Page 1
Copyright © 2004 American Academy of Pediatrics
Pediatrics 2004; 114: 234-238
July, 2004
SECTION: SPECIAL ARTICLES
LENGTH: 3458 words
TITLE: Sudden Infant Death Syndrome and Unclassified Sudden Infant Deaths: A Definitional and Diagnostic Approach
AUTHOR: Henry F. Krous, MD<1>; J. Bruce Beckwith, MD<2>; Roger W. Byard, MD<3>; Torleiv O. Rognum, MD,
PhD<4>; Thomas Bajanowski, MD<5>; Tracey Corey, MD<6>; Ernest Cutz, MD<1>; Randy Hanzlick, MD<7>;
Thomas G. Keens, MD<8>; and Edwin A. Mitchell, MD<9>
ABSTRACT: The definition of sudden infant death syndrome (SIDS) originally appeared in 1969 and was modified 2
decades later. During the following 15 years, an enormous amount of additional information has emerged, justifying
additional refinement of the definition of SIDS to incorporate epidemiologic features, risk factors, pathologic features,
and ancillary test findings. An expert panel of pediatric and forensic pathologists and pediatricians considered these
issues and developed a new general definition of SIDS for administrative and vital statistics purposes. The new definition was then stratified to facilitate research into sudden infant death. Another category, defined as unclassified sudden infant deaths, was introduced for cases that do not meet the criteria for a diagnosis of SIDS and for which alternative diagnoses of natural or unnatural conditions were equivocal. It is anticipated that these new definitions will be modified in the future to accommodate new understanding of SIDS and sudden infant death.
[SIDS, sudden infant death.]
TEXT:
ABBREVIATION. SIDS, sudden infant death syndrome.
Sudden infant death syndrome (SIDS) is a term that has been used to describe unexpected deaths of infants or
young children when subsequent investigations fail to demonstrate a definite cause of death. [n1,n2] The concept, which
was first proposed in 1969, has been controversial, and its use has been characterized by great variability in the consistency with which the requirements of standard definitions have been fulfilled. [n3-n5] Specifically, the term has been
overused and applied to cases in which there have been obvious natural or unnatural causes of death; also, the term has
been underused in favor of imprecise terms such as undetermined or unascertained. A number of other definitions that
have included quite different criteria have been proposed. [n6-n8] The most widely used definitions have made SIDS a
diagnosis of exclusion.
In 1969, at the Second International Conference on Causes of Sudden Death in Infants, it was proposed that SIDS
was "the sudden death of any infant or young child which is unexpected by history, and in which a thorough postmortem examination fails to demonstrate an adequate cause of death." [n1] In 1989, the National Institute of Child Health
and Human Development convened an expert panel to reexamine the issue of definition. The panel proposed that SIDS
was "the sudden death of an infant under one 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."
[2(p681)] This definition limited the age to <1 year and specified that a thorough examination should include examination of the death scene and review of the clinical history.
In 1992, at the SIDS International Meeting in Sydney, Australia, Bruce Beckwith proposed stratification of the definition to enable separation of cases into typical and atypical groups. [n9] The proposal was not accepted at the time,
Page 2
although others subsequently supported subclassification. [n10] In 2003, Beckwith again called for a reexamination of
the definition of SIDS, with the possibility of including positive diagnostic criteria and stratification to delineate particular subsets. [n11] As a result of this proposal, a meeting was held in San Diego, California, in January 2004; it was
sponsored by the CJ Foundation for SIDS and involved an invited panel of experts, including pediatric pathologists,
forensic pathologists, and pediatricians, all of whom had extensive experience with sudden infant death. Delegates
came from Europe, North America, and Australasia.
DISCUSSION POINTS
Presentations on a variety of topics were made first. Bruce Beckwith (Loma Linda University, Loma Linda, CA)
discussed the history of SIDS definitions, the present status, and possible proposals for the future. He indicated that the
meeting represented an attempt to formulate an approach to sudden infant death that would clarify subsets and assist
research but that it was a work in progress that would need to be revisited regularly. Data presented from Seattle in the
1960s and 1970s demonstrated that 95% of SIDS cases were within the age range of 1 to 6 months. Edwin Mitchell
(University of Auckland, Auckland, New Zealand) also discussed the need for redefinition and the possible place of risk
factors in a new definition. He pointed out that, although specific risk factors were consistent across studies, the prevalence varied among countries; therefore, he considered it better to leave risk factors out of definitions. Mitchell emphasized the importance of researchers defining their study populations carefully. Henry Krous (Children's Hospital San
Diego, San Diego, CA) presented data from the San Diego study on the frequency of risk factors and the changes that
have occurred since the Back to Sleep campaigns, specifically the decrease in the winter peak and the proportionate
increase in the number of cases <1 month of age. Torleiv Rognum (University of Oslo, Oslo, Norway) discussed the
need for a new definition and particular problems with SIDS because of its uneasy position between the health system
and the legal system. Data from Oslo in the 1980s showed an age distribution similar to the Seattle results; in later
years, however, there was a decrease in the number of cases 2 to 4 months of age, with increases in the numbers of cases involving younger and older infants. When cases >360 days of age were excluded from the latter cohort, however,
the distribution of cases was similar to data from Seattle before the Back to Sleep program. Randy Hanzlick (Fulton
County Medical Examiner's Center, Atlanta, GA) discussed the advantages and disadvantages of reporting on the death
certificate, or in other SIDS databases, those risk factors that might have been operative in causing or contributing to
death, emphasizing that reporting such conditions might enable better tracking through official documentation. Hanzlick
suggested that the group discuss the merits of abandoning the term SIDS and replacing it with "sudden unexplained
infant death." Problems with the use and scope of the International Classification of Disease coding were reviewed,
including sometimes-inadequate specificity, overlap, and variable application among coders. Roger Byard (Forensic
Science Centre, Adelaide, Australia) discussed the use of the terms undetermined and unascertained in flagging cases in
which significant parts of the investigation were lacking or in which there were questions regarding possible causes of
death. Byard also warned against the indiscriminate use of these terms to cover inadequate autopsy and case investigations.
A group discussion followed, during which the advantages of formulating and promulgating a redefinition of SIDS
were actively debated. It was agreed that creating and supporting a more inclusive SIDS definition would facilitate uniformity in diagnosis, with a resultant increase in information on current cases. It would also enable accumulated data to
be better used and would provide opportunities to propose and evaluate new theories, particularly regarding possible
SIDS subsets. Existing SIDS definitions were considered inadequate, often being applied too generally or too restrictively, and were exclusionary, failing to incorporate known features of the syndrome (such as sleep and age range). The
conclusions of the group were based on assessments of current trends and data and were intended to be fully reevaluated
in the future, when they will likely need to be modified to accommodate new developments.
The redefinition was also considered a useful step to enable more precise monitoring of changing epidemiologic
patterns in sudden infant deaths and to allow more valid international comparisons. By more clearly defining subsets
of sudden infant deaths, monitoring of the effects of public health recommendations and alterations in infant care practices can be facilitated. Finally, more precise definitions of subsets of sudden infant deaths, with specification of requirements for diagnosis, should help standardize investigative protocol development, by improving examinations of the
circumstances of death and autopsy investigations and bringing investigations more in line with recommended guidelines. [n12-n14] Providing more information and more rigorous subclassification of cases should also facilitate integrated multiagency approaches to such cases. [n15]
RESULTS
Page 3
Definitional Approach to Sudden Infant Death
The following definition and subclassification were agreed upon.
General Definition of SIDS
SIDS is defined as the sudden unexpected death of an infant <1 year of age, with onset of the fatal episode apparently occurring during sleep, that remains unexplained after a thorough investigation, including performance of a complete autopsy and review of the circumstances of death and the clinical history.
Category IA SIDS: Classic Features of SIDS Present and Completely Documented
Category IA includes infant deaths that meet the requirements of the general definition and also all of the following
requirements.
Clinical
* More than 21 days and <9 months of age.
* Normal clinical history, including term pregnancy (gestational age of >/=37 weeks).
* Normal growth and development.
* No similar deaths among siblings, close genetic relatives (uncles, aunts, or first-degree cousins), or other infants in the
custody of the same caregiver.
Circumstances of Death
* Investigation of the various scenes where incidents leading to death might have occurred and determination that they
do not provide an explanation for the death.
* Found in a safe sleeping environment, with no evidence of accidental death.
Autopsy
* Absence of potentially fatal pathologic findings. Minor respiratory system inflammatory infiltrates are acceptable;
intrathoracic petechial hemorrhage is a supportive but not obligatory or diagnostic finding.
* No evidence of unexplained trauma, abuse, neglect, or unintentional injury.
* No evidence of substantial thymic stress effect (thymic weight of <15 g and/or moderate/severe cortical lymphocyte
depletion). Occasional "starry sky" macrophages or minor cortical depletion is acceptable.
* Negative results of toxicologic, microbiologic, radiologic, vitreous chemistry, and metabolic screening studies.
Category IB SIDS: Classic Features of SIDS Present but Incompletely Documented
Category IB includes infant deaths that meet the requirements of the general definition and also meet all of the criteria for category IA except that investigation of the various scenes where incidents leading to death might have occurred was not performed and/or >/=1 of the following analyses was not performed: toxicologic, microbiologic, radiologic, vitreous chemistry, or metabolic screening studies.
Category II SIDS
Category II includes infant deaths that meet category I criteria except for >/=1 of the following.
Clinical
* Age range outside that of category IA or IB (ie, 0-21 days or 270 days [9 months] through first birthday).
* Similar deaths among siblings, close relatives, or other infants in the custody of the same caregiver that are not considered suspect for infanticide or recognized genetic disorders.
* Neonatal or perinatal conditions (for example, those resulting from preterm birth) that have resolved by the time of
death.
Circumstances of Death
Page 4
* Mechanical asphyxia or suffocation caused by overlaying not determined with certainty.
Autopsy
* Abnormal growth and development not thought to have contributed to death.
* Marked inflammatory changes or abnormalities not sufficient to be unequivocal causes of death.
Unclassified Sudden Infant Death
The unclassified category includes deaths that do not meet the criteria for category I or II SIDS but for which alternative diagnoses of natural or unnatural conditions are equivocal, including cases for which autopsies were not performed.
Postresuscitation Cases
Infants found in extremis who are resuscitated and later die ("temporarily interrupted SIDS") may be included in
the aforementioned categories, depending on the fulfillment of relevant criteria.
DISCUSSION
Before the meeting, participants were asked to submit their own definitions of SIDS. From this pool, a common
definition that incorporated agreed-upon points was formulated. Although in previous definitions SIDS was a diagnosis
of exclusion, it was recognized that there were certain repetitive features common to the majority of cases. It was these
features that led early investigators to suggest that most, but not all, sudden, unexplained, postneonatal deaths represent
a distinct syndrome, reflecting a common cause or, more likely, a common mechanism of death. The elements of this
presumptive syndrome were ignored during formulation of the previous exclusion-based definitions, however. Major
features emphasized in early SIDS studies included an association with sleep and a relatively narrow age range and frequency distribution, sparing the first weeks of extrauterine life, peaking during the 2nd to 4th months, and declining
rapidly thereafter. Deaths with onset while awake are rare and most probably involve a different mechanism of death,
compared with classic SIDS deaths. [n16] A general definition that involved these specific criteria was created, to include as many cases as possible in the SIDS classification. This was thought to be useful for certification purposes and
also for general epidemiologic studies. The phrase "death scene examination" was changed to "review of the circumstances of death," with the aim of encouraging more comprehensive assessments of the events surrounding death. For
example, although an infant's death may occur in a hospital after attempted resuscitation and thus the scene is technically an emergency department, the circumstances involve the crib, room, and house where the infant was found. Review
of the circumstances of death includes not only examination of the death scene but also assessment of all of the environments an infant might have been in before or after death.
After implementing a broad overall definition, participants decided to subcategorize SIDS cases on the basis of
specific epidemiologic features and the amount of information available. This was prompted partly by the knowledge
that the number of classic SIDS cases, typical of those occurring in the 1970s and 1980s (before the Back to Sleep and
Reduce the Risks campaigns), had decreased and it was likely that the remaining cases represented a relatively more
heterogeneous group, with varied underlying mechanisms of death.
Stratification of cases of sudden infant death into subcategories was therefore undertaken to:
* Provide recommended guidelines for general case assessment, classification, and diagnosis.
* Assist pathologists by detailing steps for infant death investigation and diagnostic categorization.
* Identify and include cases that recently have been excluded incorrectly from SIDS groups because of findings of bedsharing and prone sleeping position.
* Formalize current practices among pathologists of separating cases on the basis of the degree of certainty and the confidence with which a diagnosis of SIDS can be made.
* Reduce diagnostic confusion by introducing uniform terms.
* Provide a framework for researchers and identify the most typical cases for study.
* Assist in the evaluation of published data.
* Provide a readily accessible categorization of SIDS cases on the basis of age groups and investigative information.
Page 5
Dividing cases of sudden infant death that fit the general definition of SIDS into subgroups should not have an effect on epidemiologic studies of the syndrome as a whole, because most cases would still be classified as SIDS. However, researchers looking for classic SIDS cases to study could take them from category IA. The group considered it important that researchers specify which subgroups were used for studies, because that would enable immediate assessment of the rigor with which cases had been investigated and determination of how closely the study group represented
classic SIDS cases. It was also recommended that future research should be undertaken to examine similarities and differences among the subgroups, which might clarify specific causes.
The age range of 3 weeks to 9 months was chosen on the basis of an analysis of data from the Avon and Confidential Enquiry Into Stillbirths and Death in Infancy studies in the United Kingdom, New Zealand Health Information Services data, the Chicago and San Diego SIDS studies in the United States, studies at the Rettsmedisinsk Institutt in Oslo,
Norway, and the Westphalian and German Sudden Infant Death studies in Germany, by Edwin Mitchell. Pooling of
data from those studies showed that the 5th to 95th percentile limits for SIDS deaths were [approximately]3 weeks to 9
months.
Prone position was considered an established risk factor for SIDS deaths but not a cause of suffocation unless specific circumstances (such as a face-down position on an incompletely filled waterbed or in a thin plastic bag) could be
demonstrated. For this reason, infants found prone with no evidence of suffocation could be included in any of the categories, depending on other features. Prone sleeping involves an array of potential problems, including diaphragmatic
splinting/fatigue, rebreathing of carbon dioxide, reflex lowering of vasomotor tone with tachycardia, blunting of arousal
responses (including decreased cardiac responses to auditory stimulation), alteration of sleep patterns, upper airway
obstruction resulting from soft bedding, and overheating, and is most likely a problem only among infants with underlying susceptibilities. [n17] Similarly, although there is evidence of increased risk of infant death in shared sleeping situations, [n18] shared sleeping does not automatically exclude SIDS as a possibility, if it can be shown that the infant was
not at risk of accidental asphyxia. It should be recognized that the position in which the infant is found sometimes reflects agonal movement and is not necessarily the position of the infant at the onset of the fatal event.
It was acknowledged that a number of different and variably defined terms were being used to classify unexpected
infant deaths. Sudden unexpected death in infancy is a general term that covers SIDS and other types of unexpected
infant deaths. When an infant dies suddenly and unexpectedly and intentional or unintentional fatal injury can be excluded, death may be attributable to a specific disease entity, such as myocarditis, or to SIDS. [n19] Deaths that cannot
be precisely subcategorized or classified have been deemed undetermined, undeterminable, unascertained, or unascertainable, but this has created concerns about the specificity of these terms. The merits of replacing the term SIDS with
sudden unexplained infant death were briefly discussed during the meeting, but it was the consensus of the group that
SIDS still served a useful purpose. The term unclassified sudden infant death was proposed to account for cases in
which the criteria for category I or II SIDS were not met or an autopsy was not performed. These cases may represent
SIDS deaths, but there is insufficient information available to make that judgment or there are certain atypical features,
such as inflicted but nonlethal injury, that are insufficient by themselves to establish a cause of death but are thought to
preclude use of the term SIDS. Atypical features may also include underlying organic diseases, such as an anomalous
coronary artery without evidence of myocardial ischemia, which may also preclude a confident statement about a possible cause of death.
The investigation of infant deaths should be conducted according to established protocols [n14,n20] and should include careful evaluation of the death scene, external examination of the body with photographic documentation, radiologic examination, internal examination with photographic documentation, and histologic, microbiologic, toxicologic,
biochemical, metabolic screening, and genetic studies if indicated. Guidelines for and confirmation of the usefulness of
such stepwise examinations are available in the literature. [n21-n23]
Finally, it should be reiterated that these proposals represent nothing more than attempts to improve definitions and
to facilitate more accurate investigation, diagnosis, and categorization of cases of unexpected infant death. Considerable
public, professional, and media attention has been paid recently to certain high-profile court cases in which the standards of investigation and pathologic analyses were far from acceptable. This does not mean that the underlying concepts
are flawed; rather, it means that diagnostic terms and protocols should be more rigorously defined and standard investigative approaches should be maintained. The proposed framework is a work in progress, which will need to be continually reformulated and refined as more knowledge becomes available and our understanding of these complex and challenging cases becomes clearer. (At a pathology workshop in Canberra, Australia, in March 2004, forensic and pediatric
pathologists representing forensic institutions and hospitals from all Australian states and territories unanimously endorsed the new general San Diego definition and recommended its national implementation.)
Page 6
SUPPLEMENTARY INFORMATION: Received for publication Mar 1, 2004; accepted Mar 4, 2004.
From the <1> Department of Pathology, Children's Hospital San Diego, University of California, San Diego, School of
Medicine, San Diego, California; <2> Department of Pathology and Human Anatomy, Loma Linda University, Loma
Linda, California; <3> Forensic Science Centre and Departments of Paediatrics and Pathology, University of Adelaide,
Adelaide, Australia; <4> Rettsmedisinsk Institutt and University of Oslo, Oslo, Norway; <5> Institut fur Rechtsmedizin
and University of Essen, Essen, Germany; <6> Office of the Chief Medical Examiner and University of Louisville
School of Medicine, Louisville, Kentucky; <10> Department of Pediatric Laboratory Medicine, The Hospital for Sick
Children and the University of Toronto, Toronto, Ontario, Canada; <7> Emory University School of Medicine and the
Fulton County Medical Examiner's Center, Atlanta, Georgia; <8> Division of Pediatric Pulmonology, Children's Hospital Los Angeles, Los Angeles, California; and <9> Department of Paediatrics, University of Auckland, Auckland, New
Zealand.
Report of an expert panel convened by the CJ Foundation for SIDS, San Diego, California, January 8-9, 2004
(moderator: Henry F. Krous). Address correspondence to Henry F. Krous, MD, Children's Hospital and Health Center,
3020 Children's Way, M5007, San Diego, CA 92123. E-mail: hkrous@chsd.org
REFERENCES:
[n1.] Beckwith JB. Discussion of terminology and definition of sudden infant death syndrome. In: Bergman AB,
Beckwith JB, Ray CG, eds. Sudden Infant Death Syndrome: Proceedings of the Second International Conference on
Causes of Sudden Death in Infants. Seattle, WA: University of Washington Press; 1970:18
[n2.] Willinger M, James LS, Catz C. Defining the sudden infant death syndrome (SIDS): deliberations of an expert
panel convened by the National Institute of Child Health and Human Development. Pediatr Pathol. 1991;11:677-684
[n3.] Gilbert-Barness E. Is sudden infant death syndrome a cause of death? Am J Dis Child. 1993;147:25-26
[n4.] Meadow R. Unnatural sudden infant death. Arch Dis Child. 1999;80:7-14
[n5.] Emery JL. Is sudden infant death syndrome a diagnosis? BMJ. 1989;299:1240
[n6.] Cordner SM. The definition of sudden infant death syndrome. In: Rognum TO, ed. Sudden Infant Death Syndrome: New Trends in the Nineties. Oslo, Norway: Scandinavian University Press; 1995:18-20
[n7.] Sturner WQ. SIDS redux: is it or isn't it? Am J Forensic Med Pathol. 1998;19:107-108
[n8.] Rambaud C, Guilleminault C, Campbell PE. Definition of the sudden infant death syndrome. BMJ.
1994;308:1439
[n9.] Beckwith JB. Discussion of Terminology and Definition of the Sudden Infant Death Syndrome. Ithaca, NY: Perinatology Press; 1993
[n10.] Czegledy-Nagy EN, Cutz E, Becker LE. Sudden death in infants under one year of age. Pediatr Pathol.
1993;13:671-684
[n11.] Beckwith JB. Defining the sudden infant death syndrome. Arch Pediatr Adolesc Med. 2003;157:286-290
[n12.] Hanzlick R, Parrish RG. Death investigation report forms (DIRFs): generic forms for investigators (IDIRFs) and
certifiers (CDIRFs). J Forensic Sci. 1994;39:629-636
[n13.] Krous H. An international standardised autopsy protocol for sudden unexpected infant death. In: Rognum TO, ed.
Sudden Infant Death Syndrome: New Trends in the Nineties. Oslo, Norway: Scandinavian University Press; 1995:8195
Page 7
[n14.] Iyasu S, Rowley D, Hanzlick R. Guidelines for death scene investigation of sudden, unexplained infant deaths:
recommendations of the inter-agency panel on sudden infant death syndrome. MMWR Morb Mortal Wkly Rep.
1996;45:1-6
[n15.] Fleming PJ, Blair PS, Sidebotham PD, Hayler T. Investigating sudden unexpected deaths in infancy and childhood and caring for bereaved families: an integrated multiagency approach. BMJ. 2004;328:331-334
[n16.] Dancea A, Cote A, Rohlicek C, Bernard C, Oligny LL. Cardiac pathology in sudden unexpected infant death. J
Pediatr. 2002;141:336-342
[n17.] Byard RW, Krous HF, eds. Sudden Infant Death Syndrome: Problems, Progress and Possibilities. London, UK:
Arnold; 2001
[n18.] Scragg RK, Mitchell EA. Side sleeping position and bed sharing in the sudden infant death syndrome. Ann
Med. 1998;30:345-349
[n19.] Byard R. Sudden Death in Infancy, Childhood and Adolescence. 2nd ed. Cambridge, UK: Cambridge University
Press; 2004
[n20.] Krous HF, Byard RW. International standardized autopsy protocol for sudden unexpected infant death: appendix
I. In: Byard RW, Krous HF, eds. Sudden Infant Death Syndrome: Problems, Progress and Possibilities. London, UK:
Arnold; 2001:319-333
[n21.] Mitchell E, Krous HF, Donald T, Byard RW. An analysis of the usefulness of specific stages in the pathologic
investigation of sudden infant death. Am J Forensic Med Pathol. 2000;21:395-400
[n22.] Arnestad M, Vege A, Rognum TO. Evaluation of diagnostic tools applied in the examination of sudden unexpected deaths in infancy and early childhood. Forensic Sci Int. 2002;125:262-268
[n23.] Berry J, Allibone E, McKeever P, Moore I, Wright C, Fleming P. The pathology study: the contribution of ancillary pathology tests to the investigation of unexpected infant death. In: Fleming P, Blair P, Bacon C, Berry J, eds. Sudden Unexpected Deaths in Infancy: The CESDI SUDI Studies: 1993-1996. London, UK: The Stationery Office;
2000:97-112
Copyright © 2002 American Academy of Pediatrics
Pediatrics 2002; 109: 274-283
February, 2002
LENGTH: 8010 words
TITLE: National Underascertainment of Sudden Unexpected Infant Deaths Associated With Deaths of Unknown
Cause
AUTHOR: Mary D. Overpeck, DrPH <1> <2>; Ruth A. Brenner, MD, MPH <2>; Candace Cosgrove, MPH <2>; Ann
C. Trumble, PhD <2>; Kenneth Kochanek, MA; and Marian MacDorman, PhD
ABSTRACT. Objective. To investigate underascer-tainment of unexpected infant deaths at the national level as a result
of probable classification as attributable to unknown cause.
Page 8
Methods. Using linked birth and death certificates for all US birth cohorts from 1983-1991 and 1995-1996, we
identified 53 470 sudden infant death syndrome (SIDS) fatalities, 9071 unintentional injury deaths, 3473 injury deaths
classified with intentional or suspicious intent, and 8097 deaths with unknown underlying cause. For these deaths, we
compared relative risks (RRs) for maternal and infant variables available on birth certificates known to be predictive of
SIDS, unintentional injury, and homicides. Variables available on death certificates were compared for unlinked and
linked records. Factors related to state and national management of cases pending final cause determination are reviewed.
Results. For deaths from unknown cause, rates were consistently high among the same risk groups that have been
shown to be at increased risk for SIDS, unintentional injury, and homicides. For most risk factors, RRs for deaths attributable to unknown causes were somewhat lower than for RRs for intentional/suspicious injury deaths but higher than
for SIDS or unintentional injury, indicating combined contributions from all causes. For example, age at death from
unknown cause includes RRs that more strongly resemble patterns of intentional/suspicious injuries than SIDS or unintentional injury. Deaths from unknown cause were more likely to occur during the first week of life for unattended
births occurring outside clinical settings or when birth certificates were not found, similar to intentional/suspicious injury deaths.
Conclusions. Risk profiles indicate that deaths of unknown cause are likely to represent a mixture of unexpected
deaths. The process for determination of cause of unexpected death affects national underascertainment of SIDS and
injury deaths. Better coordination among child fatality review teams and local, state, and national officials should reduce underascertainment and improve documentation of circumstances surrounding deaths for prevention efforts.
[SIDS, infant injury fatalities, infant homicide, underascertainment, risk factors, child fatality reviews.]
TEXT:
ABBREVIATIONS. AAP, American Academy of Pediatrics; SIDS, sudden infant death syndrome; NCHS, National Center for Health Statistics; ICD-9, International Classification of Diseases, Ninth Revision; RR, relative risk.
The American Academy of Pediatrics (AAP) has called for improved comprehensive death investigation of sudden,
unexpected deaths to provide proper death certification for children. [n1] They emphasize the continuing need for
careful, timely review of deaths attributable to sudden infant death syndrome (SIDS) and trauma through appropriately
constituted review teams. Many states or local jurisdictions are performing in-depth multiple record reviews. [n2, n3]
State studies based on these reviews have documented underascertainment of fatal child abuse and neglect in young
children as reported by medical examiners, ranging from 60% to 100%. [n4-n6] Similar levels of underascertainment
have been projected to the national level. [n4, n7]
These state studies tracked possible sources of underascertainment by reviews of vital statistics, medical examiner
or coroner records, law enforcement files, and social service registries. Discrepancies among state-level data sources
may include 1) inability to report, such as when there is disagreement on designating the death as a homicide; 2) failure
of involved agencies to categorize or code information properly; and/or 3) inadequate gathering of case information.
[n6] Intensified state efforts to complete these reviews have resulted in more interagency coordination and time spent on
final determination of cause of death for most sudden, unexpected child fatalities. However, these efforts may have delayed reporting to state vital statistics offices on final determination of pending cases.
States provide annual statistical files to the National Center for Health Statistics (NCHS) based on death certificates to produce annual national mortality rates by cause and intent. [n8, n9] The need for a timely national file means
that some deaths that are under investigation at the time the death certificate is filed are marked pending in state files
submitted for national vital statistics. NCHS requests that amendments to show the cause in these pending records be
filed, but if the case is not amended when the national file is closed, then these certificates are coded to the International
Classification of Diseases, Ninth Revision (ICD-9) code 799 for cause unknown. [n9, n10] Therefore, state records may
include the final determination for reviewed cases that are shown to be attributable to unknown cause in national files.
State studies of factors associated with underascertainment have focused on reviews for child abuse and neglect. However, SIDS and traumatic injury cases all are recommended for examination by child fatality review teams as unexpected causes of death.
Since 1983, SIDS rates have dropped almost 50%, [n11, n12] with the majority of the decline probably attributable
to changes in sleep position and increased use of nonprone sleep positions after AAP recommendations in the early
1990s. [n13] Classification of possible SIDS cases to unknown cause may have increased with more death scene inves-
Page 9
tigations or fatality reviews that use more precise definitions of SIDS. [n13] Combined with increased state activities to
review all sudden, unexpected child fatalities, increased numbers of US deaths with unknown cause may result in national underascertainment of SIDS or unintentional, intentional, or undetermined intent injury classifications.
Risk factors for both natural and traumatic infant deaths are similar across most causes of infant death. [n11, n13n17] For unexpected deaths that require postmortem reviews, relative risks (RRs) are heightened for injury deaths classified as either intentional or undetermined intent compared with unintentional injury or SIDS. [n14-n19] Risk factors
consistently include maternal age or low education, lack of prenatal care, premature birth, and being a second or later
born infant. Also, males have higher risk of SIDS.
Underascertainment or misclassification of unexpected infant deaths reported through vital statistics systems may
occur for various reasons. Death of an infant who is born outside a clinical setting and/or without a trained attendant
may occur in association with a hidden pregnancy. Under these circumstances, birth certificates are less likely to be
filed or available for subsequent linkage and can be found only in total mortality files from death certificates, limiting
available risk factor information on these deaths.
Differing trends within causes of unexpected deaths and review of those deaths have occurred during the past 15
years. [n3, n13, n14] Before recent declines in SIDS rates concurrent with more systematic review based on improved
case definitions, death to unattended births and other deaths difficult to classify may also have been classified as SIDS.
[n13] Decreasing trends in SIDS deaths and unintentional injury deaths [n13, n14] and increases in rates of intentional
deaths should have affected the case mix of deaths of unknown cause if delays in reviews have resulted in more cases of
unknown cause.
This study is the first to investigate potential sources of national underascertainment of unexpected deaths. We analyzed whether infant deaths reported to NCHS with unknown cause have risk factor profiles that more closely resemble
SIDS, unintentional injury, or intentional injury deaths. Using a linked birth/infant death database, we compared rates
between 1983-1991 and 1995-1996 to assess changes in numbers, rates, and RRs of fatalities reported with unknown
cause that may be associated with underreporting of unexpected deaths. Unlinked mortality files are included to account
thoroughly for both total magnitude and differences for all unexpected deaths reported in annual mortality statistics by
cause but can be analyzed only for the limited risk factor information on the death certificate. The unlinked data are
more likely to include deaths of unknown cause for which no birth certificate was found, including for infants who were
born without trained attendants and found dead in the first few days of life. Trend analysis is included to compare relative distributions of risk factors, accounting for actual changes in cause-specific death rates and changes resulting from
effects of delayed reporting as a result of increased child fatality reviews.
METHODS
The US-linked birth/infant death data sets were analyzed for the 1983-1991 and 1995-1996 birth cohorts. Linked
files are not available for 1992-1994 birth cohorts. Methods of data linkage and file characteristics are reported elsewhere. [n20-n22] Approximately 98% of infant deaths are linked to corresponding birth certificates each year. Some
death certificates could not be linked because corresponding birth certificates were not found. In 1989, NCHS began
releasing individual record data on unlinked certificates attached to this data set. The final 1983-1997 NCHS mortality
file was used to show complete trends. [n11] The 1997 birth cohort linked file was not available during this analysis.
Variables
Causes of death are certified on death certificates by physicians, medical examiners, or coroners [n23] and are
coded according to specifications of the ICD-9 with underlying cause of death selected under rules established by the
World Health Organization. [n10] SIDS is defined as the sudden death of an infant that is unexpected by history and for
which a thorough postmortem examination fails to demonstrate an adequate cause of death (ICD code 798.0). SIDS is
considered natural, whereas injury deaths are termed traumatic. Traumatic causes describe mechanisms such as suffocation, strangulation, or blunt force from external forces (ICD E-codes).
For traumatic deaths, the certifiers assign official determination of unintentional, intentional, or undetermined intent. Criteria for assignment of undetermined intent specify that it be assigned only when, after a thorough investigation,
it cannot be determined whether the injuries are accidental or homicidal. [n10, n23] Studies of deaths classified as undetermined show that they are often considered suspicious because circumstances are similar to intentional deaths while
intent will remain unclear. [n24, n25] They include approximately 4% of all injury deaths in each year of this study.
Page 10
Intentional and undetermined intent injury deaths are combined in this analysis as intentional/suspicious like other studies to address similar underlying issues and risk factors. [n7, n18, n19, n26, n27]
Unintentional injury deaths are classified under codes E800 to E949 (excluding medical misadventures); intentional
injury deaths are classified under codes E960 to E969 (homicide and injury purposely inflicted by other persons). Undetermined intent deaths are classified under codes E980 to E989. Deaths from unknown cause are classified by code 799,
including both natural and traumatic deaths but specifically excluding all deaths that can be attributed to extreme prematurity, fetal growth retardation, congenital malformations, and SIDS. When final determination of cause remains unknown, cases are classified with code 799 or left blank. Pending cases that have not been amended by the time the national data file is closed at NCHS are converted to code 799.
Risk Factors
Factors were limited to certificate variables previously identified for SIDS, injury, and homicide deaths: mother's
age, race, education, and marital status; trimester of pregnancy that prenatal care began; and gestational age, gender, and
live birth order. [n13-n19] Missing data on fathers for fatalities (eg, missing age = 33%, race = 30% in 1983-1991) required omission of father's risk factors except for whether such data were present. We were unable to examine Hispanic
origin because before 1988, the majority of states did not report the variable. [n28] Gestational age and birth weight are
highly correlated, so only gestational age at birth was kept because of implications for prenatal timing of interventions.
Birth in a hospital, clinic/office/birthing center, place of residence, or somewhere else and with a trained birth attendant
(doctor, nurse midwife, or other midwife) are included for 1989-1991 and 1995-1996, but this detail level is unavailable
from all states before 1989.
Statistical Analysis
The number of deaths and live births represent complete counts that are not subject to sampling error. [n21] Data
are grouped for the 1983-1991 period to provide stable numbers for rates and RRs because annual random variations in
subcategories with small numbers may produce wide variations in RRs that are not meaningful. Data for 1995-1996 are
included to address recent changes but may include some unstable risk categories as a result of small numbers, as does
the unlinked data in 1989-1991 and 1995-1996. Therefore, data for unlinked files show percentages only. Denominators
for rates are per 100 000 live births. Significance tests are based on t tests for trend and Poisson tests of differences. RRs
are based on the ratio of the death rate in higher risk categories of a factor compared with the safest category as shown.
RESULTS
Trends
SIDS rates were higher than trauma or deaths from unknown causes, even after decreasing almost 50% from
146/100 000 to 77/100 000 between 1983 and 1997 (Fig 1). SIDS rates in 1995-1997 are approximately 4 times higher
than unintentional injury or unknown cause rates and 9 times higher than intentional/suspicious rates. Total death rates
from unintentional injury and unknown cause remained in a similar range (approximately 20/100 000 live births). Unintentional injury rates decreased approximately 25% (from 25/100 000 in 1983 to 19/100 000 in 1997), although rates for
some causes have increased (eg, mechanical suffocation; data not shown). [n11] Between 1983 and 1995, deaths classified with unknown causes increased almost 50%, with most of the increase occurring in the late 1980s. Rates ranged
from 16/100 000 to 24/100 000 from 1983-1995 (trend test, RHO = .01) and stabilized at 19/100 000 in 1996-1997.
Intentional injuries increased > 55% (from 5/100 000 to 8/100 000; trend test, RHO = .001), whereas suspicious injury deaths (classified as undetermined intent and representing approximately 4% of all injury fatalities) did not change.
Combined, suspicious and intentional injury deaths increased 41% overall (from 7/100 000 to 10/100 000; trend test,
RHO < .001).
Risk Factors
Mortality rates by cause, restricted to linked fatalities only, are highest for expected risk factor categories, eg, low
maternal education, maternal age < 17 or 17 to 19 years, single mother, no/late prenatal care, gestational age < 28
weeks, higher birth order, and being black or American Indian (Table 1). These categories tend to have the lowest proportion of births within each factor. Boys have higher SIDS rates than girls when compared with other causes, consistent with known SIDS risk factors. Rates tend to be high when information on a category within a risk factor from the
Page 11
birth certificate, such as timing of prenatal care or gestational age, is reported as unknown. Rates are higher when there
is no information about the father.
TABLE 1. Percentage of Births and Infant Mortality Rates From SIDS, Injuries,
and Unknown Causes by High-Risk Factors, US-Linked Birth/Infant Death Files
for 1983-1991 and 1995-1996
Factors
Percentage of
All US Births
Total number n1
Mortality rate
Mother's education n2
</= 11 y
12 y
13-15 y
>/= 16 y
Unknown
Mother's age
< 17 y
17-19 y
20-24 y
>/= 25 y
Marital status
Unmarried
Married
Any information on father
No
Yes
First prenatal visit
None
7-9 mo
4-6 mo
1-3 mo
Unknown
Gestational age at birth
< 28 wk
28-36 wk
>/= 37 wk
Unknown
Birth order
Third or more
Second child
First child
Unknown
Gender
Male
Female
Race
Black
American Indian
Asian
White
1983-1991
34 895 012
--
1995-1996
7 797 886
--
n2
n2
n2
n2
n2
22.2
33.2
21.7
21.4
1.5
2.4
10.4
28.5
58.6
2.6
10.4
24.5
62.5
24.7
75.3
32.3
67.7
12.7
87.3
13.7
86.3
1.8
4.0
17.6
74.3
2.2
1.2
2.9
14.0
79.5
2.5
0.7
9.1
87.0
3.1
0.7
10.2
88.1
1.0
25.8
32.5
41.1
0.5
26.2
32.0
41.1
0.7
51.2
48.8
51.2
48.8
16.0
0.9
3.1
79.9
15.4
1.0
4.2
79.5
Page 12
TABLE 1. Percentage of Births and Infant Mortality Rates From SIDS, Injuries,
and Unknown Causes by High-Risk Factors, US-Linked Birth/Infant Death Files
for 1983-1991 and 1995-1996
Factors
Mortality Rate (Per 100 000 Live Births)
SIDS
1983-1991
1995-1996
Total number n1
47 254
6216
Mortality rate
135.4
79.7
Mother's education n2
</= 11 y
n2
144.0
12 y
n2
85.5
13-15 y
n2
55.3
>/= 16 y
n2
26.8
Unknown
n2
110.4
Mother's age
< 17 y
272.9
173.5
17-19 y
246.3
147.0
20-24 y
180.1
116.9
>/= 25 y
88.3
50.0
Marital status
Unmarried
241.8
143.0
Married
100.5
49.6
Any information on father
No
260.7
177.0
Yes
117.2
64.3
First prenatal visit
None
373.6
284.5
7-9 mo
260.5
140.5
4-6 mo
205.8
130.0
1-3 mo
104.5
64.50
Unknown
189.8
116.4
Gestational age at birth
< 28 wk
273.5
163.7
28-36 wk
271.8
161.5
>/= 37 wk
117.1
69.1
Unknown
213.6
121.8
Birth order
Third or more
185.6
110.6
Second child
143.5
82.1
First child
97.8
58.0
Unknown
116.0
89.8
Gender
Male
160.3
93.2
Female
109.3
65.5
Race
Black
226.2
154.1
American Indian
326.6
180.9
Asian
90.1
45.1
White
116.7
65.9
TABLE 1. Percentage of Births and Infant Mortality Rates From SIDS, Injuries,
and Unknown Causes by High-Risk Factors, US-Linked Birth/Infant Death Files
Unintentional
Injury
1983-1991
7594
21.8
1995-1996
1477
18.9
n2
n2
n2
n2
n2
33.5
20.7
14.2
6.3
15.6
43.9
40.4
29.1
14.0
31.7
35.4
28.5
11.9
38.9
16.1
32.2
12.6
45.3
18.3
41.8
15.3
69.6
43.7
33.0
16.5
29.0
68.7
37.7
31.1
15.0
31.0
51.6
36.3
19.5
36.0
38.2
33.8
17.0
24.6
30.8
21.7
16.0
30.9
27.3
17.7
14.6
17.6
23.4
20.0
21.1
16.7
38.9
61.6
12.4
18.2
32.1
47.9
6.7
16.7
Page 13
for 1983-1991 and 1995-1996
Factors
Mortality Rate (Per 100 000 Live Births)
Intentional/Suspicious
Injury
Total number n1
Mortality rate
Mother's education n2
</= 11 y
12 y
13-15 y
>/= 16 y
Unknown
Mother's age
< 17 y
17-19 y
20-24 y
>/= 25 y
Marital status
Unmarried
Married
Any information on father
No
Yes
First prenatal visit
None
7-9 mo
4-6 mo
1-3 mo
Unknown
Gestational age at birth
< 28 wk
28-36 wk
>/= 37 wk
Unknown
Birth order
Third or more
Second child
First child
Unknown
Gender
Male
Female
Race
Black
American Indian
Asian
White
Unknown
Cause
1983-1991
2776
8.0
1995-1996
697
8.9
1983-1991
6552
18.8
1995-1996
1545
19.8
n2
n2
n2
n2
n2
16.5
10.7
4.4
1.6
30.4
n2
n2
n2
n2
n2
34.2
22.3
13.3
7.0
27.8
22.7
18.8
10.7
4.1
22.4
21.5
14.2
4.2
41.1
34.5
23.6
12.7
36.1
35.3
25.7
14.2
18.8
4.4
18.4
4.4
40.3
11.7
36.3
12.0
22.5
5.8
24.0
6.6
45.3
14.9
46.7
15.5
46.1
16.4
12.1
5.1
23.2
60.0
16.6
15.3
6.5
17.6
95.4
36.4
26.8
13.0
54.2
104.7
39.1
31.6
15.4
33.6
23.2
14.4
6.5
25.6
21.8
17.0
7.6
40.2
174.7
43.2
14.1
39.3
196.5
42.0
15.5
49.2
8.7
7.2
7.4
53.2
9.3
8.4
8.6
39.0
26.4
16.9
14.7
74.0
30.2
17.3
15.1
21.5
8.4
7.5
9.2
8.7
21.5
15.9
22.9
16.5
19.8
20.3
3.8
5.6
19.0
18.6
4.6
7.1
44.4
32.0
16.7
13.6
40.8
39.9
12.9
15.9
n1 Excludes deaths for which no birth certificates could be found for linkage.
Page 14
n2 Data not shown because all states did not consistently report maternal education during 1983-1991 period resulting in noncomparability of the percentage distributions and mortality rates. All states reported in 1995-1996 for which data are shown.
RRs
The RR of death for high-risk infants within each variable is greatest among intentional/suspicious injury deaths
compared with other causes, except for gestational age < 28 weeks, high birth order, and male gender (Table 2). Boys
are at 50% greater risk of SIDS compared with girls; 40% greater for unknown causes; and approximately 10% to 20%
greater for unintentional and intentional/suspicious injuries. Deaths of unknown cause have RRs similar to SIDS and
unintentional injury deaths for many high-risk categories compared with the reference categories shown on Table 2, but
RRs for unknown cause and intentional/suspicious injuries are higher when mothers are unmarried, mothers received no
prenatal care, gestational age is < 28 weeks or unknown, race is black, or information on father is unknown.
TABLE 2. RRs for High-Risk Factors for SIDS, Injury Deaths, and Deaths of
Unknown Cause, US-Linked Birth/Infant Death Files for 1983-1991 and 1995-1996
Factors
RR of Death
SIDS
Total number n1
Total rate
Mother's education (reference
value >/= 16 y)
</= 11 y
12 y
13-15 y
Unknown
Mother's age (reference value
>/= 25 y)
</= 17 y
17-19 y
20-24 y
Mother's marital status
(reference is being
married)
Unmarried
Any information on father
(reference is yes)
No
First prenatal visit (reference
value = 1-3 mo)
None
7-9 mo
4-6 mo
Unknown
Gestational age at birth
(reference value >/= 37 wk)
< 28 wk
28-36 wk
Unknown
Birth order (reference is first
child)
Third or more
Second child
Unintentional injury
1983-1991
47 254
135.4
1995-1996
6216
79.7
1983-1991
7594
21.8
1995-1996
1477
18.9
n2
n2
n2
n2
5.4
3.2
2.1
4.1
n2
n2
n2
n2
5.3
3.3
2.3
2.5
3.1
2.8
2.0
3.5
2.9
2.3
3.1
2.9
2.1
2.7
3.0
2.4
2.4
2.6
2.4
2.6
2.2
2.7
2.5
2.7
3.6
2.5
2.0
1.8
4.4
2.2
2.0
1.8
4.2
2.6
2.0
1.8
4.6
2.5
2.1
2.1
2.2
2.3
1.8
2.4
2.3
1.8
2.6
1.9
1.8
2.2
2.0
1.4
1.9
1.9
1.9
1.4
1.9
1.4
1.9
1.2
Page 15
TABLE 2. RRs for High-Risk Factors for SIDS, Injury Deaths, and Deaths of
Unknown Cause, US-Linked Birth/Infant Death Files for 1983-1991 and 1995-1996
Factors
RR of Death
SIDS
Unintentional injury
1983-1991
1995-1996
Unknown
1.2
1.5
Gender (reference is female)
Male
1.5
1.5
Race (reference is white)
Black
1.9
2.3
American Indian
2.8
2.7
Asian
0.8
0.7
TABLE 2. RRs for High-Risk Factors for SIDS, Injury Deaths, and Deaths of
Unknown Cause, US-Linked Birth/Infant Death Files for 1983-1991 and 1995-1996
Factors
1995-1996
1.2
1.2
1.4
2.1
3.4
0.7
1.9
2.9
0.4
RR of Death
Intentional/Suspicious
Injury
Total number n1
Total rate
Mother's education (reference
value >/= 16 y)
</= 11 y
12 y
13-15 y
Unknown
Mother's age (reference value
>/= 25 y)
</= 17 y
17-19 y
20-24 y
Mother's marital status
(reference is being
married)
Unmarried
Any information on father
(reference is yes)
No
First prenatal visit (reference
value = 1-3 mo)
None
7-9 mo
4-6 mo
Unknown
Gestational age at birth
(reference value >/= 37 wk)
< 28 wk
28-36 wk
Unknown
1983-1991
1.9
Unknown Cause
1983-1991
2776
8.0
1995-1996
697
8.9
1983-1991
6552
18.8
1995-1996
1545
19.8
n2
n2
n2
n2
10.2
6.6
2.7
18.8
n2
n2
n2
n2
4.9
3.2
1.9
4.0
5.5
4.6
2.6
5.3
5.1
3.4
3.2
2.7
1.8
2.5
2.5
1.8
4.3
4.2
3.4
3.0
3.8
3.7
3.0
3.0
9.0
3.2
2.1
4.5
9.2
2.6
2.4
2.7
7.4
2.8
2.1
4.2
6.8
2.5
2.1
2.2
3.6
2.2
3.9
2.9
2.3
5.3
12.4
3.1
2.8
12.7
2.7
3.2
Page 16
TABLE 2. RRs for High-Risk Factors for SIDS, Injury Deaths, and Deaths of
Unknown Cause, US-Linked Birth/Infant Death Files for 1983-1991 and 1995-1996
Factors
RR of Death
SIDS
1983-1991
Birth order (reference is first
child)
Third or more
Second child
Unknown
Gender (reference is female)
Male
Race (reference is white)
Black
American Indian
Asian
Unintentional injury
1995-1996
1983-1991
1995-1996
1.2
1.0
7.2
1.1
1.0
4.5
1.8
1.2
5.0
2.0
1.2
1.4
1.1
1.2
1.4
1.4
3.6
3.6
0.7
2.7
2.6
0.6
3.3
2.4
1.2
2.6
2.5
0.8
n1 Excludes births or deaths for which no certificates could be found for linkage.
n2 Data not shown because all states did not consistently report maternal education during 1983-1991 period, resulting in noncomparability of the percentage distributions and mortality rates. All states reported in 1995-1996, for which data are shown.
Delivery Circumstances
Approximately 99% of all births occurred in clinical settings and were attended by trained birth attendants (doctors,
nurse-midwives, or other midwives; Table 3). The percentage of deaths to births that did not occur in clinical settings or
with a trained attendant cannot be completely ascertained because some death certificates do not have a corresponding
birth certificate. However, on the basis of linked certificates, being born at a place of residence or an unknown place and
with no trained birth attendant (or attendant unknown) were associated with higher mortality rates. A higher proportion
of deaths from intentional/suspicious injury or with unknown cause occurred for births delivered in residences or other
or unknown places or when no trained birth attendant was noted. SIDS rates decreased among all categories of birth
place and attendant factors between the 2 time periods; rates for deaths of unknown cause decreased more when births
occurred in nonclinical settings or with no trained attendant than in clinical settings with trained attendants.
TABLE 3. Place of Delivery and Attendant Status for Sudden, Unexpected Deaths
and Age at Death for Unattended, Nonclinical Births, US-Linked Birth/Infant
Death Files for 1989-1991 and 1995-1996
Delivery Circumstances
Births n1
Total number
Percentage by birth place/attendant n2
Hospital/center/clinic/office
Trained attendant n2
Place of birth
Hospital/center/clinic/office
Residence/other/don't know
Nonclinical versus clinical (RR) n2
1989-1991
12 324 525
1995-1996
7 797 886
99.2
99.0
99.3
98.9
12 224 205
100 320
7 743 301
54 585
Page 17
TABLE 3. Place of Delivery and Attendant Status for Sudden, Unexpected Deaths
and Age at Death for Unattended, Nonclinical Births, US-Linked Birth/Infant
Death Files for 1989-1991 and 1995-1996
Delivery Circumstances
Births n1
1989-1991
Trained attendant n2
Attended n2
Unattended, don't know
Unattended versus attended (RR)
1995-1996
12 204 151
120 374
Unattended births in nonclinical settings by
age at death n2 (n)
<1d
1-6 d
7-28 d
1-4 mo
5-11 mo
TABLE 3. Place of Delivery and Attendant Status for Sudden, Unexpected Deaths
and Age at Death for Unattended, Nonclinical Births, US-Linked Birth/Infant
Death Files for 1989-1991 and 1995-1996
Delivery Circumstances
------
7 713 544
84 342
------
Deaths n1
SIDS
Total number
Percentage by birth place/attendant n2
Hospital/center/clinic/office
Trained attendant n2
Place of birth
Hospital/center/clinic/office
Residence/other/don't know
Nonclinical versus clinical (RR) n2
Trained attendant n2
Attended n2
Unattended, don't know
Unattended versus attended (RR)
1989-1991
16 027
Unattended births in nonclinical settings by
age at death n2 (n)
<1d
1-6 d
7-28 d
1-4 mo
5-11 mo
TABLE 3. Place of Delivery and Attendant Status for Sudden, Unexpected Deaths
and Age at Death for Unattended, Nonclinical Births, US-Linked Birth/Infant
Death Files for 1989-1991 and 1995-1996
Delivery Circumstances
1995-1996
6216
98.9
98.5
Deaths/100 000 Births
129.6
179.4
(1.4)
99.1
99.0
79.6
103.4
(1.3)
131.1
199.4
(1.5)
73.0
73.5
(1.0)
93
31
1.1
0
6.4
79.6
12.9
0
3.2
3.2
67.7
25.8
Deaths n1
Page 18
Unintentional Injury
Total number
Percentage by birth place/attendant n2
Hospital/center/clinic/office
Trained attendant n2
Place of birth
Hospital/center/clinic/office
Residence/other/don't know
Nonclinical versus clinical (RR) n2
Trained attendant n2
Attended n2
Unattended, don't know
Unattended versus attended (RR)
1989-1991
2587
1995-1996
1477
98.4
98.1
Deaths/100 000 Births
20.8
41.9
(2.0)
Unattended births in nonclinical settings by
age at death n2 (n)
<1d
1-6 d
7-28 d
1-4 mo
5-11 mo
TABLE 3. Place of Delivery and Attendant Status for Sudden, Unexpected Deaths
and Age at Death for Unattended, Nonclinical Births, US-Linked Birth/Infant
Death Files for 1989-1991 and 1995-1996
Delivery Circumstances
98.3
98.2
18.8
47.0
(2.5)
21.1
40.7
(1.9)
17.2
32.0
(1.9)
19
16
42.1
10.5
10.5
31.6
5.3
43.8
0
6.2
25.0
25.0
Deaths n1
Intentional/Suspicious
Injury
Total number
Percentage by birth place/attendant n2
Hospital/center/clinic/office
Trained attendant n2
Place of birth
Hospital/center/clinic/office
Residence/other/don't know
Nonclinical versus clinical (RR) n2
Trained attendant n2
Attended n2
Unattended, don't know
Unattended versus attended (RR)
1989-1991
1109
Unattended births in nonclinical settings by
age at death n2 (n)
<1d
1-6 d
7-28 d
1-4 mo
5-11 mo
TABLE 3. Place of Delivery and Attendant Status for Sudden, Unexpected Deaths
1995-1996
91.6
94.1
Deaths/100 000 Births
8.3
92.7
(11.2)
697
93.1
94.8
8.4
88.4
(10.5)
8.7
54.0
(6.2)
7.8
42.7
(5.5)
54
32
72.2
18.5
0
5.6
3.7
62.5
25.0
9.4
3.1
0
Page 19
and Age at Death for Unattended, Nonclinical Births, US-Linked Birth/Infant
Death Files for 1989-1991 and 1995-1996
Delivery Circumstances
Deaths n1
Unknown Cause
Total number
Percentage by birth place/attendant n2
Hospital/center/clinic/office
Trained attendant n2
Place of birth
Hospital/center/clinic/office
Residence/other/don't know
Nonclinical versus clinical (RR) n2
Trained attendant n2
Attended n2
Unattended, don't know
Unattended versus attended (RR)
Unattended births in nonclinical settings by
age at death n2 (n)
<1d
1-6 d
7-28 d
1-4 mo
5-11 mo
1989-1991
2545
1995-1996
1545
97.0
96.9
Deaths/100 000 Births
20.2
76.8
(3.8)
97.8
97.7
19.5
63.9
(3.3)
20.5
65.6
(3.2)
17.9
42.7
(2.4)
34
21
58.8
14.7
2.9
11.8
11.8
57.1
9.5
9.5
14.3
9.5
n1 Excludes deaths for which no birth certificates could be found for linkage.
n2 Trained attendant at birth was a doctor, certified nurse midwife, or other midwife; "unattended" includes "other," unknown, or attendant was not stated. "Other" may include emergency medical staff, family members, etc. Nonclinical settings include a residence, other,
or unknown.
RRs of intentional/suspicious injury deaths among infants who were born at a place of residence or unknown place
are 11 times higher than for infants who were born in a clinical setting (hospitals, birthing centers, doctor offices) and 3
to 4 times greater when cause is unknown. RRs for unattended intentional/suspicious injuries are approximately 6 times
greater than for attended births and approximately 2 to 3 times greater than when cause was unknown. RRs for unintentional injury when birth was unattended were 1.9 times attended. RRs for unattended SIDS births compared with attended births decreased to 1.0 in 1995-1996.
In linked certificate data, 0.7% of all deaths in both time periods were to infants who were born in non-clinical settings without trained attendants (801 in 1989-1991 and 414 in 1995-1996; data not shown). One fourth of these deaths
were from unexpected or unknown causes in both time periods. When limited to deaths of these infants, the age at death
differs among the unexpected or unknown causes. The majority of deaths happened during the first day or week of life
except for SIDS. Approximately 90% of the 86 intentional/suspicious injury deaths reported in both time periods in
linked files occurred during the first week, with almost three fourths of those during the first day. Approximately 70%
of the 55 deaths of unknown cause in both time periods occurred during the first week. Approximately half of the 35
unintentional injury deaths in the 2 time periods occurred during the first week, almost all during the first day. Infants
who died from unknown cause were more likely to have missing data on gestational age or a gestational age of < 28
weeks if the birth was unattended. Delivery circumstances are unavailable in unlinked files where cases of unknown
cause or injury deaths of undetermined intent may more likely to be found.
Unlinked Certificates
Page 20
The unlinked files represent approximately 2% of all infant deaths, while 2.3% of unexpected deaths are unlinked
(Table 4). Among unexpected deaths, no birth certificate was obtained for 4.0% of deaths of unknown cause, 5.0% of
intentional/suspicious injuries, 3.0% of unintentional injuries, and 1.5% of SIDS. Table 4 compares age at death in
linked certificates with deaths in unlinked files during 1989-1991 and 1995-1996. Except for SIDS in 1989-1991, total
numbers of unlinked certificates for each cause were very small, making comparisons between the time periods potentially unstable. Proportions of deaths during the first week of life are approximately 4 times greater in the unlinked file
than in the linked file. Deaths during the first day of life or before the first week are more likely to be missing birth certificates for intentional/suspicious injury deaths, deaths from unknown cause, and unintentional injury deaths than are
SIDS. The majority of intentional/suspicious injury deaths occurred during the first week of life among unlinked certificates compared with 6% to 7% for linked certificates. Among unlinked deaths from unknown causes, 43% and 48% of
deaths occurred during the first week of life in 1989-1991 and 1995-1996, respectively, compared with 11% to 12% of
linked certificates. SIDS is highly unlikely during the first week or month of life. Neither the linked nor the unlinked
deaths from unknown cause were as likely to be reported during the 1- to 4-month time period as was SIDS.
TABLE 4. Deaths From SIDS, Injuries, and Unknown Causes by Age at Death for
Linked and Unlinked Certificates, 1989-1991 and 1995-1996
Factors
SIDS
Linked
Total deaths, 1989-1991
Death rate n1
Age at death
Unlinked
Linked
Deaths in 1989-1991
16 027
130.0
<1d
1-6 d
7-28 d
1-4 mo
5-11 mo
0.2
0.6
5.6
77.7
15.9
Total deaths, 1995-1996
Death rate n1
Age at death
Unintentional Injury
6216
79.7
245
2587
2.0
21.0
Percent of Deaths, 1989-1991
0.8
0.4
3.7
74.7
20.4
Deaths in 1995-1996
18.5
3.7
3.7
29.6
44.4
83
1477
1.1
18.9
Percent of Deaths, 1995-1996
37
0.5
Intentional/Suspicious
Injury
Linked
Total deaths, 1989-1991
Death rate n1
Age at death
<1d
1.4
2.4
5.4
39.7
51.1
6.2
13.5
2.7
0
29.7
54.1
Unknown Cause
Unlinked
Linked
Deaths in 1989-1991
1109
9.0
81
0.7
1.4
1.5
4.6
38.4
54.1
<1d
0.1
0
1-6 d
0.7
0
7-28 d
6.0
4.8
1-4 mo
75.5
67.5
5-11 mo
17.7
27.7
TABLE 4. Deaths From SIDS, Injuries, and Unknown Causes by Age at Death for
Linked and Unlinked Certificates, 1989-1991 and 1995-1996
Factors
Unlinked
Unlinked
58
2545
0.5
20.7
Percent of Deaths, 1989-1991
51.7
7.7
106
0.9
36.8
Page 21
TABLE 4. Deaths From SIDS, Injuries, and Unknown Causes by Age at Death for
Linked and Unlinked Certificates, 1989-1991 and 1995-1996
Factors
SIDS
Linked
1-6 d
7-28 d
1-4 mo
5-11 mo
Total deaths, 1995-1996
Death rate n1
Age at death
<1d
1-6 d
7-28 d
1-4 mo
5-11 mo
Unintentional Injury
Unlinked
1.6
4.0
45.0
43.2
697
8.9
Linked
8.6
1.7
19.0
20.5
Deaths in 1995-1996
Unlinked
4.1
8.7
58.5
21.0
6.6
3.8
33.0
19.8
46
1545
0.6
19.8
Percent of Deaths, 1995-1996
68
0.9
4.7
1.4
4.7
47.9
41.2
45.7
17.4
0
13.0
23.9
8.5
4.1
10.7
54.4
22.3
23.5
14.7
5.9
30.9
25.0
n1 Rate per 100 000 live births. Rates based on small numbers may be unstable.
In unlinked files, deaths are more likely to occur during the first day or week for traumatic deaths and those of unknown cause compared with SIDS (Table 4). The proportions who die from traumatic and unknown causes in the first
day and week among infants who were born with no trained attendants in nonclinical settings in the linked file (Table 3)
are higher but show a pattern similar to unlinked deaths by cause. It seems that many infants for whom no birth certificate was found may have been born with no trained attendants and in nonclinical settings. This is reinforced by the relative proportions of deaths in the first day or week between unlinked and linked files: within the unlinked file, proportions of deaths during the first day are 5 to 14 times greater than the distribution in the linked file among the 4 causes
shown (Table 4). Proportions of deaths during 1 to 6 days in the unlinked file are 2 to 7 times greater than in the linked
file, except for SIDS.
Other Information on Deaths of Unknown Cause
We reviewed additional/contributing conditions from multiple cause of death files for cases with underlying cause
shown as ICD code 799 (data not shown). Fewer than 10% of the records had any additional conditions mentioned. Of
these, the most frequent additional conditions were code 799.0, asphyxia (excluding newborn code 768 for intrauterine
hypoxia or birth asphyxia); code 799.1, cardiorespiratory failure (excluding newborn syndrome codes 769 and 770.8);
code 798.0, SIDS; and codes 765.0/765.1 for extreme prematurity, other prematurity, or small for size (excluding code
764, slow fetal growth/fetal malnutrition).
Performance of an autopsy may indicate that all possible diagnoses such as SIDS or shaken infant syndrome were
carefully examined before assigning cause and/or designating cause as unknown, with possible delays in pending certificates at both the state and national levels. Since 1994, national data have not shown whether an autopsy was performed.
On the basis of linked file data, the proportion of unexpected deaths for which an autopsy was performed seems to have
increased between 1983 and 1991 (data not shown). In 1989-1991, 98% of intentional/suspicious injury deaths had an
autopsy performed as did 94% of SIDS, 87% of unknown cause deaths, and 72% of unintentional injury deaths.
DISCUSSION
Our ability to accurately track the decline in SIDS rates or changes in traumatic death rates at the national level is
diminished by the magnitude of deaths with unknown cause. The case mix of > 8000 deaths assigned to unknown cause
during the 11 years reviewed for this study is not available to reallocate these deaths to SIDS, traumatic deaths, or any
Page 22
other causes at the national level. However, death certification requirements and comparisons of risk profiles suggest
contributions from all causes of unexpected death.
Distributions of age at death probably are the best indicators of the case mix of unexpected deaths allocated to unknown cause because the patterns are relatively distinctive and usually associated with developmental stages of the infant. [n14, n18, n29] Unintentional injury deaths are infrequent in the first month of life, with the majority occurring at
>/= 5 months of age. [n14] Intentional and suspicious deaths are most likely to occur during the first week and month,
whereas SIDS occurs most frequently between 1 and 4 months. With decreases in SIDS associated with sleep position
in recent years, the AAP suggests that the proportion of reported SIDS cases that actually are attributable to infanticide
may be increasing. [n29] Our data showed that SIDS deaths in the first week of life declined very slightly between 1989
and 1991 and 1995 and 1996 (to none in the unlinked file in the latter period), with a shift of deaths at >/= 5 months to
higher proportions found primarily in the unlinked file without birth certificates. Approximately 75% of SIDS occurred
between 1 and 4 months in 1995-1996, and approximately 18% occurred between 5 and 11 months. Deaths from unknown cause that occurred between 1 and 4 months during 1995-1996, the expected timing of most SIDS, represented
54% and 31% in the linked and unlinked files, respectively. The proportions of the deaths classified with unknown
cause that occur in the first week of life (13% in the linked file and 38% in the unlinked file in 1995-1996) are inconsistent with expected timing of SIDS. [n29] On the basis of expected timing of SIDS and contributions of unintentional
injury deaths at 1 to 4 months or later, probably no more than half of the deaths of unknown causes might be SIDS cases.
Higher death rates for unknown causes since 1988 may be partly a result of closer scrutiny of possible SIDS cases,
including more autopsies, with fewer ambiguous cases assigned a SIDS designation. [n30] Conversely, because most of
the increase in the rates for deaths of unknown cause occurred before the dramatic decrease in SIDS rates after an intervention initiated in the early 1990s, [n13] only a minor portion of the decrease in SIDS could be attributable to reassignment to unknown causes.
The magnitude of the RRs of death from unknown cause for most risk factors tended to fall between levels for intentional/suspicious injury deaths and SIDS or unintentional injuries, suggesting influences from all sudden, unexpected
causes. RRs of death attributable to unknown cause were most similar to risks for intentional/suspicious injury when
mothers were unmarried, mothers received no prenatal care, gestational age is < 28 weeks or unknown, race is black, or
information on father is unknown. Particularly striking is the high proportion of deaths from unknown cause in the unlinked file that occurred during the first week of life, like the majority of unlinked intentional/suspicious injury deaths.
Although numbers are small in the unlinked file, the pattern is similar to deaths in the linked file for births delivered in
nonclinical settings without trained birth attendants. It is apparent that ignoring potential intentional or suspicious deaths
classified with either unknown cause or SIDS leads to an underestimation of the magnitude of the public health problem
of injury, abuse, neglect, or abandonment of infants.
Relation to Previous Studies
Our findings clarify and add information to results of earlier studies of underascertainment of child abuse or neglect
based on careful review of substantiated multiple record sources. [n2-n6, n31] Ewigman et al [n5] estimated that child
abuse and neglect may have been involved in 7% to 27% of injury deaths reported in 1983-1986 as unintentional and in
at least 5% of deaths classified as SIDS in Missouri. A retrospective California study of infant/toddler traumatic deaths
used final coroner reports to show that homicide estimates would increase 18% if cases that were classified as undetermined or left unspecified were examined more carefully. [n24] On the basis of state reviews in Missouri and North Carolina, 2 studies estimated underascertainment of deaths from child abuse and neglect between 60% and 100% at the
national level. [n4, n5] During 1983-1997, 4486 infants were officially reported as intentionally killed and 793 died
from injuries with undetermined but suspicious intent--a total of almost 1 a day. [n12] Because our national study shows
more than twice as many deaths classified with unknown cause compared with intentional/suspicious traumatic deaths,
approximately 30% to 50% of the deaths with unknown causes would have to result from child abuse and neglect (intentional or not) for our findings to concur with estimates by others of underascertainment of 60% to 100% at the national level. [n4-n6]
Potential Biases and Limitations
Assignment of unknown or intentional/suspicious causes by certifiers is possibly biased toward the higher risk factor profiles shown in this analysis. [n32, n33] Such profiles may increase investigations or cause the certifiers to question intent for lower socioeconomic groups more often than in cases that occur in middle- or upper-class families. Stud-
Page 23
ies that find this bias suggest the effect would be to understate fatalities attributable to intentional/suspicious injury or
from unknown causes rather than SIDS or unintentional injury. [n32, n33]
Another bias may be indicated by the relatively high proportion of infants who were born at < 28 weeks in cases of
unknown cause. These infants may actually be at higher risk of unexpected death among all causes [n13, n29] or could
include some cases in which infants are born prematurely and unexpectedly outside clinical settings and/or without preand postnatal care. This may result in ambiguity among certifiers about whether a small fetus was actually a live birth
with a subsequent death in the first few days of life and/or the underlying cause of death was prematurity. [n24] Our
data do not include adequate information on prematurity among unattended cases because of missing data or lack of
birth certificates. However, those cases with birth certificates showed that infants who died from unknown cause were
more likely to have missing data on gestational age or a gestational age of < 28 weeks if the birth was unattended.
Other limitations result from reliance on birth and death certificates. Factors identified from certificates do not describe circumstances of death that might be available from child fatality reviews. For example, death certificates are
not intended to provide relationship of perpetrators in cases of abuse or homicide. Most birth certificate information is
available only for mothers, yet state studies and police data show a majority of male perpetrators after the first week of
life, frequently a boyfriend or a relative who is caring for a child. [n34-n37] Better information on family structure or
social support is needed. [n32] These social measures may be markers for lack of readily available caregiving and parenting resources, but we also need to describe interactions among family structure, social support, alcohol, drug abuse,
and other contributors.
Future Directions
Addition of information on national files about whether determination of cause is pending or an autopsy was done
would improve assessment of changing trends of SIDS cases or deaths of unknown cause. An autopsy indicates that the
death was, indeed, unexpected and recommended procedures for ascertaining cause of death were initiated. Autopsies
may also be an indication that cases were under investigation when the certificate was originally filed, with final determination of cause and/or intent still pending. Child fatality review of unexpected deaths should improve the quality of
certificate data sources. Physicians, medical examiners, and coroners should be encouraged to file amended death certificates promptly in cases in which cause of death was initially unknown or was changed after further investigation.
States should query pending cases in a timely manner. National files can be corrected only if states amend the record
before the data year is closed for final processing. Reduced national and state underascertainment of specific causes of
death for infants and children will result from more complete and timely information from local levels.
Wider dissemination of information about fatality circumstances from child death review investigations should also
facilitate better interventions. [n18, n29] One interagency effort is under way to address the interplay among medical,
criminal, and child protective concerns of review teams while expanding and supporting a network of review teams
among states. [n38] Efforts include guidelines for development and maintenance of teams with standard procedures.
The AAP and others propose continual functioning of multiagency review teams to accelerate progress in understanding SIDS, reduce the number of fatal cases of child abuse and neglect, increase the awareness of familial genetic
diseases, focus attention on public health threats, and detect and remediate inadequate medical care. [n1, n29, n39]
States, medical providers, social services agencies, and child fatality review teams need better support to organize improved information systems to facilitate prevention programs with appropriately targeted interventions. [n2, n18, n29]
As stated in AAP recommendations on investigation and review of unexpected infant and child deaths, lack of adequate
investigations allows flawed systems to continue and are an impediment to preventing illness, injury, and the deaths of
other children at risk. [n1]
SUPPLEMENTARY INFORMATION: Received for publication Mar 22, 2001; accepted Sep 14, 2001.
From the <1> Maternal and Child Health Bureau, Health Resources and Services Administration, Rockville, Maryland;
and <2> Division of Epidemiology, Statistics, and Prevention Research, National Institutes of Child Health and Human
Development, Bethesda, Maryland.
Reprint requests to (M.D.O.) Maternal and Child Health Bureau, US Health Resources and Services Administration, 5600 Fishers Ln, Room 18-41, Rockville, MD 20857. E-mail: overpecm@hrsa.gov
REFERENCES:
Page 24
[n1.] American Academy of Pediatrics, Committee on Child Abuse and Neglect and Committee on Community Health
Services. Investigation and review of unexpected infant and child deaths. Pediatrics. 1999;104:1158-1160
[n2.] Durfee MJ, Gellert GA, Tilton-Durfee D. Origins and clinical relevance of child death review teams. JAMA.
1992;267:3172-3175
[n3.] American Academy of Pediatrics, Division of State Government Affairs. 1999 State Legislation Report. Elk Grove
Village, IL: American Academy of Pediatrics; 1999
[n4.] Herman-Giddens ME, Brown G, Verbiest S, et al. Underascertainment of child abuse mortality in the United
States. JAMA. 1999;282:463-467
[n5.] Ewigman B, Kivlahan C, Land G. The Missouri child fatality study: underreporting of maltreatment fatalities
among children younger than five years of age, 1983 through 1986. Pediatrics. 1986;91:330-337
[n6.] California Department of Justice State Child Death Review Board. Child Deaths in California, 1992-1995. Sacramento: CA: California Department of Justice; 1997
[n7.] McClain PW, Sacks JJ, Froehlke RG, Ewigman BG. Estimates of fatal child abuse and neglect, United States,
1979-1988. Pediatrics. 1993;91:338-343
[n8.] Rosenberg HM, Kochanek KD. The death certificate as a source of injury data. In: Proceedings of the International Collaboration Effort on Injury Statistics, I. Hyattsville, MD: National Center for Health Statistics; 1995;8:1-17.
DHHS Publ. No. (PHS) 95-1252
[n9.] National Center for Health Statistics. Technical appendix. In: Vital Statistics of the United States, 1992, H, Mortality, Part A. Washington, DC: Public Health Service; 1996
[n10.] World Health Organization. World Health Classification: Manual of the International Statistical Classification of
Diseases, Injuries, and Causes of Death, Ninth Revision. Geneva, Switzerland: World Health Organization; 1977
[n11.] Hoyert DL, Kochanek KD, Murphy SL. Deaths: United States, Final Data for 1997. National Vital Statistics Report. Hyattsville, MD: National Center for Health Statistics; 1999;47(No. 19):88
[n12.] National Center for Health Statistics. Mortality Public Use Data Files and Documentation, 1983-97. Hyattsville,
MD: Public Health Service; 1997
[n13.] Willinger M, Hoffman HJ, Wu KT, et al. Factors associated with the transition to nonprone sleep positions of
infants in the United States. JAMA. 1998;280:329-335
[n14.] Brenner RA, Overpeck MD, Trumble AC, DerSimonian R, Berendes H. Deaths due to injuries in infants, United
States, 1983-1991. Pediatrics. 1999;103:968-971
[n15.] Scholer SJ, Hickson GB, Ray WA. Sociodemographic factors identify US infants at high risk of injury morbidity.
Pediatrics. 1999;103:1183-1188
[n16.] Touhy PG, Counsell AM, Geddis DC. Sociodemographic factors associated with sleeping position and location.
Arch Dis Child. 1993;69:664-666
[n17.] Ponsonby AL, Dwyer T, Kasl SV, Couper D, Cochrane JA. Correlates of prone infant sleeping position by period
of birth. Arch Dis Child. 1994;72:204-208
[n18.] Overpeck MD, Brenner RA, Trumble AC, Trifiletti MA, Berendes HW. Risk factors for infant homicide in the U.
S. N Engl J Med. 1998;339:1211-1216
Page 25
[n19.] Overpeck MD, Brenner RA, Trumble AC, Smith GS, MacDorman MF, Berendes HW. Infant injury deaths with
unknown intent: what else do we know? Inj Prev. 1999;5:272-275
[n20.] National Center for Health Statistics. Birth Cohort Linked Birth/Infant Death Data Sets: Public Use Data File
Documentation for 1983 to 1991. Hyattsville, MD: US Public Health Service; 1988-1999
[n21.] National Center for Health Statistics. 1995 and 1996 Birth Cohort Linked Birth/Infant Death Data Sets. NCHS
CD-ROM, Series 20, No. 12a, 1998; and NCHS CD-ROM, Series 20, No. 14a, 1999
[n22.] Mathews TJ, Curtin SC, MacDorman MF. Infant Mortality Statistics From the 1998 Period Linked Birth/Infant
Death Data Set. National Vital Statistics Reports. Hyattsville, MD: National Center for Health Statistics; 2000;48 (No.
12)
[n23.] National Center for Health Statistics. Medical Examiners' and Coroners' Handbook on Death Registration and
Fetal Death Reporting. Hyattsville, MD: National Center for Health Statistics; 1987. DHHS Publ. No. (PHS) 87-1110
[n24.] Sorenson SB, Shen H, Kraus JF. Coroner-reviewed infant and toddler deaths. Eval Rev. 1997;21:58-76
[n25.] Jason J, Carpenter MM, Tyler CW. Underrecording of infant homicide in the United States. Am J Public Health.
1983;73:195-197
[n26.] Siegel CD, Graves P, Maloney K, Norris JM, Calonge BN, Lezotte D. Mortality from intentional and unintentional injury among infants of young mothers in Colorado, 1986-1992. Arch Pediatr Adolesc Med. 1996;150:1077-1083
[n27.] Cummings P, Theis MK, Mueller BA, Rivara FP. Infant injury death in Washington State, 1981-1990. Arch Pediatr Adolesc Med. 1994;148:1021-1026
[n28.] National Center for Health Statistics. Vital Statistics of the United States, I, Natality. Washington, DC: Public
Health Service; 1997
[n29.] American Academy of Pediatrics, Task Force on Infant Sleep Position and Sudden Infant Death Syndrome.
Changing concepts of sudden infant death syndrome: implications for infant sleeping environment and sleep position.
Pediatrics. 2000;105:650-656
[n30.] Willinger M, James LS, Catz C. Defining the sudden infant death syndrome (SIDS): deliberations of an expert
panel convened by the National Institute of Child Health and Human Development. Pediatr Pathol. 1991;11:677-684
[n31.] Meadows R. Unnatural sudden infant death. Arch Dis Child. 1999;80:7-14
[n32.] Hampton RL, Newberger EH. Child abuse incidence and reporting by hospitals: significance of severity, class
and race. Am J Public Health. 1985;75:56-60
[n33.] Dubowitz H, Hampton RL, Bithoney WG, Newberger EH. Inflicted and noninflicted injuries: differences in child
and familial characteristics. Am J Orthopsychiatry. 1987;57:525-535
[n34.] Jason J. Fatal child abuse in Georgia: the epidemiology of severe physical child abuse. Child Abuse Negl.
1983;7:1-9
[n35.] Kunz J, Bhar S. A profile of parental homicide against children. J Fam Violence. 1996;11:347-362
[n36.] Department of Health and Human Services, National Center on Child Abuse and Neglect. Child Maltreatment
1995: Reports From States to the National Child Abuse and Neglect Data System. Washington, DC: Government Printing Office; 1997:2-9
Page 26
[n37.] Bureau of Justice Statistics. Honticide Trends in the U.S. Washington, DC: National Criminal Justice; 2000.
Publ. No. 179767. Available at: www.ojp.usdoj.gov/bjs/homicide
[n38.] National Center for Child Fatality Review. "How To" Guide for Child Fatality Review Teams. Available at:
www.ican-ncfr.org
[n39.] Chukwudi OS, Forjuoh SN, West P, Brooks C. Child death reviews: a gold mine for injury prevention and control. Inj Prev. 1999;5:276-279
GRAPHIC: Fig 1, Infant mortality rates for sudden, unexpected deaths and unknown causes, United States, 1983-1997.
Copyright © 2001 American Academy of Pediatrics
Pediatrics 2001; 107: 437-441
February, 2001
SECTION: AMERICAN ACADEMY OF PEDIATRICS
LENGTH: 2781 words
TITLE: Distinguishing Sudden Infant Death Syndrome From Child Abuse Fatalities
AUTHOR: COMMITTEE ON CHILD ABUSE AND NEGLECT, 2000-2001, Steven W. Kairys, MD, MPH, Chairperson, Randell C. Alexander, MD, PhD, Robert W. Block, MD, V. Denise Everett, MD, Kent P. Hymel, MD, Carole
Jenny, MD, MBA; LIAISON REPRESENTATIVES, David L. Corwin, MD, American Academy of Child and Adolescent Psychiatry, Gene Ann Shelley, PhD, Centers for Disease Control and Prevention; SECTION LIAISON, Robert M.
Reece, MD, Section on Child Abuse and Neglect; CONSULTANT IN PEDIATRIC PATHOLOGY, Henry F. Krous,
MD, Children's Hospital of San Diego, CA; STAFF, Tammy Piazza Hurley
ABSTRACT. In most cases, when a healthy infant younger than 1 year dies suddenly and unexpectedly, the cause is
sudden infant death syndrome (SIDS). SIDS is more common than infanticide. Parents of SIDS victims typically are
anxious to provide unlimited information to professionals involved in death investigation or research. They also want
and deserve to be approached in a non-accusatory manner. This statement provides professionals with information and
guidelines to avoid distressing or stigmatizing families of SIDS victims while allowing accumulation of appropriate
evidence in potential cases of death by infanticide.
TEXT:
ABBREVIATIONS. SIDS, sudden infant death syndrome; ALTE, apparent life-threatening events.
Approximately 50 years ago, the medical community began a search to understand and prevent sudden infant
death syndrome (SIDS). [n1,n2] Almost simultaneously, medical professionals were awakened to the realities of child
abuse. [n3-n6] Since then, public and professional awareness of SIDS and fatal child abuse during infancy have increased steadily. Recently, well-validated reports of child abuse and infanticide -- perpetrated by suffocation and masqueraded as apparent life-threatening events (ALTE) and/or SIDS -- have appeared in the medical literature and in the
lay press. [n7,n8] The differentiation between SIDS and fatal child abuse can be a critical diagnostic decision. [n9] Additional funding for research into the causes and prevention of SIDS and child abuse is needed.
SIDS: EPIDEMIOLOGY, PRESENTATION, AND RISK FACTORS
Page 27
SIDS, also called crib or cot death, is the sudden death of an infant under 1 year of age that remains unexplained after thorough case investigation, including performance of a complete autopsy, examination of the death scene, and a
review of the clinical history. [n10] SIDS is the most common cause of death between 1 and 6 months of age. The incidence of SIDS peaks between 2 and 4 months of age. Approximately 90% of SIDS deaths occur before the age of 6
months. [n11]
SIDS is suspected when a previously healthy infant, usually younger than 6 months, is found dead in bed, prompting an urgent call for emergency assistance. Often, the baby is fed normally just before being placed in bed to sleep, no
outcry is heard, and the baby is found in the position in which he or she had been placed at bedtime or naptime. In some
cases, cardiorespiratory resuscitation initiated at the scene by emergency personnel is continued without apparent beneficial effect en route to the hospital, where the baby is finally declared dead. Evidence of terminal motor activity, such
as clenched fists, may be seen. There may be serosanguineous, watery, frothy, or mucoid discharge coming from the
nose or mouth. Skin mottling and postmortem lividity in dependent portions of the infant's body are commonly found.
Review of the medical history, scene investigation, radiographs, and autopsy are unrevealing.
Despite extensive research, understanding of the etiology of SIDS remains incomplete. The discovery of abnormalities in the arcuate nucleus of the brainstems of some SIDS victims suggests that true SIDS deaths likely reflect delayed
development of arousal, cardiorespiratory control, or cardiovascular control. [n12,n13] When the physiologic stability
of such infants becomes compromised during sleep, they may not arouse sufficiently to avoid the noxious insult or condition. [n14]
The SIDS rates are 2 to 3 times higher among African American and some American Indian populations. SIDS has
been linked etiologically in research studies to prone sleep position, sleeping on a soft surface, maternal smoking during
or after pregnancy, overheating, late or no prenatal care, young maternal age, prematurity, low birth weight, and male
gender. [n15-n23] To date, no definitive evidence establishes causality between SIDS and recurrent cyanosis, apnea,
ALTE, or immunizations during infancy. When recurrent cyanosis, apnea, or ALTE during infancy are reported, pediatricians should document these events objectively and determine if or not these events have occurred in the presence of
more than 1 caregiver.
In recent years, national campaigns aimed at reducing prone sleeping during infancy have dramatically decreased
the incidence of SIDS in the United States and in other countries. [n14,n24-n29] Many of these educational campaigns
have also emphasized prompt evaluation and treatment for sick infants, appropriate immunizations, breastfeeding, and
avoidance of overlying, overheating, overwrapping, gestational or postnatal passive smoke exposure, and soft sleep materials or surfaces.
SIDS: A DIAGNOSIS OF EXCLUSION
The diagnosis of SIDS is exclusionary and requires a postmortem examination, death scene investigation, [n30] and
review of case records that fail to reveal another cause of death. Infant deaths without post-mortem examination should
not be attributed to SIDS. Cases that are autopsied and carefully investigated but reveal substantial and reasonable uncertainty regarding the cause or manner of death may be designated as undetermined. Examples of undetermined cases
include suspected (but unproven) infant death attributable to infection, metabolic disease, accidental asphyxiation, or
child abuse.
A diagnosis of SIDS reflects the clear admission by medical professionals that an infant's death remains completely
unexplained. A young infant's death should be ruled as attributable to SIDS when all of the following are true:
* a complete autopsy is done, including cranium and cranial contents, and autopsy findings are compatible with SIDS;
* there is no gross or microscopic evidence of trauma or significant disease process;
* there is no evidence of trauma on skeletal survey [n31];
* other causes of death are adequately ruled out, including meningitis, sepsis, aspiration, pneumonia, myocarditis, abdominal trauma, dehydration, fluid and electrolyte imbalance, significant congenital lesions, inborn metabolic disorders,
carbon monoxide asphyxia, drowning, or burns;
* there is no evidence of current alcohol, drug, or toxic exposure; and
* thorough death scene investigation and review of the clinical history are negative.
CHILD ABUSE FATALITIES BY SUFFOCATION
Page 28
As the occurrence of cases of true SIDS has decreased, the proportion of unexplained infant deaths attributable to
fatal child abuse may be increasing. [n32] Estimates of the incidence of infanticide among cases designated as SIDS
range from <1% to 5%. [n7,n9,n33-n35]
Parents of infants with recurrent ALTEs have been observed trying to suffocate and harm their infants. [n7,n36] In
Great Britain, covert video surveillance was used to assess child abuse risk in 39 young children referred for evaluation
of recurrent ALTEs. [n7] Abuse was revealed in 33 of 39 cases, with documentation of intentional suffocation observed
in 30 patients. Among 41 siblings of the 39 infants in the studies, 12 had previously died suddenly and unexpectedly.
Although 11 of these deaths had been classified as SIDS, 4 parents later admitted to suffocating 8 of these siblings. Other cases previously thought to be multiple SIDS deaths within a family [n37] have been revealed to be cases of multiple
homicide by suffocation. [n8,n32]
It is impossible to distinguish at autopsy between SIDS and accidental or deliberate asphyxiation with a soft object.
[n38] However, certain circumstances should indicate the possibility of intentional suffocation, including:
* previous recurrent cyanosis, apnea, or ALTE while in the care of the same person;
* age at death older than 6 months;
* previous unexpected or unexplained deaths of 1 or more siblings;
* simultaneous or nearly simultaneous death of twins [n39];
* previous death of infants under the care of the same unrelated person [n40]; or
* discovery of blood on the infant's nose or mouth in association with ALTEs. [n7]
MANAGEMENT OF SUDDEN UNEXPECTED INFANT DEATH
Most sudden infant deaths occur at home. Parents are shocked, bewildered, and distressed. Parents who are innocent of blame in their child's death often feel responsible nonetheless and imagine ways in which they might have contributed to or prevented the tragedy. [n41,n42] The appropriate professional response to any child death must be compassionate, empathic, supportive, and nonaccusatory. Inadvertent comments, as well as necessary questioning by medical personnel and investigators, are likely to cause additional stress. It is important for those in contact with parents during this time to be supportive while at the same time conducting a thorough investigation.
Personnel on first-response teams should be trained to make observations at the scene, including position of the infant, marks on the body, body temperature and rigor, type of bed or crib and any defects, amount and position of clothing and bedding, room temperature, type of ventilation and heating, and reaction of the caregivers. Guidelines are available for death scene investigation of sudden, unexplained infant deaths. [n30,n33] Paramedics and emergency department personnel should be trained to distinguish normal findings, such as postmortem anal dilation and lividity, from
trauma attributable to abuse. [n42,n43]
When a previously healthy infant has died unexpectedly in the absence of external evidence of injury, a preliminary
diagnosis of "probable SIDS" can be given. To the family of a true victim of SIDS, this diagnosis conveys the health
care provider's initial impression that they could not have prevented their infant's death. Assignment of this preliminary
diagnosis should not limit or prevent subsequent thorough case investigation.
Parents should be informed that other causes of death will be excluded only by thorough death scene investigation,
postmortem examination, and review of case records. It should be explained to parents that these procedures might enable them and their physician to understand why their infant died and how other children in the family, including children born later, might be affected. Only on completion of a thorough and negative case investigation (including performance of a complete autopsy, examination of the death scene, and review of the clinical history) should a definitive
diagnosis of SIDS be assigned as the cause of death.
The family is entitled to an opportunity to see and hold the infant once death has been pronounced. A protocol
[n44] may help in planning how and when to address the many issues that require attention, including baptism, grief
counseling, funeral arrangements and religious support, cessation of breastfeeding, and the reactions of surviving siblings. [n41,n45] All parents should be provided with information about SIDS [n46,n47] and the telephone number of the
local SIDS support group. [n46]
Controversy exists in the medical literature regarding the likelihood of a repetition of SIDS within a sibship. [n48n51] When an infant's sudden and unexpected death has been thoroughly evaluated and alternate environmental or acci-
Page 29
dental causes of death have been carefully excluded, parents should be informed that the risk for SIDS in subsequent
children is not likely increased.
In many states, multidisciplinary teams have been established to review child fatalities. [n52,n53] Ideally, a multidisciplinary death review committee should include a child welfare/child protective services social worker, a law enforcement officer, a public health officer, the medical examiner/coroner, a pediatrician with expertise in child maltreatment, a forensic pathologist, a pediatric pathologist, and the local prosecutor. [n53] The proceedings of multidisciplinary death review committees should remain confidential. Sharing data among agencies helps ensure that deaths attributable to child abuse are not missed and that surviving and subsequent siblings are protected. Some child fatality
teams routinely review infant deaths attributable to apparent SIDS.
THE IMPORTANCE OF AUTOPSY, SCENE INVESTIGATION, AND CASE REVIEW
The failure to differentiate fatal child abuse from SIDS is costly. In the absence of postmortem examination, death
scene investigation, and case review, child maltreatment is missed, familial genetic diseases go unrecognized, public
health threats are overlooked, inadequate medical care goes undetected, product safety issues remain unidentified, and
progress in understanding the etiology of SIDS and other causes of unexpected infant death is delayed. Inaccurate vital
statistics lead to inappropriate allocation of limited health care resources. By thoroughly investigating apparent SIDS
deaths, the potential hazards of defective infant furniture, water beds, and beanbag mattresses have been identified and
remedied. [n54,n55]
If appropriate toxicologic tests are not done, the few infant deaths attributable to accidental or deliberate poisoning
will be missed. [n42,n56] Occult cocaine exposure is widespread and potentially lethal. One review found that 17 (40%)
of 43 infants who died before 2 days of age without an obvious cause of death at autopsy had toxicologic evidence of
cocaine exposure. [n57] A second review of 600 infant deaths revealed evidence of cocaine exposure in 16 infants
(2.7%) younger than 8 months who died suddenly and unexpectedly. [n58] Lethal concentrations of cocaine and many
other drugs in infancy are not yet established.
POSTMORTEM IMAGING
Radiographic skeletal surveys performed before autopsy in cases of suspected SIDS may reveal evidence of traumatic skeletal injury or skeletal abnormalities indicative of a naturally occurring illness. Thorough documentation of all
sites of suspected skeletal injury may require specimen resection and high-detail specimen radiography. The presence of
old and new inflicted traumatic injuries identified on skeletal survey before autopsy may lend focus to the postmortem
examination, death scene investigation, and police investigation. [n31,n59]
PATHOLOGY
The American Academy of Pediatrics endorses universal performance of autopsies on infants who die suddenly and
unexpectedly. [n60] An international standardized autopsy protocol is available for this purpose. [n61] Postmortem
findings in cases of fatal child abuse most often reveal cranial injuries, abdominal trauma (eg, liver laceration, hollow
viscous perforation, or intramural hematoma), burns, drowning, or exposure as the cause of death. [n62-n65]
Pathologists establish the diagnosis of SIDS by exclusion when they are unable to identify other specific causes for a
child's death. [n42]
Intrathoracic petechiae are identified in 80% to 85% of SIDS cases but are not pathognomonic. Substantial evidence regarding intrathoracic petechiae in human and experimental studies supports the hypothesis that upper airway
obstruction is the final event in SIDS. [n66]
Inborn errors of metabolism [n67-n69] have been implicated to cause a small percentage of sudden unexplained
deaths in infants with autopsy findings consistent with SIDS. Although cytomegaloviral inclusion bodies have been
identified in some infants who died suddenly and unexpectedly, a definitive causal link between cytomegaloviral infection and SIDS has not been established. [n70] Analysis of blood or other body fluids (urine, vitreous humor, cerebrospinal fluid, bile, and stomach contents collected and stored at -80 deg. C) and brain, liver, kidney, heart, muscle, adrenal
gland, and/or pancreas tissue may facilitate diagnosis of a fatal inborn error of metabolism. Blood tests for evaluation of
many metabolic disorders are now available at low cost.
RECOMMENDATIONS
Page 30
The Academy makes the following recommendations for evaluation of sudden, unexplained infant deaths:
* accurate history taking by emergency responders and medical personnel at the time of death and made available to the
medical examiner or coroner;
* prompt death scene investigation [n30,n33] where the infant was found lifeless and careful interviews of household
members by knowledgeable individuals (potentially including a pediatrician);
* examination of the dead infant at a hospital emergency department by a child maltreatment specialist;
* postmortem examination following established protocol [n57] within 24 hours of death, including radiographic skeletal survey, toxicologic, and metabolic screening;
* collection of medical history through interviews of caretakers, interviews of key medical providers, and review of
previous medical records;
* maintenance of a supportive approach to parents during the death review process;
* consideration of intentional asphyxiation in cases of unexpected infant death with a history of recurrent cyanosis, apnea, or ALTE witnessed only by a single caretaker or in a family with previous unexplained infant death(s);
* use of accepted diagnostic categories on death certificates as soon as possible after review;
* prompt informing sessions with parents when results indicate SIDS or medical causation of death; and
* locally based infant death review teams [n49] to review collected data with participation of the medical examiner or
coroner in the review.
SUPPLEMENTARY INFORMATION: The recommendations in this statement do not indicate an exclusive course
of treatment or serve as a standard of medical care. Variations, taking into account individual circumstances, may be
appropriate.
REFERENCES:
[n1.] Werne J, Garrow L Sudden apparently unexplained death during infancy: pathological findings in infants found
dead. Am J Pathol. 1953;29:633-652
[n2.] Adelson L, Kinney ER. Sudden and unexpected death in infancy and childhood. Pediatrics. 1956;17:663-697
[n3.] Caffey J. Multiple fractures in the long bones of infants suffering from chronic subdural hematoma. AJR Am J
Roentgenol. 1946;56:163-173
[n4.] Silverman FN. Roentgen manifestations of unrecognized skeletal trauma in infants. AJR Am J Roentgenol.
1953;69:413-427
[n5.] Adelson L. Slaughter of the innocents: a study of forty-six homicides in which the victims were children. N Engt J
Med. 1961;264:1345-1349
[n6.] Kempe CH, Silverman FN, Steele BF, Droegemueller W, Silver HK. The battered child syndrome. JAMA.
1962;181:17
[n7.] Southall DP, Plunkett MC, Banks MW, Falkov AF, Samuels MP. Covert video recordings of life-threatening child
abuse: lessons for child protection. Pediatrics. 1997;100:735-760
[n8.] Firstman R, Talan J. The Death of Innocents: A True Story of Murder, Medicine, and High-stakes Science. New
York, NY: Bantam; 1997
[n9.] Reece RM. Fatal child abuse and sudden infant death syndrome: a critical diagnostic decision. Pediatrics.
1993;91:423-429
[n10.] Willinger M, James LS, Catz. C. Defining the sudden infant death syndrome (SIDS): deliberations of an expert
panel convened by the National Institute of Child Health and Human Development. Pediatr Pathol. 1991;11:677-684
Page 31
[n11.] Peterson DR. Clinical implications of sudden infant death syndrome epidemiology. Pediatrician. 1988;15:198203
[n12.] Kinney HC, Filiano JJ, Sleeper LA, Mandell F, Valdes-Despena M, White WF. Decreased muscarinic receptor
binding in the arcuate nucleus in sudden infant death syndrome. Science. 1995;269:1446-1450
[n13.] Panigrahy A, Filiano JJ, Sleeper LA, et al. Decreased kainate binding in the arcuate nucleus of the sudden infant
death syndrome. J Neuropathol Exp Neurol. 1997;56:1253-1261
[n14.] American Academy of Pediatrics, Task Force on Infant Sleep Position and Sudden Infant Death Syndrome.
Changing concepts of sudden infant death syndrome: implications for infant sleeping environment and sleep position.
Pediatrics. 2000;105:650-656
[n15.] Hoffman HJ, Damus K, Hillman L, Krongrad E. Risk factors for SIDS. Results of the National Institute of Child
Health and Human Development SIDS cooperative epidemiological study. Ann N Y Acad Sci. 1988;533:13-30
[n16.] Hoffman HJ, Hillman LS. Epidemiology of the sudden infant death syndrome: maternal, neonatal, and postneonatal risk factors. Clin Perinatol. 1992;19:717-737
[n17.] Ponsonby A-L, Dwyer T, Gibbons LE, Cochrane JA, Wang Y-G. Factors potentiating the risk of sudden infant
death syndrome associated with prone position. N Engl J Med. 1993;329:377-382
[n18.] Kemp JS, Nelson VE, Thach BT. Physical properties of bedding that may increase risk of sudden infant death
syndrome in prone-sleeping infants. Pediatr Res. 1994;36:7-11
[n19.] Chiodini BA, Thach BT. Impaired ventilation in infants sleeping face down: potential significance for sudden
infant death syndrome. J Pediatr. 1993;123:686-692
[n20.] Jeffery HE, Megevand A, Page HD. Why the prone position is a risk factor for sudden infant death syndrome.
Pediatrics. 1999;104:263-269
[n21.] MacDorman MF, Cnattingius S, Hoffman HJ, Kramer MS, Haglund B. Sudden infant death syndrome and
smoking in the United States and Sweden. Am J Epidemiol. 1997;146:249-257
[n22.] Schoendorf KC, Kiely JL. Relationship of sudden infant death syndrome to maternal smoking during and after
pregnancy. Pediatrics. 1992;90:905-908
[n23.] Fleming PJ, Blair PS, Bacon C, et al. Environment of infants during sleep and risk of sudden infant death syndrome: results of 1993-5 casecontrol study for confidential inquiry into stillbirths and deaths in infancy. BMJ.
1996;313:191-195
[n24.] Willinger M, Hoffman HJ, Wu K-T, et al. Factors associated with the transition to non-prone sleep positions of
infants in the United States: The National Infant Sleep Position Study. JAMA. 1998;280:329-335
[n25.] Mitchell EA, Brunt JM, Everard C. Reduction in mortality from sudden infant death in New Zealand: 1986-92.
Arch Dis Child. 1994;70:291-294
[n26.] Platt MJ, Pharoah POD. Child health statistical review, 1996. Arch Dis Child. 1996;75:527-533
[n27.] Dwyer T, Ponsonby A-L, Blizzard L, Newman NM, Cochrane JA. The contribution of changes in prevalence of
prone sleeping position to the decline in sudden infant death syndrome in Tasmania. JAMA. 1995;273:783-789
[n28.] Wennergren G, Alm B, Oyen N, et al. The decline in the incidence of SIDS in Scandinavia in its relation to riskintervention campaigns. Acta Pacdiatr. 1997;86:963-968
Page 32
[n29.] US Public Health Service, American Academy of Pediatrics, SIDS Alliance, and the Association of SIDS and
Infant Mortality Programs. Reduce the Risk of Sudden Infant Death Syndrome (SIDS). Elk Grove Village, IL: American Academy of Pediatrics; 2000
[n30.] Centers for Disease Control and Prevention. Guidelines for death scene investigation of sudden, unexplained infant deaths: recommendations of the Interagency Panel on Sudden Infant Death Syndrome. MMWR Morb Mortal Wkly
Rep. 1996;45:1-22
[n31.] Kleinman PK, Blackbourne BD, Marks SC, Karellas A, Belanger PL. Radiologic contributions to the investigation and prosecution of cases of fatal infant abuse. N Engl J Med. 1989;320:507-511
[n32.] Meadow R. Unnatural sudden infant death. Arch Dis Child. 1999;80:7-14
[n33.] Bass M, Kravath RE, Glass L. Death-scene investigation in sudden infant death. N Engl J Med. 1986;315:100105
[n34.] McClain PW, Socks IJ, Frochlke RG, Ewigman BG. Estimates of fatal child abuse and negled, United States,
1979 through 1988. Pediatrics. 1993;91:338-343
[n35.] Kukull WA, Peterson DR. Sudden infant death and infanticide. Am J Epidemiol. 1977;106:485-486
[n36.] Rosen CL, Frost JD, Bricker T, Tarnow JD, Gillette PC, Dunlavy S. Two siblings with recurrent cardiorespiratory arrest: Munchausen syndrome by proxy or child abuse? Pediatrics. 1983;71:715-720
[n37.] Steinschneider A. Prolonged apnea and the sudden infant death syndrome: clinical and laboratory observations.
Pediatrics. 1972;50:646-654
[n38.] Valdes-Dapena M. The sudden infant death syndrome: pathologic findings. Clin Perinatol. 1992;19:701-716
[n39.] Groothius JR, Altemeier WA, Robarge JP, et al. Increased child abuse in families with twins. Pediatrics.
1982;70:769-773
[n40.] Meadow R. Suffocation, recurrent apnea, sudden infant death. J Pediatr. 1990;117:351-357
[n41.] Limerick S. Family and health-professional interactions. Ann N Y Acad Sci. 1988;533:145-154
[n42.] DiMaio DJ, DiMaio VJM. Forensic Pathology. New York, NY: Elsevier Science Publishing Co Inc; 1989:289321
[n43.] Kirschner RH, Stein RJ. The mistaken diagnosis of child abuse: a form of medical abuse? Am J Dis Child.
1985;139:873-875
[n44.] Sudden Infant Death Syndrome Standards for Services to Families. Association of SIDS and Infant Mortality
Programs Web site. Available at: www.asipl.org/offices.html. Accessed June 9, 2000
[n45.] American Academy of Pediatrics, Committee on Psychosocial Aspects of Child Family Health. The pediatrician
and childhood bereavement. Pediatrics. 1992;89:516-518
[n46.] The SIDS Alliance Web site. Available at: www.sidsalliance.org. Accessed June. 2000;9
[n47.] The National SIDS Resource Center Web site. Available at: www.circsol.com/sids. Accessed June 9, 2000
[n48.] Peterson DR, Chinn NM, Fisher LD. The sudden infant death syndrome: repetitions in families. J Pediatr.
1980;97:265-267
Page 33
[n49.] Oyen N, Skjaerven R, Irgens LM. Population-based recurrence risk of sudden infant death syndrome compared
with other infant and fetal deaths. Am J Epidemiol. 1996;144:300-305
[n50.] Irgens LM, Skjaerven R, Peterson DR. Prospective assessment of recurrence risk in sudden infant death syndrome siblings. J Pediatr. 1984;104:349-351
[n51.] Irgens LM, Oyen N, Skjaerven R. Recurrence of sudden infant death syndrome among siblings. Acta Paediatr
Suppl. 1993;82(suppl 389):23-25
[n52.] Kaplan SR. Granik LA, eds. Child Fatality Investigative Procedures Manual. Chicago, IL: American Bar Association; 1991
[n53.] Granik LA, Durfee M, Wells SJ. Child Death Review Teams: A Manual for Design and Implementation. Chicago, IL: American Bar Association; 1991
[n54.] Kemp JS, Thach BT. Sudden death in infants sleeping on polystyrene-filled cushions. N Engl J Med.
1991;324:1858-1864
[n55.] Ramanatha R, Chandra S, Gilbert-Barness E, Franciosi R. Sudden infant death syndrome and water beds. N
Engl J Med. 1988;318:1700
[n56.] Perrot LJ, Nawojczyk S. Nonnatural death masquerading as SIDS (sudden infant death syndrome). Am J Forensic Med Pathol. 1988;9:105-111
[n57.] Rogers C, Hall J, Muto J. Findings in newborns of cocaine-abusing mothers. J Forensic Sci. 1991;36:1074-1078
[n58.] Mirchandani HG, Mirchandani IH, Hellman F, English-Rider R, Rosen S, Laposata EA. Passive inhalation of
free-base cocaine ('crack') smoke in infants. Arch Pathol Lab Med. 1991;115:494-498
[n59.] Kleinman PK. Postmortem imaging. In: The Diagnostic Imaging of Child Abuse. 2nd ed. St Louis, MO: Mosby
Inc; 1998:242-245
[n60.] American Academy of Pediatrics, Committee on Child Abuse and Neglect and Committee on Community Health
Services. Investigation and review of unexpected infant and child deaths. Pediatrics. 1999;104:1158-1160
[n61.] Krous H. Instruction and reference manual for the international standardized autopsy protocol for sudden unexpected infant death. J Sudden Infant Death Syndrome Infant Mortal. 1996;1:203-246
[n62.] Brown RH. The battered child syndrome. J Forensic Sci. 1976;21:65
[n63.] Lauer B, Ten Brock E, Grossman M. Battered child syndrome: review of 130 patients with controls. Pediatrics.
1974;54:67
[n64.] Scott PD. Fatal battered baby cases. Med Sci Law. 1973;13:197
[n65.] Wecht CH, Larkin GM. The battered child syndrome: a forensic pathologist's viewpoint. Med Trial Tech Q.
1981;28:1-24
[n66.] Beckwith JB. Intrathoracic petechial hemorrhage: a clue to the mechanism of death in sudden infant death syndrome? Ann N Y Acad Sci. 1988;533:37-47
[n67.] Howat AJ, Bennett MJ, Variend S, et al. Defects of metabolism of fatty acids in the sudden infant death syndrome. Br Med J (Clin Res Ed). 1985;290:1771-1773
Page 34
[n68.] Vawter GF, McGraw CA, Hug G, et al. An hepatic metabolic profile in sudden infant death (SIDS). Forensic
Sci Int. 1986;30:93-98
[n69.] Harpey JP, Charpentier C, Paturneau-Jonas M. Sudden infant death syndrome and inherited disorders of fatty
acid beta-oxidation. Metab Proh Newborn. 1990;58(suppl):70-80
[n70.] Variend S, Pearse RG. Sudden infant death and cytomegalovirus inclusion disease. J Clin Pathol. 1986;39:383390