DEATH CERTIFICATION

Although the origins of the death certificate go as far back as 1538, the US Standard Certificate of Death was adopted in 1900 (1). In 1910, a requirement to record manner of death on all certificates for deaths not due to natural causes was initiated (2, 3). Despite its long history, the death certificate is sometimes misunderstood and incorrectly completed. Completion of the medical portion of the certificate requires entry of the cause of death (COD) and manner of death (MOD) by the certifier of death (i.e., the medical provider, medical examiner, or coroner). The demographic portion of the certificate is completed by funeral directors with the assistance of informants (usually next of kin). The state office of vital records is responsible for quality assurance and maintenance of certificates.

Accurate completion of the medical portion of the death certificate is dependent on a basic understanding of cause and manner of death. The certified underlying cause of death, defined as the disease, injury, or combination thereof that initiates an ultimately lethal sequence of events, and without which the death would not have occurred, must be etiologically specific (3). The underlying cause of death and conditions in the sequence of events leading to death are entered in a section of the certificate titled “Part I. Cause of Death” (Figure 10.1).

Figure 10.1

Portion of the US Standard Certificate of Death where cause and contributory cause statements are entered (4).

Contributory cause(s) of death, entered in the certificate field titled “Part II” (Figure 10.1), are conditions that added to the physiologic or anatomic processes leading to death, but did not initiate the lethal sequence of events. While a correctly certified death has only one underlying COD, there may be multiple contributors. Part II should be reserved for true contributors and is not intended to be a repository for extraneous diagnoses or information (2).

Manner of death reflects the circumstances of death and is limited, in most medicolegal jurisdictions, to five choices: Natural, Accident, Suicide, Homicide, and Could not be determined (commonly referred to as Undetermined) on the US Standard Death Certificate (3). Certification of manner requires knowledge of both the cause of death and the circumstances provided by death investigation. If there is any contribution to death by an extrinsic, unnatural factor, no matter how small, the manner should not be certified as Natural. Stating the converse, certifying manner as Natural implies death was due entirely to natural causes, with no contribution by trauma, environment, medication, or toxins (2, 3).

Certification of a death as Accident, Homicide, or Suicide requires a brief description of the circumstances in a designated field of the certificate titled “Describe How Injury Occurred.” Requirements for completion of that field vary by jurisdiction when the manner is Undetermined. A simple example is an individual who is shot in the chest by another individual. The cause of death is “Gunshot wound of chest” and the manner of death is Homicide. For the “Describe how injury occurred” section of the death certificate, “Shot by other” may be entered (3).

Death certificates are key data sources for mortality surveillance which informs health policies, prevention strategies, and resource allocation for public health research and interventions. High quality and consistent certification practices would benefit surveillance of sudden, unexpected death in infancy, which has been designated as a high-priority health issue in Healthy People 2020 (5).

CODING AND CLASSIFICATION

The National Center for Health Statistics (NCHS) compiles and coordinates mortality data from the states into a national file via the National Vital Statistics System, the oldest example of intergovernmental sharing of public health data in the US. Data in the medical portion of the death certificate (e.g., data in fields titled “Part I. Cause of Death”, “Part II. Other significant conditions contributing to death”, and “Describe how injury occurred”) are used to generate cause of death codes, which allow for easier death classification by nosologists and end users. The data can be accessed through online query tools such as CDC WONDER and WISQARS and other public use data files (6, 7). Users of National Vital Statistics System mortality data and data query tools – including health and medical researchers, federal and state agencies, private entities, and academic institutions – may classify/categorize deaths according to their own needs using the provided codes. Some users will obtain original certificates through other means to use literal text in their own analyses and classifications. The classification of a death may not necessarily be based on the underlying cause of death. For example, surveillance of lethal infections may be accomplished by tracking deaths classified as sepsis, which is a mechanism of death caused by many different underlying causes. Certifiers are not responsible for coding or classification, but the accuracy and phrasing of the certification directly influence subsequent coding, reported national health statistics, and the results of studies based on those statistics.

Causes of death are tracked using International Statistical Classification of Diseases and Related Health Problems (ICD) codes, published by the World Health Organization (WHO) and currently in the tenth revision (ICD-10) (8). Cause of death statements from the death certificate (often referred to as the literal text of the COD) are assigned codes based on the WHO ICD classification rules and supplemental coding instructions by the National Center for Health Statistics (9). The National Center for Health Statistics uses both automated and manual systems to apply ICD coding rules. Manual coding by nosologists is used when terminology is not recognizable by the automated system or when coding rules are complex; this often occurs for infant death certificates with complicated listings of risk factors and circumstances. One code is assigned for the underlying COD and secondary codes capture contributory causes and other diagnostic information. In the US, the ICD–10 codes most frequently assigned to sudden, unexpected infant deaths include R95 (Sudden Infant Death Syndrome), R99 (other ill-defined and unspecified causes of mortality), and W75 (accidental suffocation and strangulation in bed) (Table 10.1). Sudden unexplained deaths in childhood are most frequently assigned the R99 ICD-10 code, with fewer coded as R96 (other sudden death, cause unknown), neither of which is specific to childhood. A broad range of cause of death statements will result in an infant death being coded as R95 (Sudden Infant Death Syndrome), including sudden infant death, sudden infant death syndrome, sudden unexplained death in infancy, sudden unexpected infant death, and the much older terms cot death and crib death. Because of coding rules, the certifier’s cause of death may not be accurately reflected in the resultant underlying cause of death code (explained further below).

Table 10.1

Most Frequent Codes for Sudden Unexplained Pediatric Deaths.

DIAGNOSTIC SHIFT IN CERTIFICATION OF SUDDEN UNEXPLAINED PEDIATRIC DEATHS

In the 1980s, several case-comparison studies showed that prone sleeping was an important risk factor for sudden infant death syndrome (10–13). In response to these findings, supine sleep position was promoted by the American Academy of Pediatrics and the “Back to Sleep” campaign in the early 1990s (14, 15). Following campaign initiation, a dramatic increase in infants placed to sleep supine was reported (17% in 1993 versus 72% in 2001) with a corresponding decline in sudden infant death syndrome rates per 1000 live births (1.1 in 1993 to 0.6 in 2001, a 50% decline) (16). However, adherence to supine, infant sleep position recommendations and the corresponding decline in sudden, unexplained infant death rates appears to have slowed or stagnated and there are large variations in death rates by state (17). The reasons for the apparent plateau in death rates is unknown but a topic of much discussion.

Since 1999, US deaths with sudden infant death syndrome (R95) as the underlying cause of death code have shifted downward in frequency and infant deaths with explained causes (e.g., asphyxiation in an unsafe sleep environment [W75]) and other unexplained and undetermined causes (R99) have shifted upwards (17–19). Reasons for the shift in coding are likely multifactorial and related to improved cause determination through better death investigations, increased recognition that many unexplained infant deaths may not be due to a natural syndrome, and changes over time in individual or jurisdictional diagnostic preferences. A 2014 survey of US medical examiners and coroners found that only about 50% of certifiers would use the term Sudden Infant Death Syndrome to describe an infant death that was sudden and remained unexplained following a thorough investigation (20). Standardization of infant death reporting practices has been repeatedly called for, but no reporting algorithm has achieved general acceptance. Consistent reporting practices are critical for reliably comparing mortality rates and trends across jurisdictions, states, and countries.

In an attempt to mitigate the effect of certification variation due to individual or jurisdictional diagnostic preferences and allow for comparisons over time and between geographic areas, the ICD-10 codes R95, R99, and W75 are often grouped together as Sudden Unexpected Infant Deaths (19, 21). The explained deaths coded as W75 (accidental suffocation and strangulation in bed) are grouped with unexplained deaths in an unsafe sleep environment coded as R95 and R99 to reflect the lack of uniformly accepted and applied diagnostic criteria for asphyxial deaths.

Death certificate data is also used to monitor sudden, unexplained deaths in childhood. The ICD-10 codes R96 and R99 reflect deaths with ill-defined and unknown causes. When used in combination with age at death, R96 can be used track sudden, unexplained deaths in children after infancy. R96 specifically captures use of the term “sudden”; cause statements such as “sudden death of undetermined etiology” or “unexplained sudden death” will be coded as R96 when age at death is ≥ 365 days. In contrast, cause statements that indicate an undetermined or unexplained death but lack the word “sudden”, will be coded as R99 regardless of age. These codes (R96 and R99) will only be assigned if there are no other codable conditions reported on the death certificate.

In addition to using ICD-10 codes to understand the causes of death, some researchers manually review the literal text of death certificates. In recent years, this process has been supplanted by automated analysis of text in electronic versions of the death certificates (22–24). These techniques have allowed researchers to access more information concerning causes of death when the ICD-10 codes lack the detail needed or mask information due to coding rules. Studies have applied these techniques to sudden, unexpected infant deaths (25, 26).

CHALLENGES AND CONTROVERSIES IN CERTIFICATION

Certification of sudden, unexpected death in infants and children can be challenging. Many unwitnessed and initially unexplained deaths remain unexplained despite thorough investigation, full autopsy, and comprehensive laboratory studies. In other cases, multiple possible but unproven causes or abnormal conditions of uncertain lethality are identified. In these cases, it is impossible to determine a single, specific cause of death and the most truthful, accurate, and understandable certification is to enter “Undetermined” in Part I of the death certificate. This strategy should be appropriate for every age and circumstance of unexplained death, yet there is tremendous confusion and variation in certification of unexplained infant and child deaths.

The recent reluctance by many US and some international medical examiners to certify the cause of a sudden, unexplained infant death as Sudden Infant Death Syndrome is two-fold. Firstly, Sudden Infant Death Syndrome is not an etiologically specific cause of death. It is a cluster of circumstances that commonly occur together: an unexplained sudden death, an investigation that fails to identify a cause, and a negative autopsy and ancillary tests. Sudden Infant Death Syndrome is not due to one as-of-yet unidentified etiology. Secondly, some incorrectly believe that infant deaths which remain unexplained after thorough investigation and autopsy are natural in manner, i.e., a natural medical entity. In many cases, factors in the sleep environment with potential for causing asphyxia are identified and are the reason the cause of death remains unexplained and the manner undetermined (i.e., the certifier is unable to distinguish with enough certainty between an accidental suffocation or a natural disease). Sudden, unexplained infant deaths cannot be assumed to be natural in manner.

The shift away from use of the term Sudden Infant Death Syndrome as a certified cause of death has worried many of those who focus on working with these bereaved families, often arguing that families need the acronym to find information and identify with support groups. Fortunately, online sources of information and support for families have shifted their wording as the medicolegal system has focused on more thorough investigation, urged attention to environment, and decreased use of the term Sudden Infant Death Syndrome as a certified cause of death. An Internet search for undetermined or unexplained infant death brings one to similar sites as a search for Sudden Infant Death Syndrome.

Another reason for the variation in certification of sudden unexplained infant and child deaths is the lack of a universally accepted alternative to Sudden Infant Death Syndrome that allows these deaths to be uniquely identifiable for epidemiologic studies through coding yet also be accurate and adhere to the basic principles of certification. Many certification recommendations have been proposed for sudden unexplained infant deaths (3, 27–34) but none have been proposed for the same situation in children ≥365 days old. Certification of infant deaths in the US has undergone many changes in recent years (19, 31, 35–37). The largest change in the US came after the Sudden Unexplained Infant Death Investigation National Training Academies sponsored by the Centers for Disease Control and Prevention were endorsed by the American Board of Medicolegal Death Investigators, National Association of Medical Examiners, and National Sheriff’s Association beginning in 2006. These train-the-trainer sessions were attended by representative medicolegal professionals from all states and relied in part on a publication by Corey et al. titled “NAME Ad Hoc Committee on Sudden Unexplained Infant Death. A functional approach to sudden unexplained infant deaths” (28) and several training materials (Sudden Unexplained Infant Death Investigation. A Systematic Training Program for the Professional Infant Death Investigation Specialist [38], The Infant Death Investigation: Guidelines for the Scene Investigator [39], and Sudden Unexplained Infant Death Reporting Form [40]). The recommendation for certification of deaths consistent with the definition of Sudden Infant Death Syndrome was to report the cause of death as “Sudden Unexplained Infant Death” and the manner as Undetermined (28). Additionally, reporting of conditions, risk factors, and possible external stressors (e.g., focal bronchiolitis, prenatal tobacco exposure, bed-sharing, prone found position, excessive blanketing etc.), when present, in Part II of the death certificate was recommended (41). Certification of cause of death as “sudden unexplained infant death” remains in use by many medical examiners and coroners in the United States, with a roughly equal portion of medical examiners and coroners using “Undetermined”, with or without the risk factors etc. in Part II (unpublished data from a 2018 survey of offices accredited by the National Association of Medical Examiners). Thus, deaths previously certified as Sudden Infant Death Syndrome or Sudden Unexplained Infant Death are increasingly certified as Undetermined or as due to an asphyxial process (31, 36, 42, 43).

Whether risk factors or conditions that may or may not have contributed to an infant death should be entered onto a death certificate remains controversial. In the National Association of Medical Examiners “A Guide for Manner of Death Classification,” conditions causing interpretive difficulties, including focal, natural disease (for example, a focal area of bronchiolitis), or external factors such as sleep environment, may be listed in Part II of the death certificate as “other significant conditions.” However, there is no clear recommendation regarding additional risk factors (e.g., cigarette smoke exposure, bed-sharing, etc.) (30). Proponents of listing risk factors on the death certificate, either in the Part II or “Describe How Injury Occurred” section, argue it is a mechanism to communicate modifiable behaviors (28). The limitation of this approach is that entities appearing in Part II that have corresponding ICD-10 codes (e.g., bronchiolitis, smothering) may cause unintended shifts in ranking of the underlying causes of death, as the death will be coded as being due to the condition that appears in Part II if the Part I cause was a nonspecific condition such as “undetermined.” Others abstain from listing any risk factors on the death certificate because of their uncertain significance for cause of death. A reasonable interpretation of the definition for death certificate Part II (“conditions contributing to death but not resulting in the underlying cause of death in Part I”) suggests that risk factors or circumstances of uncertain significance are not to be entered. The death certificate is not designed as a tool to communicate complex data unrelated to cause for purposes of research. While some certifiers may want to convey the presence or absence of risk factors using the death certificate, specific risk factors should be identified in the autopsy and scene investigation reports, extracted by child fatality review teams or other public health groups, and collated at the national level (31). Such infrastructure is already in place in the US but resources are lacking to expand the number of deaths for which data are collected.

The risk factors that cause the most controversy in certification are unsafe sleep environments with potential for asphyxia. Recommendations for a safe sleeping environment have become quite detailed (44). However, there is no consensus on interpreting the degree and role of an unsafe sleep environment in causing an asphyxiation. Despite the growing literature and prevention efforts to warn caregivers that an unsafe sleep environment increases the risk of sudden death, there remains a paradoxical resistance to having infant deaths certified as asphyxial accidents. Some argue that unsafe sleep environments are merely risk factors for sudden unexplained death that should not be conflated to a cause. It is well known that many unsafe sleep circumstances do result in an asphyxia death; what was a risk factor prior to death becomes the cause of death. While it may be difficult to evaluate the risk of a specific circumstance in a specific case and differentiate between an unexplained death and asphyxia due to the unsafe sleep environment, the task cannot be avoided. Unfortunately, there is no objective test to prove when asphyxiation resulted in death. Recent articles and textbooks describe the conundrum (36, 37, 43, 45).

Arguments are made that asphyxia due to an unsafe sleep environment should be the certified cause of death only when investigation shows “strong evidence of full, external obstruction of both the nose and mouth or external neck or chest compression with a reliable and non-conflicting witnessed account, and no other potentially fatal findings or concerning conditions, and suffocation is probable given the infant age and likely stage of development” (45). If the circumstances fall short of the above diagnostic criteria but asphyxia remains reasonably possible, what cause of death statement results in appropriate coding and still communicates this risk factor for statistical tracking? In many cases, if certified as “sudden unexplained infant death”, the National Center for Health Statistics will code the death as ICD-10 code R95 (Sudden Infant Death Syndrome) (28, 31, 45). If the cause of death statement is “undetermined” the death will be coded as ICD-10 code R99 and not counted, in terms of leading/ranked causes, as an unexplained sudden infant death (see Table C in [46]). Another criticism of “undetermined” as a certified cause of death is that it carries a social stigma of parental blame since it is also used as a manner of death when homicide cannot be excluded. A 2004 report from the United Kingdom stated that use of the similar term, “unascertained”, “has now come to have a stigma attached to it” because it implies the death is suspicious (47, 48). A study from the US also mentions that the shift in cause of death from Sudden Infant Death Syndrome to undetermined has negatively affected surviving families (35). Recommendations for cause of death statements and coding that may alleviate some of these problems are discussed in greater detail below.

THE EFFECT OF LITERAL TEXT ON CODING

As important members of the public health system, certifiers should have a basic understanding of how their death certificates are coded by the National Center for Health Statistics. Cause of death statements are assigned ICD codes according to specific ICD rules. Each code has a title that summarizes the content of that bin of data (e.g., R95. Sudden Infant Death Syndrome; R99. Other ill-defined and unspecified causes of mortality) and instructions for use of the code. At present, ICD-10 is in use and will be for several years. However, ICD-11 is in development and new ICD-11 codes have already been proposed (49, 50). Codes most frequently encountered in cases of sudden, unexplained pediatric deaths are summarized in Table 10.1.

Assignment of the underlying cause code is influenced by the content, sequence, and combination of conditions entered in Part I and Part II of the certificate. If the cause of death is certified as sudden infant death syndrome, sudden unexplained infant death, or unexplained sudden death (any combination that includes the words sudden and death), and is not followed by a better-defined, coded condition in Part I, the underlying cause code will be R95 (MH11 in ICD-11 as proposed) (25, 26). The same is not true for infant deaths certified as undetermined. If an infant cause of death is entered as “undetermined”, the underlying cause code will be R99 (MH14 in ICD-11 as proposed), except when another better-defined, coded condition is present. In which case, “undetermined” is ignored. Examples are shown in Table 10.2.

Table 10.2

Effect of Sequence and Part II on Coding When Cause of Death Includes “Sudden” and “Death”.

When Part I is “Undetermined” and Part II contains an entity with an ICD code, the underlying cause of death will be coded according to the codable entity listed in Part II. This is how inclusion of risk factors with an uncertain contribution to the death may result in a cause of death that does not reflect the certifier’s intent. For example, if the cause of death is listed as Part I: Undetermined, Part II: Risk Factors – Bed-sharing, Prematurity, the case will be coded as P073 for the prematurity (thus being coded and classified as an explained natural death). There are no codes for many known risk factors, including bed-sharing, prone sleep, and tobacco smoke exposure. But overlaying, suffocation, and smothering are coded. Table 10.3 illustrates coding when risk factors or other conditions are listed in Part II of a certificate with undetermined in Part I. More uniform reporting practices and a better understanding of coding may result in a reduction in coding discrepancies and an improvement in the accuracy and reliability of surveillance of sudden, unexpected pediatric deaths.

Table 10.3

Effect of Part II on Coding When Cause of Death is “Undetermined”.

CERTIFICATION OF UNEXPECTED DEATHS THAT BECOME EXPLAINED BY INVESTIGATION AND AUTOPSY

A sudden, unexpected death that becomes explained (a cause of death can be determined) after investigation and autopsy is certified with the etiologically specific disease or injury that caused death, according to standard procedures. An example would be a 3-year-old who collapses, is transported to the emergency room, and subsequently pronounced deceased. Autopsy does not reveal evidence of trauma. The heart appears dilated and flabby, and histology reveals lymphocytic myocarditis. Toxicologic testing is negative. This unexpected death is explained by a single, definitive cause. The death certificate will be certified as cause of death: Lymphocytic myocarditis; manner of death: Natural. Certification when an underlying etiology is identified is not an age-specific process that needs further discussion in this publication, with exception of certification of infant deaths due to asphyxia by extrinsic factors which is described below.

Many times, suffocation (homicidal or accidental) leaves no signs on the body. There are no findings or diagnostic tests at autopsy that can rule in or rule out asphyxia as the mechanism of death (36, 42, 43, 51). In such cases, the diagnosis must be made based upon historical information and scene investigation. An example is a 3-month-old infant sleeping on a couch with an adult. The infant is found dead, wedged between the cushions of the couch, with compression of the chest confirmed by scene investigation and doll reenactment. The death is explained and should be certified as cause of death: Asphyxia due to wedging; manner of death: Accident. In many cases, the level of evidence for an asphyxial death is less conclusive; the certifier must weigh the evidence and offer an opinion concerning cause. With this process comes variability between certifiers, which can be frustrating for families and other professionals using the data.

Prior to this publication, diagnostic criteria have not been established for certification of death by asphyxia due to extrinsic factors. This has resulted in tremendous overlap among fully explained asphyxial deaths and undetermined/unexplained deaths in data for sudden unexpected deaths in infancy. The classification system used by the Centers for Disease Control and Prevention Sudden Unexpected Infant Death Case Registry is a useful model to differentiate explained suffocations from possible suffocations and other sudden, unexplained infant deaths with and without unsafe risk factors (45, 52). A death explained by asphyxia due to an extrinsic cause is one where there is convincing evidence of full, external obstruction of both nose and mouth, or external compression of the neck or chest, and the proposed mechanism causing asphyxia is probable given the infant’s age and stage of development. Death scene investigation, including doll reenactment when appropriate, is essential. A reliably witnessed and detailed account of the obstruction or compression, with no conflicting accounts, may be sufficient for diagnosis in the absence of a doll reconstruction. Other potentially fatal findings or concerning medical conditions are reasonably excluded by thorough autopsy with ancillary testing and review of the medical history. Recommended criteria for certification of an infant death as being caused by an asphyxia etiology are listed in Table 10.4. Absent these criteria, simple presence of bed/sleep-surface sharing and/or prone sleep are considered insufficient to certify the cause of death as due to asphyxia, or even probable asphyxia. While use of “probable”, “possible”, or “consistent with” are appropriate for acknowledging diagnostic uncertainty, these modifiers are ignored by coding rules and the death will be coded as an asphyxia.

Table 10.4

Criteria for Certification of an Infant Death as Being Caused by an Asphyxial Etiology.

CERTIFICATION AND RECOMMENDED CODING OF UNEXPECTED DEATHS THAT REMAIN UNEXPLAINED

Often, multiple, questionably causal factors are identified during investigation and autopsy of a sudden unexpected pediatric death. Thus, the death is unexplained. On rare occasion, competing causes of death (two or more potentially fatal conditions) are revealed and the certifier will need to determine whether the death is certified as undetermined (it cannot be determined which of the two potential causes resulted in death) or if an underlying cause and a contributory cause can be discerned (useful when one cause has higher fatal potential and the two causes may act in synergy).

Causal factors revealed by investigation and/or autopsy typically fall into two categories: intrinsic factors (threats to life originating inside the child’s body) and extrinsic factors (threats to life originating outside of the child’s body). These include known risk factors for sudden unexplained pediatric deaths as well as pathologies and circumstances with fatal potential that are not necessarily published risk factors. In trying to arrive at the cause of death, the certifier integrates all data from the investigation and autopsy and weighs the relative contributions of these intrinsic and extrinsic factors. While the level of evidence for any single factor being the cause of death may be uncertain, it is important to acknowledge (in the autopsy report) and convey (to end-users of the information) the intrinsic and extrinsic factors identified.

Intrinsic and Extrinsic Factors

Intrinsic factors are natural conditions or risk factors associated with abnormal physiology or anatomy that are concerning as contributors to death but are insufficient as a cause (e.g., low birth weight, preterm birth, small for gestational age, concurrent nonlethal illness, history of febrile seizures), or natural conditions of unknown significance (e.g., cardiac channelopathy, or seizure gene variants of unknown significance).

Extrinsic factors are defined as conditions in the child’s immediate environment that are a potential threat to life but cannot be deemed the cause of death with reasonable certainty (e.g., side or prone sleep if unable to roll to supine, over-bundling without documented hyperthermia, objects in immediate sleep environment, sleep environment not specifically designed for infant sleep, soft or excessive bedding, and sleep-surface sharing), injuries or toxicologic findings that are either nonlethal or of unknown lethality, or circumstances/findings otherwise concerning for unnatural death.

Consensus recommendations from the authoring panel (see Chapter 1 for panel creation and composition) attempt to address the above challenges and controversies by providing a certification scheme that may improve surveillance of pediatric deaths by identifying fully investigated, unexplained sudden deaths and allowing tracking of intrinsic and potentially modifiable extrinsic factors. The six recommended cause of death statements, criteria for the certifier to consider, and expected ICD codes are listed in Table 10.5, with examples of their usage in Appendix 7. The cause of death statements in Table 10.5 are recommended literal text for Part I of the death certificate. Part II of the death certificate should be left blank, with the specific intrinsic and extrinsic factors identified being listed in the final diagnoses portion of the autopsy report.

Table 10.5

Recommended Cause of Death Statements, Criteria for the Certifier to Consider, and Resulting ICD Codes.

SUMMARY

To fully realize the epidemiologic benefits of improved pediatric death investigation and autopsy practices, cause of death reporting must also be improved and standardized. Criteria are proposed to strengthen the level of evidence used for determination of asphyxia due to extrinsic factors as cause of death. Sudden, unexpected deaths that remain unexplained after thorough investigation and autopsy and meet criteria above should be certified as “unexplained sudden death” with indication of intrinsic and/or extrinsic factors as appropriate. Use of “undetermined” as the cause of death statement in pediatric deaths should be limited to cases unexplained because investigation, death scene examination, or autopsy were substantially limited, incomplete, or insufficient or the unexplained death was not sudden. If the guidance introduced here achieve broad acceptance and implementation, which can be complimented by the abandonment of confusing acronyms, use of non-stigmatizing terminology, and recognition and tracking of potentially modifiable extrinsic factors, surveillance of sudden pediatric deaths may significantly improve. Such improved surveillance may allow us to better understand the true causes of death, and thus develop effective prevention strategies.

REFERENCES

1.

National Center for Health Statistics. Milestones in national vital statistics. In: Vital statistics: summary of a workshop. Washington: National Academies Press; 2009. p. 89. Available from: https://www​.nap.edu/read​/12714/chapter/9#89.

2.

Spitz WU, Spitz DJ. Spitz and Fisher’s medicolegal investigation of death: guidelines for the application of pathology to crime investigation. 4th ed. Springfield (IL): Charles C. Thomas Pub; 2006. 1325 p.

3.

Hanzlick R. Cause of death and the death certificate: important information for physicians, coroners, medical examiners, and the public. Northfield (IL): College of American Pathologists; 2006. 235 p.

4.

U.S. Standard certificate of death [Internet]. Hyattsville (MD): National Center for Health Statistics; [updated 2003 Nov; cited 2019 Mar 1]. 4 p. Available from: https://www​.cdc.gov/nchs​/data/dvs/death11-03final-acc.pdf.

5.

ODPHP: Office of Disease Prevention and Health Promotion [Internet]. Washington: Department of Health and Human Services; c2018. Healthy people 2020; [updated 2018 Apr 27; cited 2018 Apr 28]. Available from: http://www​.healthypeople.gov/.

6.

Centers for Disease Control and Prevention [Internet]. Atlanta: Centers for Disease Control and Prevention; c2019. Under standing death data; [updated 2018 Aug 24; cited 2019 Apr 4]. Available from: https://www​.cdc.gov/surveillance​/projects​/understanding-death-data.html.

7.

Centers for Disease Control and Prevention [Internet]. Atlanta: U.S. Department of Health & Human Services; c2019. National vital statistics system: mortality data; [updated 2019 May 7; cited 2019 Apr 4]. Available from: https://www​.cdc.gov/nchs/nvss/deaths.htm.

8.

ICD-10: International statistical classification of diseases and related health problems. 2nd ed. Geneva (Switzerland): World Health Organization; 2004. 125 p.

9.

Centers for Disease Control and Prevention [Internet]. Atlanta: Department of Health & Human Services; c2018. National vital statistics system: instruction manuals; [updated 2017 Aug 30; cited 2018 Apr 28]. Available from: https://www​.cdc.gov/nchs​/nvss/instruction_manuals.htm.

10.

Mitchell EA. Sleeping position of infants and the sudden infant death syndrome. Acta Paediatr Suppl. 1993 Jun; 82 Suppl 389:26–30. PMID: 8374186. [PubMed: 8374186]

11.

Fleming PJ, Gilbert R, Azaz Y, et al. Interaction between bedding and sleeping position in the sudden infant death syndrome: a population based case-control study. BMJ. 1990 Jul 14; 301(6743):85–9. PMID: 2390588. PMCID: PMC1663432. https://doi​.org/10.1136/bmj.301.6743.85. [PMC free article: PMC1663432] [PubMed: 2390588]

12.

Mitchell EA, Ford RP, Taylor BJ, et al. Further evidence supporting a causal relationship between prone sleeping position and SIDS. J Paediatr Child Health. 1992; 28 Suppl 1:S9–12. PMID: 1524882. https://doi​.org/10.1111/j​.1440-1754.1992.tb02732.x. [PubMed: 1524882]

13.

Beal SM, Finch CF. An overview of retrospective case-control studies investigating the relationship between prone sleeping position and SIDS. J Paediatr Child Health. 1991 Dec; 27(6):334–9. PMID: 1836736. https://doi​.org/10.1111/j​.1440-1754.1991.tb00414.x [PubMed: 1836736]

14.

Kattwinkel J, Brooks J, Keenan ME, Malloy M. Infant sleep position and the sudden infant death syndrome (SIDS) in the United States: joint commentary from the American Academy of Pediatrics and selected agencies of the federal government. Pediatrics. 1994 May; 93(5):820. PMID: 8165086. [PubMed: 8165086]

15.

Willinger M, Hoffman HJ, Wu KT, et al. Factors associated with the transition to nonprone sleep positions of infants in the United States: the National Infant Sleep Position Study. JAMA. 1998 Jul 22–29; 280(4):329–35. PMID: 9686549. https://doi​.org/10.1001/jama.280.4.329. [PubMed: 9686549]

16.

Colson ER, Rybin D, Smith LA, et al. Trends and factors associated with infant sleeping position: the National Infant Sleep Position Study, 1993–2007. Arch Pediatr Adolesc Med. 2009 Dec; 163(12):1122–8. PMID: 19996049. PMCID: PMC2898125. https://doi​.org/10.1001/archpediatrics​.2009.234. [PMC free article: PMC2898125] [PubMed: 19996049]

17.

Erck Lambert AB, Parks SE, Shapiro-Mendoza CK. National and state trends in sudden unexpected infant death: 1990–2015. Pediatrics. 2018 Mar;141(3). pii: e20173519. PMID: 29440504. https://doi​.org/10.1542/peds.2017-3519. [PMC free article: PMC6637428] [PubMed: 29440504]

18.

Malloy MH, MacDorman M. Changes in the classification of sudden unexpected infant deaths: United States, 1992–2001. Pediatrics. 2005 May; 115(5):1247–53. PMID: 15867031. https://doi​.org/10.1542/peds.2004-2188. [PubMed: 15867031]

19.

Shapiro-Mendoza CK, Tomashek KM, Anderson RN, Wingo J. Recent national trends in sudden, unexpected infant deaths: more evidence supporting a change in classification or reporting. Am J Epidemiol. 2006 Apr 15; 163(8):762–9. PMID: 16582034. https://doi​.org/10.1093/aje/kwj117. [PubMed: 16582034]

20.

Shapiro-Mendoza CK, Parks SE, Brustrom J, et al. Variations in cause-of-death determination for sudden unexpected infant deaths. Pediatrics. 2017 Jul; 140(1). pii: e20170087. PMID: 28759406. PMCID: PMC5599098. https://doi​.org/10.1542/peds.2017-0087. [PMC free article: PMC5599098] [PubMed: 28759406]

21.

Mathews TJ, MacDorman MF, Thoma ME. Infant mortality statistics from the 2013 period linked birth/infant death data set. Natl Vital Stat Rep. 2015 Aug 6; 64(9):1–30. PMID: 26270610. [PubMed: 26270610]

22.

Trinidad JP, Warner M, Bastian BA, et al. Using literal text from the death certificate to enhance mortality statistics: characterizing drug involvement in deaths. Natl Vital Stat Rep. 2016 Dec; 65(9):1–15. PMID: 27996933. [PubMed: 27996933]

23.

Koopman B, Zuccon G, Nguyen A, et al. Automatic ICD-10 classification of cancers from free-text death certificates. Int J Med Inform. 2015 Nov; 84(11):956–65. PMID: 26323193. https://doi​.org/10.1016/j​.ijmedinf.2015.08.004. [PubMed: 26323193]

24.

Holman RC, Belay ED, Christensen KY, et al. Human prion diseases in the United States. PLoS One. 2010 Jan 1; 5(1):e8521. PMID: 20049325. PMCID: PMC2797136. https://doi​.org/10.1371/journal​.pone.0008521. [PMC free article: PMC2797136] [PubMed: 20049325]

25.

Kim SY, Shapiro-Mendoza CK, Chu SY, et al. Differentiating cause-of-death terminology for deaths coded as sudden infant death syndrome, accidental suffocation, and unknown cause: an investigation using US death certificates, 2003–2004. J Forensic Sci. 2012 Mar; 57(2):364–9. PMID: 21981558. https://doi​.org/10.1111/j​.1556-4029.2011.01937.x. [PubMed: 21981558]

26.

Shapiro-Mendoza CK, Kim SY, Chu SY, et al. Using death certificates to characterize sudden infant death syndrome (SIDS): opportunities and limitations. J Pediatr. 2010 Jan; 156(1):38–43. PMID: 19782997. https://doi​.org/10.1016/j​.jpeds.2009.07.017. [PubMed: 19782997]

27.

Beckwith JB. Defining the sudden infant death syndrome. Arch Pediatr Adolesc Med. 2003 Mar 1; 157(3):286–90. PMID: 12622679. https://doi​.org/10.1001/archpedi​.157.3.286. [PubMed: 12622679]

28.

Corey TS, Hanzlick R, Howard J, et al. A functional approach to sudden unexplained infant deaths. Am J Forensic Med Pathol. 2007 Sep 28(3):271–7. PMID: 17721183. https://doi​.org/10.1097/01​.paf.0000257385.25803.cf. [PubMed: 17721183]

29.

Bajanowski T, Brinkmann B, Vennemann M. The San Diego definition of SIDS: practical application and comparison with the GeSID classification. Int J Legal Med. 2006 Nov; 120(6):331–6. PMID: 16237562. https://doi​.org/10.1007​/s00414-005-0043-0. [PubMed: 16237562]

30.

Hanzlick R, Hunsaker JC, Davis GJ. A guide for manner of death classification. Atlanta: National Association of Medical Examiners; 2002. 29 p.

31.

Nashelsky MB, Pinckard JK. The death of SIDS. Acad Forensic Pathol. 2011 Jul; 1(1): 92–9. https://doi​.org/10.23907/2011.010.

32.

Blair PS, Byard RW, Fleming PJ. Sudden unexpected death in infancy (SUDI): suggested classification and applications to facilitate research activity. Forensic Sci Med Pathol. 2012 Sep; 8(3):312–5. PMID: 22076788. https://doi​.org/10.1007​/s12024-011-9294-x. [PubMed: 22076788]

33.

Randall B, Donelan K, Koponen M, et al. Application of a classification system focusing on potential asphyxia for cases of sudden unexpected infant death. Forensic Sci Med Pathol. 2012 Mar; 8(1):34–9. PMID: 22076787. https://doi​.org/10.1007​/s12024-011-9291-0. [PubMed: 22076787]

34.

Randall BB, Wadee SA, Sens MA, et al. A practical classification schema incorporating consideration of possible asphyxia in cases of sudden unexpected infant death. Forensic Sci Med Pathol. 2009 Dec; 5(4): 254–60. PMID: 19484508. PMCID: PMC3274765. https://doi​.org/10.1007​/s12024-009-9083-y. [PMC free article: PMC3274765] [PubMed: 19484508]

35.

Crandall LG, Reno L, Himes B, Robinson D. The diagnostic shift of SIDS to undetermined: are there unintended consequences? Acad Forensic Pathol. 2017 Jun; 7(2):212–20. https://doi​.org/10.23907/2017.022. [PMC free article: PMC6474538] [PubMed: 31239975]

36.

Matshes EW, Lew EO. An approach to the classification of apparent asphyxial infant deaths. Acad Forensic Pathol. 2017 Jun; 7(2):200–11. https://doi​.org/10.23907/2017.021. [PMC free article: PMC6474537] [PubMed: 31239974]

37.

Duncan JR, Byard RW, editors. SIDS -- Sudden infant and early childhood death: the past, the present and the future. Adelaide (Australia): University of Adelaide Press; 2018. 830 p. [PubMed: 30024688]

38.

Sudden unexplained infant death investigation: a systematic training program for the professional infant death investigation specialist [Internet]. Atlanta: Centers for Disease Control and Prevention; [date unknown] [cited 2018 Jun 28]. 248 p. Available from: http://www​.suidi.org​/training/SUIDITextbook.pdf.

39.

Hanzlick RL, Jentzen JM, Clark SC. Sudden, unexplained infant death investigation, infant death investigation: guidelines for the scene investigator [Internet]. Atlanta: Department of Health and Human Services; 2007 Jan [cited 2018 May 6]. 39 p. Available from: https://www​.cdc.gov/sids​/pdf/508suidiguidelinessingles​_tag508.pdf.

40.

Centers for Disease Control and Prevention [Internet]. Sudden unexpected infant death and sudden infant death syndrome: SUIDI reporting form. Atlanta: Department of Health & Human Services; [updated 2018 Feb 13; cited 2018 May 6]. Available from: https://www​.cdc.gov/sids/SUIDRF.htm.

41.

Hanzlick R. Sudden unexplained infant death investigation: a systematic training program for the professional infant death investigation specialist [Internet]. Atlanta: Centers for Disease Control and Prevention; [date unknown]. Chapter 9, Certification of unexplained infant deaths: cause of death and the death certificate; [cited 2018 Jun 28]; p. 212–27. Available from: http://www​.suidi.org​/training/SUIDITextbook.pdf.

42.

Hunt CE, Darnall RA, McEntire BL, Hyma BA. Assigning cause for sudden unexpected infant death. Forensic Sci Med Pathol. 2015 Jun; 11(2):283–8. PMID: 25634430. PMCID: PMC4415994. https://doi​.org/10.1007​/s12024-014-9650-8. [PMC free article: PMC4415994] [PubMed: 25634430]

43.

Pasquale-Styles MA, Regensburg M, Bao R. Sudden unexpected infant death certification in New York City: intra-agency guideline compliance and variables that may influence death certification. Acad Forensic Pathol. 2017 Dec; 7(4):536–50. https://doi​.org/10.23907/2017.046. [PMC free article: PMC6474437] [PubMed: 31240005]

44.

Task Force on Sudden Infant Death Syndrome. SIDS and other sleep-related infant deaths: updated 2016 recommendations for a safe infant sleeping environment. Pediatrics. 2016 Nov; 138(5):e20162938. PMID: 27940804. https://doi​.org/10.1542/peds.2016-2938. [PubMed: 27940804]

45.

Shapiro-Mendoza CK, Camperlengo L, Ludvigsen R, et al. Classification system for the sudden unexpected infant death case registry and its application. Pediatrics. 2014 Jul; 134(1):e210–9. PMID: 24913798. PMCID: PMC4311566. https://doi​.org/10.1542/peds.2014-0180. [PMC free article: PMC4311566] [PubMed: 24913798]

46.

ICD-10 Underlying cause-of-death lists for tabulating mortality statistics: instruction manual [Internet]. Hyattsville (MD): Department of Health and Human Services; 2017 Jan [cited 2019 Apr 5]. 67 p. Available from: https://www​.cdc.gov/nchs​/data/dvs/Part9InstructionManual2006.pdf.

47.

Working Group Convened by the Royal College of Pathologists and the Royal College of Paediatrics and Child Health. Sudden unexpected death in infancy and childhood: multi-agency guidelines for care and investigation. London (UK): Royal College of Pathologists and the Royal College of Paediatrics and Child Health; 2004.

48.

Crisp E, Julious SA. The analysis of the use of ‘unascertained’ for sudden unexpected deaths in infancy from 1988 to 2010. Arch Dis Child. 2014 Mar; 99(3):300–1. PMID: 24265413. https://doi​.org/10.1136​/archdischild-2013-305196. [PubMed: 24265413]

49.

ICD-11 Maintenance Platform [Internet]. Geneva (Switzerland): World Health Organization; 2018 Dec 9 [cited 2018 Dec 10]. Available from: https://icd​.who.int/dev11/l-m/en.

50.

The 11th Revision of the International Classification of Diseases (ICD-11) is due by 2018! [Internet]. Geneva (Switzerland): World Health Organization; 2018 Mar 16 [cited 2018 Dec 10]. Available from: https://www​.who.int/classifications​/icd/revision/en/.

51.

Brown TT, Batalis NI, McClain JL, et al. A retrospective study of the investigation of homicidal childhood asphyxial deaths. J Forensic Sci. 2018 Jul;63(4):1160–1167. PMID: 29044506. https://doi​.org/10.1111/1556-4029.13666. [PubMed: 29044506]

52.

Shapiro-Mendoza CK, Camperlengo LT, Kim SY, Covington T. The sudden unexpected infant death case registry: a method to improve surveillance. Pediatrics. 2012 Feb; 129(2):e486–93. https://doi​.org/10.1542/peds.2011-0854. [PubMed: 22232303]

*

Disclaimer: The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.