Child
Abuse
by Drowning
Karen
J. Griest,
MD, and
Ross
E. Zumwalt,
MD
From the Office of the Medical Investigator, University of New Mexico School of Medicine,
Albuquerque
ABSTRACT. Drowning as a form of subtle fatal child abuse is difficult to distinguish from accidental immer-sion or from sudden unexpected natural death when the circumstances of immersion are concealed. Homicidal
drownings are unwitnessed, usually occurring in the
home, and the victims are young, either infants or tod-dlers. Accidental drownings are more likely to involve toddlers or older children in public areas such as swim-ming pools, drainage ditches, lakes, and rivers. This is especially true in rural areas. In cities, bathtubs remain a major site of accidental childhood drownings. Perpetra-tors of deliberate drownings often fit the sociopathologic profile of a child abuser. Because there is often a survival
interval between immersion and death, pathologic find-ings consistent with postimmersion syndrome suggest the cause of death. Foreign material in the lungs, if immer-sion was other than in clear tap water, and injuries of the face are other positive correlating factors. A thorough investigation of the circumstances and cooperation be-tween the investigating agency and the pathologist are essential to determine the correct manner of death in these cases. Pediatrics 1989;83:41-46; child abuse, drown-ing, immersion, nonaccidental death.
Since the publication of the landmark article concerning nonaccidental trauma in 1946 by Caf-fey,’ numerous forms of child abuse have been described in the medical !iterature.2’1 Victims of fatal child abuse run the gamut from multiple in-juries of varying ages (including battered baby syn-drome) to few or no injuries. Because infants and small children can be fatally injured without leaving physical signs, determination of a homicidal man-ner of death often necessitates circumstantial as we!! as anatomical evidence.’2’3 Willful drowning
Received for publication Jan 18, 1988; accepted March 17, 1988. Reprint requests to (K.J.G.) Office of the Medical Investigator,
University of New Mexico School of Medicine, Albuquerque, NM 87131.
PEDIATRICS (ISSN 0031 4005). Copyright © 1989 by the
American Academy of Pediatrics.
as a form of subtle fatal abuse embodies this
prob-lem.
Deliberate immersion with or without a fatal outcome is a phenomenon that may be encountered but not recognized by many physicians who treat children, by coroners and medical examiners, and by law enforcement and other protective service personnel. It is no doubt underreported and under-diagnosed because of the lack of physical evidence or criteria on which to base a diagnosis of abuse. There are few articles concerning nonaccidental immersion of children in the medical literature,’
16 a reflection of the inherent difficulties of these types of cases.
Nixon and Pearn’4’6 in 1977 described several features that they believed helped to differentiate accidental from nonaccidental immersion. All of the nonaccidental drownings or near-drownings re-ported by them occurred in the bathtub at an un-usual time of day with the drowned child alone in the bath. The parents fit the usual sociopathologic profile of abusing guardians, and a precipitating crisis, often a domestic problem, was present. These abused children were older, usually between 15 and 30 months of age compared with 9 to 15 months of age for accidental immersion. Accidental immer-sion often occurred during the usual bath time when more than one child was present in the tub, the older ones left the youngest, and, significantly, the household routine was upset, resulting in a lapse of adequate supervision. The depth of the bath water ranged from 5.0 to 35.0 cm (2 to 14 in). Children who were immersed four minutes or less survived, whereas those immersed five minutes or more died. Nixon and Pearn concluded that in cases of delib-erate immersion, the child was held under water until unconsciousness ensued.
We describe six childhood homicides by drown-ing, four seen in 1 year in a jurisdiction of 1.5 million people. These cases have a number of fea-tures in common with the previously reported cases
CASE REPORTS
Case I
A 2#{189}-year-old boy was found unresponsive in his bed by his 11-year-old uncle who said that the upper part of
the front of the boy’s body was wet. The boy lived with his unmarried mother, biologic father, and several other
related individuals. The parents were known abusers of alcohol and drugs. Pencyclicine was used by various members ofthe household on the night ofthe boy’s death.
No one in the house claimed knowledge of the events leading to the boy’s death, although the father stated that
the boy’s head may have been submerged in water. When the child was autopsied, there were nonpat-terned contusions of the lips, lacerations and abrasions of the buccal mucosa, and frontal subscalpular hemor-rhages. Internally, the brain and lungs were edematous. The only significant microscopic finding was pulmonary edema with occasional polymorphonuclear leukocytes in alveolar spaces. The bruises of the lips and scalp and the injuries of the mouth indicated force applied to the face. The cerebral and pulmonary findings were consistent with a short survival after immersion in water. This, together with the wet condition of the body and state-ments of the father, pointed to drowning as the cause of
death.
The father was indicted for involuntary manslaughter but acquitted at his trial because of lack of evidence.
Case 2
A 3-year-old girl was found on the couch in the living room of her home by paramedics summoned by a con-cerned neighbor. In spite ofthe fact that the girl appeared to be dead, resuscitation was attempted at the scene and at a local hospital.
The paramedics reported that the girl’s clothing was wet and her hands and feet were “water-logged.” In the home, the police found a group of nude women, including the girl’s mother and grandmother, seated on the floor in another room, conducting a religious ritual with chanting and swaying.
At the autopsy, contusions of the lips and cheeks were present as were several small recent contusions of the extremities. On the conjuctivae, eyelids, and upper face were prominent petechiae (Fig 1). Except for cerebral and pulmonary edema, there were no other injuries or natural diseases found after a complete autopsy. Micro-scopically, there was amorphous pale-staining foreign material in some bronchioles (Fig 2).
The following postmortem chemical analysis of vitre-ous humor was reported: sodium 119 mEqJL and chloride
105 mEqJL.
The petechial hemorrhages of the face suggested an
asphyxial mechanism of death. The cerebral edema in-dicated some survival after the asphyxia! episode. The wet clothing and water-logged hands and feet suggested drowning as the cause of asphyxia. The concentrations of the electrolytes in the vitreous were consistent with dilution by water absorbed through the lung and/or
in-testinal tract.
Fig I. Petechiae on conjunctiva and eyelids.
Fig 2. Amorphous foreign material in a bronchiole (he-matoxylin and eosin, x200).
Based on the autopsy findings, the police requestioned the people present in the house at the time of the girl’s death. The grandmother was a religious extremist who had practiced a religious cleansing of the devil from the child which consisted of holding the child down, forcing her mouth open while holding her nose closed, and pour-ing water from a 2-L soft drink container down her throat. This was repeated until the child became unresponsive.
The asphyxia was caused by water being forced into the airway, as evidenced by the foreign material in the bronchioles, a probable contaminant of the water or soft drink container, and perhaps by blockage of the airway during manipulation of her mouth and head. The grand-mother was convicted of second-degree murder.
Case 3
A 2-year-old male child in a comatose state was brought to an emergency room by his father. He and his 4- and 1-year-old siblings were in the sole custody of their father while their mother was hospitalized. The father told the police that the child had fallen from the
monkey bars at the park earlier in the day, striking the back of his head. The child did not lose consciousness at that time but collapsed suddenly later in the evening. The boy survived in a coma for two days.
con-tusions of the chest, abdomen, and extremities. There were numerous white small scars on the abdomen, back, helix of the left ear, and extremities. The brain was markedly edematous and soft, with microscopic evidence of hypoxia. The lungs were moderately edematous and congested, with moderately abundant numbers of foamy macrophages in the alveoli, extensive intraalveolar poly-morphonuclear cell infiltrates in all lobes, and hyaline membranes (Fig 3). Microscopically, the bruises appeared to be at least several days old.
When the police confronted the father with the au-topsy findings consistent with postimmersion syndrome, he confessed to having held the child’s head underwater in the bathtub until the boy lost consciousness. The father, who had a previous history of abusing his sons, was charged and convicted of child abuse resulting in
death, a second-degree felony in New Mexico.
Case 4
A newborn infant was found in the toilet of a private
home by paramedics who responsed to a summons for
help for the 17-year-old unwed mother who had vaginal bleeding and had fainted.
Autopsy findings indicated that the infant boy was 34
to 35 weeks’ gestation, viable, nonmacerated, and without congenital abnormality or natural disease. A hematoma
occupying approximately one third of the maternal sur-face of the placenta was present. Microscopic studies of the infant and placenta revealed a third-trimester pla-centa without abnormality except for the hematoma. The infant’s organs were consistent with 35 weeks’ gestational
age. The lungs were mature, and present within the
bronchi, bronchioles, and alveoli were eosinophilic amor-phous appearing vegetable cell walls consistent with the microscopic appearance of the toilet paper in the water sampled from the toilet (Fig 4). The lungs were
edema-tous. There was generalized partial inflation of the al-veoli.
The infant was live-born and drowned in the toilet as evidenced by the partial inflation of the lungs and foreign
Fig 3. Intraalveolar polymorphonuclear cells and hya-line membranes (hematoxylin and eosin, x200).
Fig 4. Eosinophilic amorphous appearing vegetable cell wall, consistant with toilet paper, in alveolus (hematox-ylin and eosin, x400).
material from the toilet water found deep within the lungs. He was of sufficient maturity to survive on his own. His premature delivery was probably precipitated by partial detachment of the placenta with formation of a hematoma under the placenta.
The events surrounding the delivery and death of the infant given by the mother and family members present in the household were conflicting. The position of the infant in the toilet was not consistent with delivery into the toilet. The mother was charged with child abuse.
Case 5
The body of a 3-month-old girl was recovered from a shallow grave in a rural area. The mother and her boy-friend had at first reported the baby kidnapped but later admitted that the baby had died and they had buried her. When the baby was autopsied, there was edema of the lungs and brain, a healing fracture of the left eighth rib, and periosteal reaction of the right humerus and left and right femora, as well as soft tissue swelling of the right upper arm. A moderate diaper rash was present. Micro-scopically, the lungs were edematous with large areas of moderately abundant intraalveolar macrophages, intraal-veolar edema, and polymorphonuclear cells in all lobes. There were a few intraalveolar giant cells. A small group of intraalveolar vegetable cells with mild surrounding cellular reaction and several intraalveolar round struc-tures consistent with vegetable spores were present (Fig 5). The lungs were mature and without other lesion. There were no other findings except for stress changes in the thymus and adrenal glands. The rib fracture was healing and greater than ten days old.
Analysis of the water from hot water springs at the last encampment site of the couple showed vegetable matter and spores similar to those in the child’s lungs. The abundant polymorphonuclear leukocyte reaction present in all lung lobes indicated diffuse damage to the
Fig 5. Intraalveolar vegetable cells with mild surround-ing cellular reaction (hematoxylin and eosin, x40).
injury alone could have been accidental or birth trauma, the number and type of injuries strongly suggested the battered baby syndrome.
The boyfriend had a police record in another state for attempted drowning of a child of another girlfriend in the same hot water spring. The mother and boyfriend were charged with child abuse resulting in death.
Case 6
A 9-month-old boy was found unresponsive in the bathtub by his mother. He had been left in the tub with his 2-year-old brother for 15 minutes while his mother went to get some clothes for the children. The rescue squad was not called until 20 minutes later.
There was a history of child abuse, including neglect and sexual abuse, in this family.
When the child was autopsied, the lungs and brain were edematous. Watery fluid was present in the small bowel. On the forehead were a few recent small abrasions associated with some subscalpular hemorrhage (Fig 6). Microscopically, distended alveoli alternated with partial atelectasis. The contusions of the forehead were both acute and healing.
Because of conflicts in the mother’s story and the delay
in rendering aid, the mother was charged with child
neglect.
DISCUSSION
Homicide by drowning is a crime perpetuated on infants and small children in an unwitnessed set-ting. In our group of six drownings, although only three occurred in the bathtub, five occurred in the home. The sixth occurred in a private area, a rural hot spring, used by the transient guardians as a wash area. The children ranged from newborn to 3 years of age, matching the peak age range for death from all forms of abuse (Table 1).
Cooperation between law enforcement personnel
Fig 6. Small recent abrasions on forehead.
TABLE 1. Child Abuse by Drowning
Case
No.
Age Sex Site of
Drowning
Contributing
Factors
1 2#{189}yr M Bathtub Alcohol and drug abuse
2 3 yr F Living room
3 2 yr M Bathtub History of child abuse 4 Newborn M Toilet
5 3 mo F Hot spring History of child abuse 6 9 mo M Bathtub History of child
abuse
and the forensic pathologist was necessary to make the correct determination of cause and manner of death in all cases. There was a history of previous abusive behavior by the parent or guardian in three of the cases and a history of alcohol and drug abuse in another. In two cases, the wet state of the body and clothing provided valuable clues to the final diagnosis, because the scene had been disturbed and could not be evaluated. In four of the cases there was a delay between the event and a sum-moning of help or of finding the expired child.
Three of the cases had postmortem pulmonary features of the postimmersion syndrome. The postimmersion syndrome is characterized by wide-spread intraa!veo!ar polymorphonuclear leukocyte infiltrates and edema.’7’9 Hyaline membranes may be present. In those cases in which the child sur-vives in the hospital for a short period of time, early intubation and oxygen effects must be distin-guished from the postimmersion syndrome. This is not difficult because cellular infiltrates are not a feature of the respirator syndrome or oxygen tox-icity.20’2’ Superimposed bronchopneumonia has a distinctive microscopic pattern that is different from postimmersion syndrome.
Scene investigation Position of body Conditions of environment Presence of blood and location Statements of witnesses Statements of rescue personnel Social history
Sample of water from drowning site Documentation by photographs and diagrams
External examination at autopsy
Condition of clothing, including areas of wet-ness
Presence of injuries, with location, color, size, and shape
Condition of body to include livor mortis, rigor
mortis, and temperature
Presence of edema foam at nose and mouth Placental examination in newborns
Internal examination at autopsy
Degree of pulmonary edema and edema foam in trachea and bronchi
Presence of water in stomach
Presence of underlying congenital and other diseases
Complete microscopic studies, including multi-ple sections of lungs and water in stomach Samples for cultures, toxicologic studies, and
paternity testing
Special studies Cultures Toxicology Dating of injuries
Chemical analysis of water from drowning site, lungs, and stomach
Special incisions to disclose hidden injuries
Full-body roentgenograms for infants Flotation test of newborn lung tissue indicated delay between the immersion and death.
Important in the cases that did not occur in the bathtub was the presence of foreign material in the lungs. This foreign material confirmed both the cause of death as drowning and, in some cases, the location of the drowning. Of course, analyzing the water from the potential immersion site for foreign material and matching it to that in the lung was crucial.
A
review of the series of accidental drownings of children between birth and 12 years of age we saw from 1982 through November 1987 are summarized in Table 2. The ruling of accident was made only after a complete investigation. Questionable or in-determinate cases were excluded.New Mexico is a predominately rural state with a total population of 1.5 million people. Agriculture
is a primary occupation, accounting for the fact
that irrigation ditches are the primary site of child-hood drownings in the state. There were a total of 66 cases of accidental drownings; these were equally divided into two distinct groups: newborns to 3-year-old children and 4 year- to 12-year old chi!-dren. There were more boys than girls in both groups, 20 boys to 13 girls in the newborn to 3-year-old group and 20 boys to three girls in the older group. Irrigation ditch drownings prevailed among the toddler group, whereas lakes, rivers, and swimming pools were the prime location among the older children. In both of the bathtub drownings of older children (4 and 8 years of age), there was a major contributing factor, cerebral palsy and epi-lepsy. In the three bathtub drownings of younger children (10 to 17 months of age), the child was left alone in the tub by the parent for a short period of time. In one case, an older sibling was in the bath with the infant. In the two bucket deaths, the child was a toddler who was tall enough to tip himself into the pail, but not large enough to knock it over.
A
sufficient quantity of water was present in the bucket to weigh it down and cover the child’s nose and mouth.Toilet drownings of newborns are a unique oc-currence necessitating special documentation (Fig 7). The usual history is that the child was delivered into the toilet by a mother who was young, unaware that she was pregnant, and/or in labor. The role of the police and pathologist is to verify the circum-stances surrounding the infant’s death and deter-mine the cause of death. At the scene, the position of the infant and placenta in the toilet must be noted and documented with photographs. The pres-ence and quantity of blood in the toilet and other areas of the bathroom and house should be dia-gramed and photographed. If the child were re-moved from the toilet by the family or the mother,
TABLE 2. Accidental Drowning of Children, 1982-1987
Age/Sex (No.) of Children
Location of Drowning (No. of Children)
Newborn
M (1) Toilet (2) F(1)
10 mo-3 yr
M (20) Irrigation ditch (15) F (13) Lake or pond (2)
River (4); swimming pool, Jacuzzi (3); septic tank (1); bathtub (3); bucket (2)
4-12 yr
M (30) Irrigation ditch (5)
F (3) Lake or pond (10); river or creek (7); swimming pool, hot tub (6); sep-tic tank, etc (3); bathtub (2)
Fig 7.
syn-copa! episodes or bleeding, witnesses and assistance at the delivery, other people in the house, and postnatal aid given to the mother and baby. The placenta is examined grossly and microscopically for pathology and the umbilical cord for tearing or cutting. Things to note on the infant’s body are edema foam at the nostrils and in the trachea, injuries, infections, congenital anomalies, and signs of decomposition or maceration. Complete micro-scopic examination of organs and injuries is neces-sary. Any discoloration of the skin should be incised to disclose hidden injuries. The gestational age is calculated using the baby’s height, weight, foot length, organ maturity, and bone maturity accord-ing to roentgenogram. Water from the toilet should
be
compared with any water in the stomach and deep within the lungs. Chemical additives and for-eign material in the toilet water may be very useful in this respect. Because of the position of the infant in the toilet, water present in the stomach is not due to passive diffusion.The
color and consistency of a newborn’s lungs are not reliable indicators of respiration unless the lungs are pink and fluffy. Microscopic examination of the lungs for aeration is also difficult and shouldbe
approached with caution.’7 The flotation test-dropping a section of lung in water, to see whether it floats or sinks-is useful if there is no decompo-sition or history of resuscitation.Other
special procedures useful in the case of newborn deaths include cultures of lung and blood. Full-body roentgenograms, in addition to showing bone and soft tissue anomalies, will reveal fractures and gas patterns in lung and bowel. Samples should be drawn for paternity and maternity typing and drug analysis (narcotics, alcohol, and CNS depres-sants) when indicated.In summary, the successful investigation of a nonaccidenta! drowning of a child requires cooper-ation between investigative agencies and the pa-thologist, a complete autopsy, a social history, and
special studies such as analysis of the water at the site of the drowning and aging of injuries.
REFERENCES
1. Caffey J: Multiple fractures in the long bones of infants
suffering from chronic subdural hematoma. Am J Roent-genoi 1946;56:163-173
2. Kempe CH, Silverman FN, Steele BF, et al: The
battered-child syndrome. JAMA 1962;181:17-24
3. Ellerstein NS: The cutaneous manifestations of child abuse and neglect. Am J Dis Child 1979;133:906-909
4. Enos WF, Conrath TB, Byer JC: Forensic evaluation of the sexually abused child. Pediatrics 1986;78:385-398
5. Caffey J: On the theory and practice of shaking infants. Am JDiS Child 1972;124:161-169
6. Ayoub C, Pfeifer D: Burns as a manifestation of child abuse and neglect.Am J Di.s Child 1979;133:910-914
7. Adelson L: Homicide by starvation: the nutritional variant
of the “battered child.” JAMA 1963;186:458-460
8. Zumwalt RE, Hirsch CS: Subtle fatal child abuse. Hum Pat/wi 1980;11:167-174
9. Meadow R: Munchausen syndrome by proxy: The hinterland of child abuse. Lancet 1977;3:343-345
10. Adelson L: The battering baby. JAMA 1972;222:159-161
11. Rosenn SW, Loeb LS, Jura MB: Differentiation of organic from non-organic failure to thrive. Pediatrics 1980;66:689-704
12. Norman MG, Newman DE, Smialek JE, et al: The
post-mortem examination on the abused child: pathological, ra-diography and legal aspects. Perspect Pediatr Pat/wi
1984;8:313-343
13. Zumwalt RE, Hirsch CS: Pathology of fatal child abuse and neglect, in Helfer RE, Kempe RS (eds): The Battered Child, ed 4. Chicago, The University of Chicago Press, 1987, p 280 14. Nixon J, Pearn J: Non-accidental immersion in bathwater:
Another aspect of child abuse. Br Med J 1977;1:271-272
15. Pearn J, Nixon J: Bathtub immersion accidents involving
children. Med JAust 1977;1:211-213
16. Pearn J, Nixon J: Prevention of childhood drowning
acci-dents. Med J Aust 1977;1:616-618
17. Adelson L: The Pat/wiogy ofHomicide, ed 1. Springfield, IL, Charles C Thomas Publisher, 1974, pp 573-574, 628-632 18. Fuller RH: Drowning and the post-immersion syndrome.
Miiit Med 1963;128:22-36
19. Modell JH: The Pathophysioiogy and Treatment of Drowning
and Near-Drowning. Springfield, IL, Charles C Thomas Publisher, 1971, pp 79-81
20. Katzenstein AA, Bloor CM, Leibow AA: Diffuse alveolar
damage-The role of oxygen, shock, and related factors. Am
J Pat/wi 1976;85:210-224
21. Nash G, Blennerhassett JB, Pontoppidan H: Pulmonary
lesions associated with oxygen therapy and artificial
