Epidemiology of child homicide in Jefferson County, Alabama

Child Abuse & Neglect 27 (2003) 1063–1073

Epidemiology of child homicide in Jefferson County, Alabama Jacquelyn M. Lyman a , Gerald McGwin Jr. a,b,c,∗ , Darren E. Malone b , Allison J. Taylor b , Robert M. Brissie d,e , Gregory Davis d,e , Loring W. Rue III b,c a

Department of Epidemiology, School of Public Health, University of Alabama at Birmingham, Birmingham, AL, USA b Center for Injury Sciences, University of Alabama at Birmingham, Birmingham, AL, USA c Section of Trauma, Burns, and Surgical Critical Care, Department of Surgery, School of Medicine, University of Alabama at Birmingham, Birmingham, AL, USA d Coroner/Medical Examiner Office, Jefferson County, Birmingham, AL, USA e Department of Pathology, Forensic Division, School of Medicine, University of Alabama at Birmingham, Birmingham, AL, USA Received 28 May 2002; received in revised form 17 December 2002; accepted 17 December 2002

Abstract Objective: The purpose of this study was to present the epidemiology of homicide among children younger than 6 years of age in Jefferson County, Alabama. This study focused on obtaining great detail on homicides and suspicious deaths occurring within a fixed population. Methods: For purposes of this study, cases included Jefferson County deaths attributed to “homicide” or that were “undetermined” as noted in the coroner files among children younger than 6 years of age who were born and died between January 1, 1988 and December, 31, 1998. Victim and offender characteristics were obtained from the Jefferson County Coroner/Medical Examiner Office records. Environmental factors and circumstances surrounding the death were noted as well. Results: The 53 study subjects were mainly female (55%), Black (69%), younger than 2 years of age (85%), had single mothers (38%), and a history of abuse (53%). Offenders were more likely to be male (64%), Black (73%), and a parent of the victim (53%). Homicides primarily resulted from an angry impulse (61%), with hands the most common weapon (61%). Conclusions: The majority of deaths in this study occurred among children younger than age 2, with a high proportion of fatalities among Black children of unmarried mothers. The offender most often knew the victim, with half of all homicides and two-thirds of all infant homicides involving a parent. ∗

Corresponding author address: 120 Kracke Building, 1922 7th Avenue South, Birmingham, AL 35294-0016, USA. 0145-2134/$ – see front matter © 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S0145-2134(03)00166-2

1064

J.M. Lyman et al. / Child Abuse & Neglect 27 (2003) 1063–1073

More than half of the homicides resulted from an angry impulse, while the most common scenario for deaths with undetermined intent involved the caretaker finding the child unresponsive. © 2003 Elsevier Ltd. All rights reserved. Keywords: Epidemiology; Homicide; Children; Coroner data

Introduction The United States has the highest rates of childhood homicide, suicide, and firearm-related deaths among industrialized countries (Centers for Disease Control and Prevention, 1997). During the 1980s homicide surpassed motor vehicle crashes as the leading cause of infant injury death (Hicks & Gaughan, 1995), accounting for almost one-third of such fatalities in the US by 1996 (Ventura, Peters, Martin, & Maurer, 1997). In 1998, homicide was the twelfth leading cause of death among infants and the fourth leading cause among children age 1 to 5 years (Centers for Disease Control and Prevention, 2001; National Center for Health Statistics, 2000). Homicide rates in the US differ according to age and geographic location (Jason, Gilliland, & Tyler, 1983; McClain et al., 1994; National Center for Health Statistics, 2000). The overall US homicide rate in 1998 was 6.8 per 100,000 persons, compared with Alabama’s rate of 11.1 (National Center for Health Statistics, 2000). Among children, homicide occurs most frequently during the first year of life, with similar or higher rates occurring only during later adolescence (National Center for Health Statistics, 2000; Overpeck, Brenner, Trumble, Trifiletti, & Berendes, 1998). According to Jason et al. (1983), homicide among infants aged 1 week or less occurred in the US more often in rural areas than did other types of child homicide, with the regional distribution higher in the Northeast and South. Homicide among infants aged 1 week to 1 year was highest in the North Central Region. McClain et al. (1994) estimated the occurrence of fatal child abuse or neglect among children age 5 years or younger for each state in the US from 1979 to 1988. Rates were highest in the South and West and lowest in the Northeast. Alabama ranked nineteenth, with a fatal child abuse or neglect rate of 11.2 per 100,000 persons younger than 6 years; the US estimate was 10.4. In 2000, approximately 879,000 children were found to be victims of child maltreatment (National Clearinghouse on Child Abuse and Neglect Information, 2002). One review of medical examiner reports suggests that 80–90% of homicides among children younger than 6 years of age result from child abuse (Herman-Giddens et al., 1999). More than 90% of fatal child abuse cases are among children younger than 5 years old, with more than 40% occurring in infants (McClain et al., 1994). Prior studies have shown that both homicides and fatal cases of child abuse are undercounted (Christoffel, Zieserl, & Chiaramonte, 1985; Herman-Giddens et al., 1999; Jason, 1983a). This underascertainment is due to restrictions and inaccuracies in coding causes of death, incomplete or inaccurate information on death certificates and police reports, variable case definitions, and the absence of a national system for reviewing child homicides. Hall, Reyes, Meller, and Stein (1993) reviewed medical examiner records to determine the causes of deaths due to traumatic injury in children younger than 16 years of age. Among

J.M. Lyman et al. / Child Abuse & Neglect 27 (2003) 1063–1073

1065

infants, homicide was the leading cause of traumatic death and accounted for nearly half of all deaths in the study. Collins and Nichols (1999) reviewed forensic cases of children younger than 6 years old to develop a profile of pediatric homicide. The majority of the victims were male, Black, and 1 year of age or younger, with head trauma the most common cause of death. The offender was usually known to the victim, frequently a family member, and often female. Other studies have shown that parents or stepparents carry out most infanticides, and a slight majority of offenders are male (Department of Health and Human Services, National Center on Child Abuse and Neglect, 1997; Jason, 1983b; Kunz & Bahr, 1996). The purpose of this study is to present the epidemiology of homicide among children younger than 6 years of age in Jefferson County, Alabama. This study focused on obtaining great detail on homicide cases from medical examiner files. Much of the information obtained would not have been available if data were based solely on death certificates. Some previous studies based on medical examiner data have focused on specific causes of death or determination of fatality rates. Results from this study may assist Jefferson County and other similar geographic areas in the understanding of the problem and allocation of resources to at-risk populations.

Methods Deaths included in the study were cases investigated by the Jefferson County Coroner/Medical Examiner Office (JCCMEO) between January 1, 1988 and December 31, 1998 where the manner of death was listed as “homicide” or “undetermined.” Previous studies have demonstrated that deaths with undetermined intent have characteristics similar to homicides (Overpeck et al., 1999; Sorenson, Shen, & Kraus, 1997). Therefore, inclusion of these cases in this study is consistent with the methods of other studies of homicide and traumatic injury in children (Cummings, Theis, Mueller, & Rivara, 1994; Emerick, Foster, & Campbell, 1986; McClain et al., 1994; Overpeck et al., 1998; Siegel et al., 1996; Vanamo, Kauppi, Karkola, Merikanto, & Räsänen, 2001). Cases were included in the study if the decedent was (a) born and died between January 1, 1988 and December 31, 1998, (b) younger than 6 years of age at the time of death, and (c) a resident of Jefferson County, Alabama at the time of death. Six fetal deaths were excluded because they were not considered homicides in this jurisdiction. During the study time period, the JCCMEO was operating under a consistent medical examiner’s statute. All deaths investigated by the office were certified by one of five forensic pathologists. The medical examiner’s statute charges the office with the responsibility of investigating all sudden and unexpected deaths that have occurred in Jefferson County. An autopsy is performed for every homicide victim and for any individual for whom careful investigation and external examination cannot determine a reasonable probable cause of death. Case files were reviewed and included investigative reports, autopsy protocols, toxicology and serology reports, medical records, and police reports. The following information was obtained for each case when available: subject characteristics (age, sex, race), factors related to the death (cause of death, manner of death, types of injuries sustained, toxicological and serological findings, autopsy findings), and environmental factors (mother’s marital status,

1066

J.M. Lyman et al. / Child Abuse & Neglect 27 (2003) 1063–1073

history of prior abuse or neglect). In cases where the manner of death was ruled a homicide, additional information was collected on offender characteristics (age, sex, race, relationship to the victim) and crime related factors (type of weapon used, method of assault, and circumstances of the incident). For cases where manner of death was undetermined, a description of the circumstances surrounding the death was recorded. Homicide cases were categorized using a detailed classification system of filicide subtypes (Guileyardo, Prahlow, & Barnard, 1999). Filicide is defined as the killing of a child by one or both parents. This classification provides a framework of motives and precipitating factors leading to filicide. Many of the categories also apply to general cases of child homicide in which the offender is not the parent. The main categories used to classify homicides in this study include “angry impulse,” “innocent bystander,” “negligence and neglect,” and “unwanted pregnancy or neonaticide.” Homicides due to an angry impulse occur during a beating or other violent outburst. These children may be the victims of a single incident or multiple nonfatal episodes of violence that precede death. In the innocent bystander subtype, the child is inadvertently killed during an attempt to harm someone else. Death of a viable fetus due to death of the mother is included in this category. Negligence has been defined as conduct that falls below the standard established by law for the protection of others against unreasonable risk of harm; neglect can be defined as prolonged failure to provide adequate care. These homicides occur by acts of commission or omission that lie outside the behavior of a reasonable person. Neonaticide is defined as the killing of a child by its parent within the first 24 hours of life (Resnick, 1970). Standard statistical parameters were used to describe subject and offender characteristics. Means and standard deviations were calculated for continuous variables, and frequency distributions and proportions were computed for categorical variables. A comparison of characteristics of homicides and deaths of undetermined intent was made with corresponding p values computed using Pearson’s Chi-square. If expected cell frequencies were less than five, Fisher exact test was used.

Results Birmingham, the largest city in Alabama, is located in Jefferson County, which is the most populous county in Alabama. The total population in 1990 was 651,525, with approximately 64% of the population White and 35% Black. Among children younger than age 6, 56% were White and 43% were Black. In 1994, 38.5% of all births were to unmarried women. There were 53 eligible cases in Jefferson County between 1988 and 1998. Twenty-eight cases (53%) were classified as homicides, and 25 were deaths where intent could not be determined. Table 1 displays demographic and environmental characteristics of all study subjects. Among the 30 victims who died before their first birthday, 27% (8) died during the first week of life. Of the eight victims who died during the first week of life, five died on the day they were born. Those who died during infancy (less than 1 year of age) were predominantly female (19, 63%) and Black (19, 63%). Marital status of the parent was noted in 42 of the 53 cases. The highest proportion of fatalities was among Black children of unmarried mothers (20, 48%), followed by White children of unmarried mothers (10, 24%), and White children with married mothers (7, 17%).

J.M. Lyman et al. / Child Abuse & Neglect 27 (2003) 1063–1073

1067

Table 1 Demographic and environmental characteristics of subjects Characteristic

Total

Homicide

Undetermined

p value

N

%

N

%

N

%

Total

53

100.0

28

100.0

25

100.0

Sex Male Female

24 29

45.3 54.7

11 17

39.3 60.7

13 12

52.0 48.0

.35

Race White Black

17 36

31.2 68.8

9 19

32.1 67.9

8 17

32.0 68.0

.99

Age 1 day 2–365 days 1 year 2–5 years

5 25 15 8

9.4 47.2 28.3 15.1

3 9 11 5

10.7 32.1 39.3 17.9

2 16 4 3

8.0 64.0 16.0 12.0

.13a

Mother’s marital status Married Living with boyfriend/girlfriendb Single

12 14 16

28.6 33.3 38.1

6 9 10

24.0 36.0 40.0

6 5 6

35.3 29.4 35.3

.79a

History of abuse/neglect Notation in file No notation

28 25

52.8 47.2

17 11

60.7 39.3

11 14

44.0 56.0

.22

a b

Fisher exact test used to calculate p value due to expected cell frequencies <5. Includes one subject living with the father and his girlfriend.

In 28 cases, evidence of prior abuse and/or neglect was noted in the case file. Nearly half of these cases (12) occurred in households where the mother was living with her boyfriend. In comparison, 28% of all live births in Jefferson County in 1990 were to Black unmarried women, while 5% were to White unmarried women. Table 1 presents these characteristics according to the manner of death. Of particular interest were the age distributions of these two groups. Homicide victims were nearly equally divided between infants (12, 43%) and children age 1 to 5 years (16, 57%). However, deaths where intent was undetermined were over two times more likely to occur among infants (18, 72%). A history of abuse was more likely to be noted in homicide cases (17, 61%) than intent undetermined cases (11, 44%). There were no statistically significant differences between the homicide and intent undetermined groups. Thirty-four offenders were identified for 26 cases of homicide. Table 2 presents the characteristics of these individuals including demographic information and their relationship to the victim. Over 90% (31) of offenders knew their victims, the majority of whom were relatives. There were 14 cases of filicide in which one or both parents were involved. Over one-third (5) of these cases involved the mother acting alone, while two cases (14%) involved only the father. Both parents were implicated in four cases (29%), with one additional case (7%)

1068

J.M. Lyman et al. / Child Abuse & Neglect 27 (2003) 1063–1073

Table 2 Characteristics of known offenders in homicide cases Characteristics

N

% (Min–Max)

Total offenders Mean age in years (SD)

34a 23.6 (7.3)

100.0 (8–41)

Sex Male Female

21 12

63.6 36.4

Race White Black

4 11

26.7 73.3

Relationship to the victim Stranger Known to victim Parent Boyfriend/girlfriend of parentb Friend of parent/victim

3 31 19 7 5

8.8 91.2 61.3 22.6 16.1

a b

Represents 26 homicide cases. Included 6 boyfriends and 1 girlfriend.

involving both parents and a friend. In each of the remaining two cases, a parent acted with the aid of a boyfriend (7%) or girlfriend (7%). Of the 12 remaining homicides where an offender was identified, three deaths occurred at the hands of a stranger. In the other nine cases, the mother’s boyfriend (5) or someone known to the family (4) was implicated. In over 60% (17) of homicides, offenders used their hands to commit the homicide. Within this category, the method of assault ranged from beating to smothering to throwing the child into a ravine. In four of these cases, violent shaking was thought to have contributed to death. The next most common weapon was a firearm, which was used in four cases (14%). A teenaged family friend shot one subject, unknown assailants shot two subjects, and one subject died from hypoxia, which resulted from the death of the mother by firearm just prior to the infant’s delivery. Neglect was cited in two cases (7%), while a knife and overmedication were used in one case each. The 25 intent undetermined deaths were categorized according to the circumstances surrounding the death. The most common scenario involved a parent or caretaker finding the child unresponsive or having trouble breathing (10, 40%). The possibility of sudden infant death syndrome was noted in the files of four cases (16%). Three cases (12%) had a history of medical problems that were thought to have contributed to the subject’s death, while possible neglect was cited in three other cases. Two cases involved delivery of the infant into a toilet (8%). The three remaining cases involved abandonment, possible Shaken Baby Syndrome, and a tub drowning. The 28 homicides were categorized according to the classification scheme suggested by Guileyardo et al. (1999) with the results shown in Table 3. The majority of cases (61%) were classified as due to an angry impulse. Among the four cases of unwanted pregnancy or neonaticide, three cases involved the mother, while the fourth case involved both parents.

J.M. Lyman et al. / Child Abuse & Neglect 27 (2003) 1063–1073

1069

Table 3 Classification of homicide cases Classification

No. of cases

%

Number of homicide cases classified Angry impulse Innocent bystander Negligence and neglect Unwanted pregnancy/neonaticide

28 17 5 2 4

100.0 60.7 17.9 7.1 14.3

The results of toxicological analyses were noted for each subject. Five subjects (9%) tested positive for various drugs. Acetaminophen was found in two subjects, with phenobarbital also present in one of these subjects. Diazepam, diphenhydramine, phenobarbital, and alcohol were discovered in one subject. In this case, the mother overmedicated the subject in an attempt to stop the child’s crying. The remaining two subjects had cocaine in their systems, which was received in utero. One of these subjects died after being delivered into a toilet; the other subject died of hypoxia as a result of the mother being beaten by her boyfriend.

Discussion In this study of 53 homicides and deaths where intent was undetermined, the majority of deaths occurred among children younger than 2 years of age. Approximately two-thirds of the children were Black. The offender most often knew the victim, with half of all homicides and two-thirds of all infant homicides involving a parent. An angry impulse by the offender was the most common homicide scenario. A history of abuse or neglect was noted for over half of all subjects. A larger proportion of deaths occurred among Black children, a finding similar to some studies (Collins & Nichols, 1999; Muscat, 1988) but dissimilar to others (Emerick et al., 1986; Jason et al., 1983). The victims in this study were also slightly more likely to be female, which is contrary to prior studies (Collins & Nichols, 1999; Jason et al., 1983). Noteworthy was the high proportion of fatalities among Black children of unmarried mothers, which has not been reported previously. Strong evidence supports the concept that both homicides and fatal cases of child abuse are undercounted (Christoffel, 1990; Christoffel et al., 1985; McClain, Sacks, Froehlke, & Ewigman, 1993; Sorenson et al., 1997). For this reason, intent undetermined cases were included in this study and should provide a better representation of the true number of homicides that occurred in children in this age group during the study period in Jefferson County. As in other studies, most offenders knew their victim (Collins & Nichols, 1999; Department of Health and Human Services, National Center on Child Abuse and Neglect, 1997; Jason, 1983a; Jason et al., 1983; Kunz & Bahr, 1996; Overpeck et al., 1998; Vanamo et al., 2001). Offenders in the current study were more likely to be male, which has been demonstrated in some studies (Department of Health and Human Services, National Center on Child Abuse and Neglect, 1997; Jason, 1983a; Jason et al., 1983; Kunz & Bahr, 1996), but not others (Collins

1070

J.M. Lyman et al. / Child Abuse & Neglect 27 (2003) 1063–1073

& Nichols, 1999; Vanamo et al., 2001). Studies have shown that mothers and fathers are predominantly the offenders in cases of infanticide, with mothers predominantly responsible for neonaticides (Creighton, 1995; Guileyardo et al., 1999; Jason et al., 1983; Kotch, Chalmers, Fanslow, Marshall, & Langley, 1993). Parents remain the most likely perpetrators of homicides among children age 1 to 5 years (Browne & Lynch, 1995; Crittendon & Crain, 1990; Kotch et al., 1993). Our findings are consistent with these studies in that half of the homicides involved a parent. Among infants, two-thirds of homicides involved at least one parent. Previous literature describes several broad categories of child homicide. Christoffel characterized two types of homicide for children in this age group (Christoffel, 1984). The first type is infanticide, which represents impatience with the child that turns into violence. The second type is fatal child abuse and neglect after infancy, which describes toddlers and preschool children who do not live up to their parents’ developmental goals and are punished or neglected. Fatal child abuse is intrafamilial and occurs most often in children 3 years of age or younger. Studies have shown that many homicides occur during quarrels (Jason, 1983b; Paulson & Rushforth, 1986); this scenario epitomizes the majority of homicides in this study, which occurred as a result of an “angry impulse.” Sorenson and Peterson investigated the social and child protective service history of child homicide victims in Los Angeles, finding that only one in six children was investigated prior to death (Sorenson & Peterson, 1994). Another study of 14 cases of fatal child abuse or neglect found that less than half of the identified families had prior protective service involvement (Hicks & Gaughan, 1995). In the current study, history of abuse was based on JCCMEO investigation files, which included police and autopsy reports. Fifty-three percent of subjects had a history of abuse or neglect. The current study was limited by the relatively small sample. Although Jefferson County represents the largest metropolitan area in Alabama, it is only one of 67 counties in the State. Because of the urban and suburban nature of Jefferson County, the results of this investigation may not be applicable to other parts of Alabama. The main strength of this study was its focus on obtaining details on homicide cases occurring within a fixed population. Findings from this analysis will allow Jefferson County and other geographic areas with similar cultural and demographic profiles to direct intervention resources toward those most likely to benefit from them. The current study utilized coroner records, which are a valuable data source for the study of fatal injuries (Berkelman et al., 1985; Conroy & Russell, 1990; Haddix, Harruff, Reay, & Haglund, 1996), including homicide (Blaser et al., 1984; Goodman et al., 1986; Herman-Giddens et al., 1999). Coroner records usually contain detailed information describing the victim’s death and surrounding circumstances. While a death certificate is limited to short narrative phrases, the coroner’s report will usually explain, if known, how the person was injured and the nature of any injuries (Conroy & Russell, 1990). Whereas previous studies have concentrated on infants, future research should be expanded to include children ages 1 to 5 years. Future studies should focus on the identification of risk factors for homicide in children younger than age 6 and attempt to link multiple sources of data (e.g., vital statistics, crime reports, and coroner reports) in order to obtain a complete picture. The impact of community-level factors and social stressors has not been addressed previously. The inclusion of these ecological variables in an appropriate analysis will increase

J.M. Lyman et al. / Child Abuse & Neglect 27 (2003) 1063–1073

1071

the understanding of how contextual factors translate into differences in individual-level risk. Concurrent evaluation of such data sources should provide an enhanced profile of children at risk for homicide. In summary, among children younger than 6 years of age in Jefferson County, Alabama, deaths due to homicide and deaths where intent was undetermined were similar in nature with respect to demographic and environmental characteristics. Overall, victims were more likely to be younger than 2 years of age and to have a history of abuse. The most common scenario among intent undetermined deaths was the parent or caretaker finding the child unresponsive or having trouble breathing. Among homicide cases, over 90% of the offenders knew their victim, with the majority of offenders being parents. Over 60% of homicides occurred as the result of an angry impulse. Crisis childcare is an intervention method that could be used to focus on this specific type of situation. Home visitation by trained nurses is another strategy that has been shown to be effective in reducing child abuse and neglect among very young children.

Implications Use of a system of child homicide classification subtypes can assist in the interpretation of difficult cases. In order to prevent future child homicides, it is necessary to know how and why children are killed. Sixty-one percent of the homicides in this study were categorized as resulting from an angry impulse. Crisis childcare is a possible intervention method that could address this type of at-risk family. These childcare programs are designed to provide parents in crisis with a break from the stresses of child rearing and provide at-risk children with a safe environment. This type of program has been shown to be an effective preventive intervention (Cowen, 1998). Home visitation by trained nurses during pregnancy and the first 2 years of life has been shown to reduce the frequency of child abuse and neglect among firstborn children of unmarried adolescents of low socioeconomic status (Olds, Henderson, Kitzman, & Cole, 1995; Olds et al., 1997). This is important considering 85% of the cases in this study involved children younger than 2 years of age. A prior study found that 60% of parents who killed their children had been in contact with a physician, nurse, or social worker prior to the act (Scott, 1973). For this reason, pediatricians, family practitioners, nurses, and social workers need to be aware of potential warning signs and risk factors for homicide. Therefore, other prevention strategies may include community access to psychiatric care, prompt intervention at the first report of child abuse, and social agencies receptive to assuming responsibility for unwanted children (Guileyardo et al., 1999). References Berkelman, R. L., Herndon, J. L., Callaway, J. L., Stivers, R., Howard, L. B., & Bezjak, A. (1985). Fatal injuries and alcohol. American Journal of Preventive Medicine, 1, 21–28. Blaser, M. J., Jason, J. M., Weinger, B. G., Elsea, W. R., Finton, R. J., Hanson, R. A., & Feldman, R. A. (1984). Epidemiologic analysis of a cluster of homicides in children in Atlanta. Journal of the American Medical Association, 251, 3255–3258.

1072

J.M. Lyman et al. / Child Abuse & Neglect 27 (2003) 1063–1073

Browne, K., & Lynch, M. (1995). The nature and extent of child homicide and fatal child abuse. Child Abuse Review, 4, 309–316. Centers for Disease Control and Prevention. (1997). Rates of homicide, suicide, and firearm-related death among children—26 industrialized countries. MMWR—Morbidity & Mortality Weekly Report, 46, 101–105. Centers for Disease Control and Prevention. (2001). National Center for Injury Prevention and Control. Leading causes of Death Reports. Retrieved October 2001, from: http://webapp.cdc.gov/sasweb/ncipc/leadcaus.html. Christoffel, K. K. (1984). Homicide in childhood: A public health problem in need of attention. American Journal of Public Health, 74, 68–70. Christoffel, K. K. (1990). Violent death and injury in US children and adolescents. American Journal of Diseases of Childhood, 144, 697–706. Christoffel, K. K., Zieserl, E., & Chiaramonte, J. (1985). Should child abuse and neglect be considered when a child dies unexpectedly? American Journal of Diseases of Childhood, 139, 876–880. Collins, K. A., & Nichols, C. A. (1999). A decade of pediatric homicide: A retrospective study at the Medical University of South Carolina. American Journal of Forensic Medicine & Pathology, 20, 169–172. Conroy, C., & Russell, J. C. (1990). Medical examiner/coroner records: Uses and limitations in occupational injury epidemiologic research. American Journal of Forensic Science, 35, 932–937. Cowen, P. S. (1998). Crisis child care: An intervention for at-risk families. Issues in Comprehensive Pediatric Nursing, 21, 147–158. Creighton, S. J. (1995). Fatal child abuse—how preventable is it? Child Abuse Review, 4, 318–328. Crittendon, P. M., & Crain, S. E. (1990). Developmental trends in the nature of child homicide. Journal of Interpersonal Violence, 5, 202–216. Cummings, P., Theis, M. K., Mueller, B. A., & Rivara, F. P. (1994). Infant injury death in Washington State, 1981 through 1990. Archives of Pediatrics & Adolescent Medicine, 148, 1021–1026. Department of Health and Human Services, National Center on Child Abuse and Neglect. (1997). Child maltreatment 1995: Reports from the states to the National Child Abuse and Neglect Data System. Washington, DC: US Government Printing Office, 2–9. Emerick, S. J., Foster, L. R., & Campbell, D. T. (1986). Risk factors for traumatic infant death in Oregon, 1973 to 1982. Pediatrics, 77, 518–522. Goodman, R. A., Mercy, J. A., Loya, F., Rosenberg, M. L., Smith, J. C., & Allen, N. H. (1986). Alcohol use and interpersonal violence: Alcohol detected in homicide victims. American Journal of Public Health, 76, 144–149. Guileyardo, J. M., Prahlow, J. A., & Barnard, J. J. (1999). Familial filicide and filicide classification. American Journal of Forensic Medicine & Pathology, 20, 286–292. Haddix, T. L., Harruff, R. C., Reay, D. T., & Haglund, W. D. (1996). Asphyxial suicides using plastic bags. American Journal of Forensic Medicine & Pathology, 17, 308–311. Hall, J. R., Reyes, H. M., Meller, J. L., & Stein, R. J. (1993). Traumatic death in urban children, revisited. American Journal of Diseases of Childhood, 147, 102–107. Herman-Giddens, M. E., Brown, G., Verbiest, S., Carlson, P. J., Hooten, E. G., Howell, E., & Butts, J. D. (1999). Underascertainment of child abuse mortality in the United States. Journal of the American Medical Association, 282, 463–467. Hicks, R. A., & Gaughan, D. C. (1995). Understanding fatal child abuse. Child Abuse & Neglect, 19, 855–863. Jason, J. (1983a). Fatal child abuse in Georgia: The epidemiology of severe physical child abuse. Child Abuse & Neglect, 7, 1–9. Jason, J. (1983b). Child homicide spectrum. American Journal of Diseases of Childhood, 137, 578–581. Jason, J., Gilliland, J. C., & Tyler, C. W. (1983). Homicide as a cause of pediatric mortality in the United States. Pediatrics, 72, 191–197. Kotch, J. B., Chalmers, D. J., Fanslow, J. L., Marshall, S., & Langley, J. D. (1993). Morbidity and death due to child abuse in New Zealand. Child Abuse & Neglect, 17, 233–247. Kunz, J., & Bahr, S. (1996). A profile of parental homicide against children. Journal of Family Violence, 11, 347–362. McClain, P. W., Sacks, J. J., Ewigman, B. G., Smith, S. M., Mercy, J. A., & Sniezek, J. E. (1994). Geographic patterns of fatal abuse or neglect in children younger than 5 years old, United States, 1979 to 1988. Archives of Pediatrics & Adolescent Medicine, 148, 82–86.

J.M. Lyman et al. / Child Abuse & Neglect 27 (2003) 1063–1073

1073

McClain, P. W., Sacks, J. J., Froehlke, R. G., & Ewigman, B. G. (1993). Estimates of fatal child abuse and neglect, United States, 1979 through 1988. Pediatrics, 91, 338–343. Muscat, J. E. (1988). Characteristics of child homicide in Ohio, 1974–84. American Journal of Public Health, 78, 822–824. National Center for Health Statistics. (2000). Deaths and death rates for the 10 leading causes of death in specified age groups, by race and sex: United States, 1998. National Vital Statistics Reports, 48, 26–36. National Clearinghouse on Child Abuse and Neglect Information. (2002). National Child Abuse and Neglect Data System (NCANDS) summary of key findings from calendar year 2000. Washington, DC: US Department of Health and Human Services. Olds, D. L., Eckenrode, J., Henderson, C. R., Jr., Kitzman, H., Powers, J., Cole, R., Sidora, K., Morris, P., Pettitt, L. M., & Luckey, D. (1997). Long-term effects of home visitation on maternal life course and child abuse and neglect: Fifteen-year follow-up of a randomized trial. Journal of the American Medical Association, 278, 637–643. Olds, D., Henderson, C. R., Jr., Kitzman, H., & Cole, R. (1995). Effects of prenatal and infancy nurse home visitation on surveillance of child maltreatment. Pediatrics, 95, 365–372. Overpeck, M. D., Brenner, R. A., Trumble, A. C., Smith, G. S., MacDorman, M. F., & Berendes, H. W. (1999). Infant injury deaths with unknown intent: What else do we know? Injury Prevention, 5, 272–275. Overpeck, M. D., Brenner, R. A., Trumble, A. C., Trifiletti, L. B., & Berendes, H. W. (1998). Risk factors for infant homicide in the United States. New England Journal of Medicine, 339, 1211–1216. Paulson, J. A., & Rushforth, N. B. (1986). Violent death in children in a metropolitan county: Changing patterns of homicide, 1958 to 1982. Pediatrics, 78, 1013–1020. Resnick, P. J. (1970). Murder of the newborn: A psychiatric review of neonaticide. American Journal of Psychiatry, 126, 1414–1420. Scott, P. (1973). Parents who kill their children. Medicine, Science and the Law, 13, 120–126. Siegel, C. D., Graves, P., Maloney, K., Norris, J. M., Calonge, B. N., & Lezotte, D. (1996). Mortality from intentional and unintentional injury among infants of young mothers in Colorado, 1986 to 1992. Archives of Pediatrics & Adolescent Medicine, 150, 1077–1083. Sorenson, S. B., & Peterson, J. G. (1994). Traumatic child death and documented maltreatment history, Los Angeles. American Journal of Public Health, 84, 623–627. Sorenson, S. B., Shen, H., & Kraus, J. F. (1997). Coroner-reviewed infant and toddler deaths. Many “undetermineds” resemble homicides. Evaluation Review, 21, 58–76. Vanamo, T., Kauppi, A., Karkola, K., Merikanto, J., & Räsänen, E. (2001). Intra-familial child homicide in Finland 1970–1994: Incidence, causes of death, and demographic characteristics. Forensic Science International, 117, 199–204. Ventura, S. J., Peters, K. D., Martin, J. A., & Maurer, J. D. (1997). Births and deaths: United States, 1996. Monthly Vital Statistics Report, 46(Suppl. 2), 33–34.

Résumé/Resumen French- and Spanish-language abstracts not available at time of publication.