths (SUID)—unintentional suffocation and sudden infant death syndrome (SIDS)—in a large urban county in Wisconsin.

Design Retrospective CDR data were analysed, 2007–2008, for Milwaukee County, Wisconsin.

Patients or subjects Unintentional suffocation and SIDS infant deaths under 1 year of age in Milwaukee County, Wisconsin, 2007–2008, with a CDR record indicating a death in a sleep environment.

Main outcome measure Study examined demographic characteristics, bed-sharing, incident sleep location, position of child when put to sleep, position of child when found, child's usual sleep place, crib in home, and other objects found in sleep environment.

Results Unintentional suffocation (n=11) and SIDS (n=40) classified deaths with CDR data made up 18% (51/283) of all infant deaths in Milwaukee County from 2007 to 2008. The majority of infants who died of unintentional suffocation (n=9, 81.8%) or SIDS (n=26, 65.0%) were black and under the age of 3 months. Bed-sharing was involved in most of the unintentional suffocation deaths (n=10, 90.9%) and the SIDS deaths (n=28, 70.0%). All unintentional suffocation deaths (n=11, 100%) and the majority of SIDS deaths (n=31, 77.5%) took place in a non-crib sleeping environment.

Conclusions The study demonstrates how CDR provides enhanced documentation of risk factors to help steer prevention efforts regarding SUID deaths in a community and reaffirms infants in an unsafe sleep environment have an increased risk of death.

  • Child
  • community
  • public health

Statistics from


Sudden unexplained infant deaths (SUID) account for approximately 4600 infant deaths annually in the USA.1 The SUID category includes, but is not limited to, deaths due to unintentional suffocation and sudden infant death syndrome (SIDS).1 An analysis of death certificates across the USA found the rate of fatalities attributed to accidental suffocation and strangulation in bed in the first year of life quadrupled between 1984 and 2004.2 Investigators have noted that some increase in unintentional suffocation deaths may represent a shifting of classification and reporting of deaths within the SUID group, with fewer deaths reported as due to SIDS and more attributed to unintentional suffocation or unknown cause.2–4 This shift may reflect efforts over the last two decades to adhere more strictly to SIDS as a diagnosis of exclusion, and more thorough case investigations revealing unintentional suffocation as a more appropriate cause of death.2 The SUID classification and reporting challenges also reflect the overlap that exists regarding the risk factors for these events. Identifying these risk factors across the spectrum of SUID events is an essential first step in understanding where best to place prevention resources targeted at reducing the burden of SUID deaths within a community.

In 2008, Milwaukee County, Wisconsin had a SIDS rate of 0.85 per 1000 live births, whereas the USA as a whole had a SIDS rate of 0.57.5 These rates prompted community advocates to question why rates are higher in the Milwaukee area. This study demonstrates the utility of the child death review (CDR) reporting system by examining Milwaukee County CDR records from 2007 to 2008 to quantify and describe infant deaths in two SUID subgroups with overlapping risk factors—unintentional suffocation, and SIDS. The goal of CDR is to identify areas of prevention aimed at protecting the health of children in a community by analysing the circumstances surrounding the death of a child. By using CDR case review information, this study was able to more fully identify and describe the risk factors associated with unintentional suffocation and SIDS deaths than if death certificate information had been used alone.

CDR teams are multidisciplinary review teams seeking to understand the circumstances surrounding the death of a child. Team members typically include the medical examiner/coroner office, child protective services, law enforcement, public health, paediatricians, district attorneys, school districts, mental health professionals, and others relevant to a specific case. Each of these members brings unique information about the child death to the review meeting. Wisconsin CDR teams are based at the local level in individual counties or multiple counties who share resources. The multidisciplinary review of child deaths allows for comprehensive data collection, capturing demographics, risk factors, and information about the circumstances. Wisconsin participates in the National Center for Child Death Review's Case Reporting System (CDR reporting system), a web-based system collecting comprehensive information from CDR meetings.6 A person from each local team is designated to enter the data from the meetings. Cases are entered into the CDR reporting system where built-in skip patterns prompt the data entry person for information based on the manner and cause of death taken from the death certificate. A data dictionary is provided to all teams to assist with data entry and defining terms.7 Following data entry, this web-based system allows teams to query 33 different standardised reports organised by cause of death to track local trends. The data collected are used to inform community based prevention.


We queried the CDR reporting system for all infant deaths (<1 year of age) from 1 January 2007 to 31 December 2008 occurring in Milwaukee County. The Milwaukee County CDR team reviews all childhood deaths reported to the medical examiner's office. We examined all the unintentional suffocation and SIDS deaths within the time reference; the analysis included information from the auxiliary CDR data field ‘death occurring in a sleeping environment’.

Demographic information used in the analysis included gestational age in weeks, race, and family income status. Incident circumstances analysed included: position of the child at time of sleep and position of child when found (on back, on stomach, on side, unknown), usual sleep place (adult bed, couch, crib, other), and incident sleep location (adult bed, couch, crib, other), whether a crib was found in the home (yes/no), and objects found in the sleep environment (pillow, mattress, wall, blankets, comforter, other). Objects found in the sleep environment pertain to the object(s) that were found to contribute to the death. For example, in the case of ‘mattress’, this may mean that the child was pressed into or wedged into the mattress. The occurrence of bed sharing was noted in the CDR reporting system as ‘Child sleeping on same surface with persons(s) or animal(s)’ with responses categorised as with adults, with other children, with animal(s), unknown, or a combination of these.

Frequency counts were tabulated using the CDR reporting system data via export at the Wisconsin Department of Health Services. SAS version 9.1 was used for analyses. The Wisconsin Child Death Review project is approved by the Children's Hospital Institutional Review Board in Milwaukee.


The CDR reporting system contained 203 infant deaths in Milwaukee County during 2007–2008, representing 72% of all infant deaths that occurred within the county. There were 52 unintentional sleep environment deaths comprising 11 unintentional suffocation deaths and 40 SIDS deaths. One additional infant sleep environment death was classified as ‘undetermined whether injury or medical condition’; this case was excluded from the analysis.

The majority of infants who died of unintentional suffocation (n=9, 81.8%) or SIDS (n=26, 65.0%) were black (table 1); Milwaukee County population was 25% black during the reference period.8 Income level was known in 28 of the SIDS cases, of which 96% (n=27) lived below the federal poverty level. Of the 11 unintentional suffocation deaths, income level was known in seven of the cases; 86% (n=6) lived below the poverty level. During the study interval, 13% of families in Milwaukee County lived below the poverty level.8 Most of the infants who died of unintentional suffocation with known gestational ages were born full term, while most of the SIDS cases were born prematurely (table 1). The majority of children were 3 months of age or less at time of death, regardless of gestational age and regardless of whether the cause of death was identified as SIDS or unintentional suffocation.

Table 1

Race, gestational age, and age at time of death of unintentional suffocation and SIDS deaths, Milwaukee County Wisconsin, 2007–2008

Characteristics of unintentional suffocation deaths

Within the CDR reporting system, sleep location was recorded in all unintentional suffocation deaths. All unintentional suffocation deaths took place in a non-crib sleeping environment, with 54% of the deaths occurring on a couch. A non-crib sleep environment was the child's usual place for sleep in at least 10 of the 11 cases (table 2). Approximately half of the children (n=6, 54.5%) reportedly were placed on their back to sleep, but only three children were found on their backs (27.3%) after the incident (table 2). Although the usual sleep environment and incident sleep location was not a crib in all incidents, the majority (n=6, 54.5%) did have a crib in the home (table 2). Additionally, at least one object (eg, blanket, pillow, adult, or child) contributed to the obstruction of the child's airway in all 11 cases. Bed sharing was involved in almost all unintentional suffocation deaths (n=10, 90.9%) (table 2). The majority of bed sharing was done with adults only (n=8, 72.7%).

Table 2

Characteristics of infant unintentional suffocation and SIDS deaths in Milwaukee County Wisconsin, 2007–2008

Characteristics of SIDS deaths

Within the CDR reporting system, sleep location was recorded in all SIDS deaths. A non-crib sleep environment was the incident location for most deaths (n=34, 85.0%) (table 2). A crib was noted as the usual sleep environment for one fifth of the infants (n=9, 22.5%); however, a crib was present in the home for over half of the infants (n=23, 57.5%) (table 2). The majority of infants dying of SIDS were placed on their backs when put to sleep (n=29, 72.5%), while at least seven infants (17.5%) were not placed on their backs and, at a minimum, 12 (30%) were found not on their back after the event (table 2).

Additionally, 17 of the infants were reported to have fully or partially obstructed airways due to the presence of a pillow, blanket, other bedding, other children, an adult, or a combination of these items. Of the infants with an unobstructed or unknown status of their airway, 25 were compromised in their sleeping environment with one or more of these objects or persons. Bed sharing was an identified factor in 70% of SIDS deaths. SIDS deaths included adult only bed sharing (n=13, 32.5%) and bed sharing with adults and children (n=14, 35.0%) (table 2).


Improving child death data

The public health model of prevention emphasises the use of data to inform community interventions. Using the CDR reporting system data to identify and target the modifiable risk factors for SUID deaths is a logical prevention approach. Shapiro-Mendoza and colleagues have identified what contributing risk factors can be obtained from death certificate text information for both SIDS and unintentional suffocation infant deaths.2 9 Many of their findings are similar to what this study reveals for risk factors based on the Milwaukee CDR data. However, contributing information is often not listed on the death certificate, being absent in almost 80% of the SIDS cases reviewed in the national study conducted by Shapiro-Mendoza et al.2 9

Locally, it is the policy of the Milwaukee County Medical Examiner's office to conduct a death scene investigation by a trained investigator for all deaths occurring in a sleeping environment following the National Association of Medical Examiners' recommendations using an internally developed standardised form.10 The lack of nationwide and worldwide standardisation of how SUID cases are investigated and reported, including the documenting of contributing factors, supports the call for a national SUID registry. The Center for Disease Control's (CDC) SUID Initiative calls for a uniform national approach for SUID cases. The CDC recently began a pilot study examining the CDR reporting system as the platform for an SUID registry.11 This SUID registry may answer many of the questions regarding a possible diagnostic shift from SIDS deaths to suffocation deaths.

Our study illustrates the advantages of using the CDR reporting system as a registry for SUID deaths compared to an analysis of death certificates. In the CDR reporting system, the data are entered into a web based system and can be retrieved in real time by local CDR teams. The CDR reporting system promotes standardisation of data collection, and helps to move local public health prevention efforts towards data and evidence based programming in a more timely fashion than death certificates alone.9 12 Electronic death certificate data are available in Wisconsin as a tool to tabulate broad categories of deaths and basic demographics. However, the level of detail available is not adequate to understand fully the circumstances of the death and, subsequently, identify prevention strategies. The majority of information available on electronic death certificates relates to demographic variables of the decedent. The paper copy of the death certificate provides more information about the circumstances of the death, but availability of the paper copy is limited due to confidentiality and feasibility concerns. Additionally, there is often a cost associated with obtaining a paper copy of the death certificate.

This project also provided an opportunity to learn how to improve the CDR reporting system, to ease the process of manipulating and extracting data, identify key contributory factors, and clarify terminology. Expanded definitions in the CDR data dictionary would remove ambiguity, allowing teams to input data more consistently. Identifying clear, universally agreed upon definitions for SUID cases would also bring greater standardisation to the process. Current CDR data sources include all SUID deaths (whether defined as SIDS or undetermined cause) under the SIDS category, while asphyxia is a separate category. Once asphyxia is chosen as the cause of death, the data entry skip patterns embedded in the CDR reporting system do not allow for the entry of risk factors found in the SIDS data field. SIDS risk factors, such as the presence of secondhand smoke or overheating, also relevant to asphyxia, become unavailable for electronic entry. The skip patterns ease data input, but also limit the functionality of collecting valuable data across cause of death categories. The addition of a child developmental assessment field would also be helpful—for example, knowing if a child has reached the stage of being able to roll to their side or prone when initially placed on their back would aid in the evaluation of sleep related deaths. The CDC supported SUID Case Registry Pilot project, based on the CDR reporting system, will benefit from addressing these concerns.

Using data to inform prevention

Using CDR reporting system data to understand the burden of SUID deaths in Milwaukee County rather than relying on death certificate data alone will improve prevention efforts. The SUID cases presented here represented 18% (51/283) of all infant deaths in Milwaukee County from 2007 to 2008, a size worthy of prevention efforts aimed at reducing infant mortality in the Milwaukee community.

There is an increasing awareness that an unsafe sleep environment increases an infant's risk for SUID events. Although both prone sleeping and a shared sleeping surface are risk factors for unintentional suffocation, their presence independently does not exclude SIDS as a cause of death.13 The CDR reporting system data allows for a more robust evaluation of SUID events regardless of SUID subtype. Local prevention efforts can be more appropriately tailored using data from this analysis. The ‘Back to sleep’ campaign initiated in the early 1990s by the American Academy of Pediatrics and National Institute of Child Health and Human Development led to a 50% decline in SIDS cases in the USA by identifying a ‘modifiable’ risk factor, the prone sleeping position.14 Providing public health education in a targeted fashion to change caregiver practice dramatically reduced prone sleeping. Unfortunately, our analysis identified that at least 20% of the infants were placed in a position other than on their backs prior to sleep. Nationwide, SIDS occurs at a rate of 99.4 per 100 000 live births in blacks compared to a rate of 55.4 per 100 000 live births in whites.15 Similar disparities were identified in Milwaukee. These national and local findings indicate the need to re-emphasise the importance of a culturally appropriate ‘Back to sleep’ message in the Milwaukee community.16

As the decline in SIDS deaths has levelled off,17 other modifiable risk factors have emerged as possible reasons for the lack of further decline in deaths. Bed sharing is identified in 75% (38/51) of all the infant deaths we reviewed. The majority of these deaths occurred within the first 3 months of life. Bed sharing is now well described as a risk factor for SUID.18–21 This analysis also found that 54% of the unintentional suffocations in Milwaukee County occurred on a couch. The benefits of standardisation of collection and subsequent use of this local CDR data, layered with the national and international findings regarding the risk of couch sleeping and bed sharing, can be shared with the Milwaukee community in an attempt to promote education regarding these modifiable risk factors further.

Programmes aimed at ensuring families have available a crib in which their child can sleep safely are a logical prevention approach. Using CDR reporting system data, we found over 40% of the environments did not have a crib present as a sleep option. Unfortunately, our CDR data also support the findings of other studies22 that the existence of a crib does not ensure the device will be used, as only 20% of the families who had a crib chose to use the device on the day of the infant's death. Programmes providing cribs to the community will have to explore ways of influencing adult behaviour such that the device is used. The data also indicate the need to expand the current outcomes research on community crib access programmes such as the S.I.D.S. of Pennsylvania's Cribs for Kids programme.23 For families continuing to bed share with their infant, efforts should be focused on evaluating harm reduction strategies directed at decreasing the risk of SUID while bed sharing.


Milwaukee began using the CDR National Reporting System in June 2009, therefore the data used for this study are the result of reviewing data previously collected by the medical examiner's CDR team and retrospectively entered into the CDR National Reporting System. We anticipate that prospective CDR review of sleep environment cases will enhance data collection and limit incomplete or missing data. In addition, classification of cases may have changed over the course of our study period due to the evolving diagnostic criteria of SUID and SIDS and by the presence of two different Milwaukee County medical examiners over the study period. The Milwaukee County medical examiners also did not preclude the use of ‘SIDS’ as a cause of death in cases where a sleep environment risk factor was noted during the investigation. Lastly, CDR data and death certificate data were not directly linked, so we could not fully explore the benefit of CDR review over matched death certificate data alone.


This analysis demonstrates the utility of using information from the National Child Death Review Data System as a method of more fully understanding risk factors for infant death, while providing an example of how the CDR reporting system may serve as a national SUID registry. Additionally, we have documented data on potentially modifiable risk factors that can be used to inform prevention practices in a local community with a high rate of SUID. The use of local data can help inform questions and diffuse conflicts of opinion about what is occurring in one's own community. With local data as a foundation, it should become easier to develop relationships between diverse partners in prevention, engage appropriate community leaders, and ultimately develop supportive and appropriate prevention messages that are aimed at the goal of CDR—to prevent future childhood death and injury.

What is already known on this subject

  • Unsafe sleep environment increases an infant's risk of sudden unexplained infant death (SUID).

  • The child death review (CDR) reporting system makes it possible to collect information about the sleep environment surrounding infant deaths.

What this study adds

  • CDR reporting system data will enhance death certificate data by providing supplemental information on the sleep environment that will inform the cause of death.

  • This study demonstrates the ability and value of CDR to reveal the risk factors for SUID deaths regardless of whether they are classified as unintentional suffocation or sudden infant death syndrome (SIDS).

  • Bed sharing is reaffirmed as a common risk factor for SUID deaths.


The authors of this paper would like to thank Julia McCallum and Shannon Baumer for their assistance in entering Milwaukee County data into the National Child Death Review Database. Additionally, we would like to thank Dr Christopher Happy, former Medical Examiner for Milwaukee County, for the provision of data and assistance with this manuscript, and Mary Czinner, Program Coordinator at the IRC, for her thoughtful editorial assistance.


View Abstract


  • Funding This work was partially supported by Centers for Disease Control and Prevention Grants R49/CE00175 and 5U17CE524815-05 as well as the Keeping Kids Alive grant provided by the University of Wisconsin School of Medicine and Public Health, The Wisconsin Partnership Program.

  • Competing interests None.

  • Ethics approval This study was conducted with the approval of the Children's Hospital of Wisconsin Institutional Review Board.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.