Customize your JAMA Network experience by selecting one or more topics from the list below.
Herman-Giddens ME, Smith JB, Mittal M, Carlson M, Butts JD. Newborns Killed or Left to Die by a Parent: A Population-Based Study. JAMA. 2003;289(11):1425–1429. doi:10.1001/jama.289.11.1425
Author Affiliations: North Carolina Child Advocacy Institute, Raleigh, and Department of Maternal and Child Health, University of North Carolina School of Public Health, Chapel Hill (Dr Herman-Giddens); Injury Prevention Research Center, University of North Carolina, Chapel Hill (Mr Smith); State Center for Health Statistics, Raleigh, NC (Dr Mittal); and Office of the Chief Medical Examiner, Chapel Hill, NC (Ms Carlson and Dr Butts).
Context Interest in the discarding or killing of newborns by parents has increased
due to wide news coverage and efforts by states to provide Safe Haven legislation
to combat the problem.
Objective To describe the characteristics of these cases in North Carolina.
Design, Setting, and Population Case series derived from data on all deaths among liveborn infants 0
to 4 days of age reported to the North Carolina medical examiner from 1985
Main Outcome Measures Incidence of newborns known to have been killed or discarded by a parent;
epidemiological characteristics of newborns and parents.
Results There were 34 newborns known to have been killed or discarded by a parent,
comprising 0.002% of all liveborn infants during the 16-year study period,
giving a rate of 2.1 per 100 000 per year. A total of 58.8% were male,
41.1% were white, and 52.9% were black. For 29 cases, the perpetrator was
determined to be the mother. Among mothers, 50% were single and 20.6% were
married (marital status of the remainder was unknown). Thirty-five percent
had had other children. Eight mothers (23.5%) were known to have received
some prenatal care. The mean age of the mothers was 19.1 years (range, 14-35
years) and more than half were aged 18 years or older. The most common causes
of death were asphyxiation/strangulation (41.1%) and drowning (26.5%).
Conclusions In North Carolina, at least 2.1 per 100 000 newborns are known
to be killed or left to die per year, usually by their mothers. It is unknown
how many of these deaths might be prevented by Safe Haven laws. Efforts to
educate the public about these laws need to target the general public. Where
resources are limited, the focus should be on on adolescent pregnancy prevention
programs, young adults, prenatal care clinics, and married women.
The risk of homicide on the first day of life (neonaticide) is 10 times
greater than the rate during any other time of life.1 Neonaticide
has been attributed to motives and circumstances associated with the mother's
perceived need to keep the pregnancy and birth unknown to her family and associates.2,3 Theories about the causes of this behavior
include shame about an unwanted or illegitimate infant and/or the obvious
sexual activity, mental illness,4 difficult
relationships with parents,2 immaturity,5 and a culture that promotes self-destructive impulses.6 The mother may deny to herself that she was pregnant,
and thus be surprised by the birth.5,7,8 Public
and professional conceptions of the situations involving these cases are often
based on the stereotype of the unwed adolescent girl, especially one who is
from a religious family or feels she cannot confide in her parents.2,9 As of 2002, 42 states have passed "Safe
Haven" or "Safe Surrender" laws to prevent neonaticide (William L. Pierce,
PhD, written communication, February 21, 2003).10 These
laws allow parents to anonymously transfer an unwanted newborn to certain
persons or authorities, such as hospitals, without being charged with infant
To estimate the incidence of newborn infants killed or discarded and
left to die, we investigated all deaths of newborns 0 to 4 days of age in
North Carolina and characterized the perpetrators and circumstances of these
deaths to inform Safe Haven policy.
We reviewed the case records of all deaths regardless of cause of all
liveborn infants 4 days of age or younger listed in the North Carolina Medical
Examiner (ME) database for the 16-year period from January 1, 1985, through
December 31, 2000.13 The ME case files include
the death certificate, the name of involved law enforcement officers, and
the ME's report, which contains details of the circumstances surrounding the
death, the autopsy, and any toxicology studies. The manner of death (homicide,
accident, natural, suicide, undetermined) was determined for each case by
the Office of the Chief Medical Examiner. The designation of a death as a
homicide depended on a determination made by the ME that the death was the
consequence of deliberate actions on the part of the perpetrator to cause
harm to the infant. Intent was inferred in most cases from scene and historical
information or, in some instances, from statements made to the authorities
by the perpetrator. It was assumed that the parent(s) of a newborn would know
that some level of immediate care and support to the infant was necessary
and that an apparent failure to provide that care to the infant constituted
a deliberate act, albeit of omission. In instances where there was serious
question in regard to whether or not the parent was able to appreciate the
need for such care because of mental impairment, homicide was not ruled. In
one instance not ruled a homicide, the infant had what would have been a rapidly
(if not immediately) lethal cardiac anomaly, so it was inconclusive that lack
of care or abandonment caused the death.
For liveborn infants, all homicide cases and cases determined by our
review to be deaths related to discarding of the newborn and committed or
likely to have been committed (from the circumstances) by a parent regardless
of manner were coded into the data set. Variables abstracted included the
victim's name, date of birth, age, sex, race, county of residence, and death;
date, time, circumstances, and location of death; and location of the body,
the manner of death, and the means of death (such as suffocation, stabbing,
or drowning). Medical examiner narratives were reviewed for perpetrator and
other pertinent information. When necessary, law enforcement officials were
contacted for additional perpetrator information. Homicide cases were further
reviewed to identify those where neither parent was noted as a perpetrator.
We supplemented ME data with data obtained from linked birth and death files
from the State Center for Health Statistics. Variables extracted from the
birth certificate part of the infant death file consisted of the mother's
name, age, race, marital status, identity of the father, number of live children,
children now dead, last menstrual period, plurality, and month prenatal care
began. Statistical analysis was performed using SAS, version 8.2 (SAS Institute
Inc, Cary, NC).
Thirty-four cases of newborn deaths by homicide or by discarding where
the perpetrator was either confirmed to be a parent (n = 29) or likely to
have been a parent (n = 5) were identified over the 16-year period. Thirty-one
cases were ruled homicides and 3 deaths were ruled either as accidental or
natural but related to conditions of discarding and lack of providing care
for the newborn (Table 1). During
this time period, North Carolina had a total of 1 644 718 liveborn
infants; therefore, 0.002% of all liveborn infants were known to have been
killed or discarded, for a rate of 2.1 per 100 000 newborns per year.
In 2001, there were 4 028 000 liveborn infants in the United States.
Therefore, if the United States as a whole had a similar proportion of newborns
actively or passively killed, as many as 85 lives per year might have been
saved if Safe Haven laws were available in all states and in the unlikely
event that all parents of these infants took advantage of these laws.
Table 2 and Table 3 show the characteristics of the newborns, circumstances
of their deaths, mothers' characteristics, and fathers' identities. Newborn
males comprised 58.8% of the cases and females 35.3% of the cases, with 5.9%
indeterminate due to advanced decomposition of the bodies. Testing for differences
between the population proportion, the male to female proportion was statistically
significant (test statistic: z = 1.99, P = .045). Of all newborns in this 16-year period, 51% were female
and 49% were male. Forty-one percent of the infants were white. Black infants,
52.9% of the cases, were overrepresented compared with the overall population
of newborns, of whom approximately 28% were black. Data were not available
for Hispanic ethnicity until 1988. None of the cases from 1988 through 2000
were known to be Hispanic. The mean age of the mothers was 19.1 years (range,
14-35 years), and 20.6% were married. The mother was determined to be the
perpetrator in 29 of the 34 cases. Information was available about prenatal
care for 17 of the cases. Eight of the 17 had received prenatal care. Thirty-five
percent of the infants were second or third born. Although it is not a routinely
collected variable, in 8 cases (23.5%), the ME narratives noted that the mother
denied being pregnant. Analysis of the legal outcomes is beyond the scope
of this study, but we noted where we were able to obtain information on legal
outcomes that sentences of perpetrators to imprisonment varied from none to
In the process of case reviews, we found 7 other cases coded as homicides
of infants aged 0 to 4 days that did not meet the inclusion criteria for the
study (Table 4). These 7 cases
were not included in further analyses.
Over a 16-year period in North Carolina, we report on 34 newborns known
to be killed or discarded in the first day of life by (or likely by) the parents,
an annual rate of 2.1 per 100 000 newborns. The majority of the cases
(91.2%) were ruled homicides, while 3 were found to be unintentional or due
to natural causes. Mothers were the perpetrators in all cases where a perpetrator
was positively identified. Approximately one fifth were married and a similar
proportion had received some prenatal care. To our knowledge, this study is
the only population-based case series that has been conducted in the United
States. A Rio de Janeiro case series from 1900 to 1995 identified a total
of 72 newborn homicides, of which 53 were determined to be homicides by the
mothers of the newborns.14 Cases not ruled
homicide were not studied. Twelve percent of those mothers (mean age, 22.5
years; range, 17-48 years) were married, a lower percentage than in this study.
As we found, there were more male victims than female (56.6% vs 43.4%). In
our study, about one third of the mothers were 21 years of age or older. Studies
and news reports tend to focus on adolescent girls.2,15,16 Moreover,
35% of our cases were second or third pregnancies. Perhaps the most unexpected
finding was that 20.6% of the mothers were married and slightly more had received
some prenatal care, albeit usually late. These findings suggest that targets
for educational programs about Safe Haven options need to be populationwide
(rather than focused exclusively on pregnant adolescents) and that family
planning and prenatal clinics be targeted specifically.
The phenomenon of mothers needing to surrender newborns anonymously
became such a problem in Europe in the Middle Ages17 that
many churches contained slots where the mother could slip the baby in and
leave unseen, perhaps after ringing a bell. This practice of "baby drops"
has recently been reinstituted in California and several other countries,
including Germany and South Africa.10,11
Our study is limited by several factors. Even though all known neonaticides
or deaths due to discarding within a 16-year period were identified, small
numbers and missing values may restrict generalizability to the United States
as a whole. Our study lacked some information about the mothers such as their
mental state at the time of the birth, substance abuse or mental health history,
socioeconomic profile, or whether they were already under observation by the
state child protective services system. Therefore, it was not possible to
estimate how many of the mothers might have been willing to take advantage
of a Safe Haven law, had one been in effect. Finally, the number of newborns
killed or discarded may actually be higher since it is unknown how many deaths
may have occurred without ever being discovered.
In conclusion, this study provides an opportunity to understand the
epidemiology of cases of active or passive killing of newborns by their parents.
Data from this study may aid prevention efforts such as Safe Haven programs
so that states, health departments, adolescent pregnancy prevention programs,
and other related organizations can use their limited resources more effectively.
In addition, for known cases, the study provides an upper estimate for the
number of lives that could be saved under Safe Haven programs in the event,
although unlikely, that all susceptible parents participated.