Abusive Head Trauma Outcomes
Abusive head injury (shaken baby syndrome) is the most frequent cause of traumatic mortality and morbidity in infants. Although advancements have been made in the recognition of the clinical, radiographical and pathological findings in abusive head injury, little is known about the long-term outcomes in these patients.
Specifically, the prognosis of infants and children with less severe acute injuries is not known. This review highlights case reports and prospective or retrospective studies that discuss the clinical outcomes of infants and children following abusive head injury.
1. Oliver JE. Microcephaly following baby battering and shaking. British Medical Journal. 1975;2:262-264.
This author describes three infants who had microcephaly following shaking injuries. The author does not qualify what is meant by "shaking injuries".
The first infant suffered inflicted injuries when she was 3 months old. At the age of 7 years, she had profound mental retardation, bilateral spasticity and limited vision and hearing. The second infant was injured when he was 3 ½ months old. At 5 years of age, he had severe behavior problems, a right hemiparesis, right homononymous hemianopia and partial deafness. The third infant was injured when he was 7 months old. At 4 years of age he had bilateral spasticity and profound mental retardation.
2. Ludwig S, Warman M. Shaken baby syndrome: a review of 20 cases. Annals of Emergency Medicine. 1984;13:105-107.
Ludwig and Warman performed a chart review of 20 cases of shaken baby syndrome. Only cases in which there was a perpetrator confession or the medical staff felt that there was no other diagnosis to account for the patient's injuries were included. Children with evidence of other signs of abuse were excluded. The mean age at the time of injury was 5.8 months.
Three patients in this series died. Of the 17 remaining patients, ten had what the authors call significant morbidity including blindness, motor deficits, seizures or developmental delay. Seven patients had no apparent deficit. It is not mentioned how far out from injury that these outcomes were recorded.
3. Frank Y, Zimmerman R, Leeds NMD. Neurological manifestations in abused children who have been shaken. Developmental Medicine Child Neurology. 1985;27:312-316.
These authors describe four cases of infants who were diagnosed with shaking injuries because they presented with neurological and ophthalmologic abnormalities. The ages of the infants at the time of injury were 3 months, 5 weeks, 6 weeks and 4 months.
The first patient was found to have mental retardation, quadriplegia and was blind on follow-up. The second patient was blind and quadriplegic on follow-up examination. For these two patients it is not mentioned when the follow-up occurred. The third patient was blind, had poor motor control and a right gaze preference; she made no improvement up to five months following the injury. The fourth patient was found to be partially blind and spastic six months following the shaking injury.
4. Sinal SH, Ball MR. Head trauma due to child abuse: serial computerized tomagraphy in diagnosis and management. South Medical Journal. 1987;80:1505-1512.
These authors reviewed the charts of 17 children less than 5 years of age with "whiplash shaken infant syndrome". Children were diagnosed with whiplash shaken infant syndrome if there was a history of shaking, or if there were intracranial and intraocular hemorrhages in the absence of signs of external trauma to the head. The mean age at the time of injury was 4.6 months.
Developmental evaluations were performed by a developmental specialist or a pediatric neurologist. Psychological testing and assessment by a physical therapist and a language therapist were also performed. Two children died and one child was found to be normal. Five children were found to have profound mental retardation and four children had mild to moderate mental retardation. Five children had a spastic quadriplegia or diplegia. Six children had epilepsy, one child had behavior problems. Two children had blindness and another had visual problems. These numbers add up to more than seventeen because some children had more than one finding.
5. Wilkinson WS, Han DP, Rappley MD, Owings CL. Retinal hemorrhage predicts neurologic injury in the shaken baby syndrome. Archives of Ophthalmology. 1989;107:1472-14 74.
Wilkinson and colleagues prospectively examined 14 cases of shaken baby syndrome, which they defined as intraocular hemorrhage and intracranial injuries in the absence of external signs of head trauma. The mean age at the time of injury was 10.5 months.
All patients underwent ophthalmic examination, clinical examination and computed tomagraphy imaging. After reviewing the clinical examination and computed tomagraphy findings, two of the authors assigned each patient an initial and late neurological score. The late neurological score was assigned based on the presence of motor or cognitive delay, seizures or residual neurological deficit; one point was assigned for each finding. Additionally, points were assigned for the presence of hydrocephalus or substance loss as demonstrated on CT. A late neurological score was assigned to 8 of the 12 patients who survived; patients were followed for 12 to 56 months following injury. Five patients had a mean late neurological score of 3 and two patients had a mean score of 1 and one child had a late neurological score of 0.
6. Aspan T, Narborough G, Punt JAG, Lowe J. Cerebral contusional tears as a marker of child abuse-detection by cranial sonography. Pediatric Radiology. 1992;22:237-245.
This paper includes six infants younger than six months of age who suffered non-accidental head injury. The infants all had cerebral contusional tears that were detected by intracranial sonography. The mean age of the infants at the time of injury was 10.6 weeks,
The clinical outcomes of these infants were recorded. Outcomes were assessed from 4 months-1 year (mean 6.4 months) following injury. The first infant had global developmental retardation, spastic quadraplegia and severe visual impairment. The second infant had a seizure disorder, developmental retardation and hypertonic lower extremities. The third infant had motor developmental retardation with spastic quadriplegia. The fourth infant was showing evidence of spastic diplegia. The fifth infant died and the sixth infant had visual impairment, hypertonicity and psychomotor developmental retardation.
7. Fischer H, Allasio D. Permanently damaged: long-term follow-up of shaken babies. Pediatrics. 1994;33:696-698.
Fischer and Allasio reviewed the charts of 25 infants admitted to their institution with the diagnosis of shaken baby syndrome. The diagnosis was made if infants had intracranial and retinal hemorrhages in the absence of signs of head trauma or a skull fracture. The mean age of the infants at the time of injury was 4.1 months.
The authors were able to trace ten patients to collect current information through a phone interview of the patient's caregivers and by reviewing their medical records. The authors recorded the patients' neurological, cognitive, sensory and behavior status. Follow-up was obtained when the children were between 8 to 15 years (mean 10.1 years) of age. Three of the ten patients had no problems identified by their caretaker or identified in the medical chart. Seven patients had some impairment including paralysis, seizures, mental retardation, blindness and behavior problems.
8. Bonnier C, Nassogne MC, Evrard P. Outcome and prognosis of whiplash shaken infant syndrome: late consequences after a symptom-free interval. Developmental Medicine Child Neurology. 1995;37:943-956.
Bonnier and colleagues followed 12 children diagnosed with "whiplash shaken infant syndrome" (WSIS). The children ranged in age from 3 weeks to 21 months at the time of diagnosis. The diagnosis of WSIS was made if the children had intraocular and intracranial lesions in the absence of any other traumatic or non-traumatic mechanism.
The children were evaluated on average once a year; evaluations consisted of general, neurological, psychological and social assessments. Developmental tests used include the Gesell, WISC and the Terman. The children were evaluated for an average of four to fourteen years (mean 7.2 years) following injury.
All but one of the twelve children were left with some disability. Disabilities included mental retardation, epilepsy, blindness, quadroplegia, developmental delay and behavioral problems.
9. Johnson DL, Boal D, Baule R. Role of apnea in non-accidental head injury. Pediatric Neurosurgery. 1995;23:305-310.
Johnson and colleagues performed a retrospective review of the medical records of 28 children with a confirmed diagnosis of non-accidental head injury. There were two groups of patients that were studied together: a shaken-impact group with physical signs of impact injury and a shaken group who did not have evidence of direct trauma to the head and neck.
Patients were categorized into good, moderate or severe disability outcome groups, but how these categories were defined is not discussed. Four children died. One child was left in a persistent vegetative state. Seven children had a severe disability, seven were left with a moderate disability and nine had a good recovery.
10. Duhaime AC, Christian C, Moss E, Seidl T. Long-term outcome in infants with the shaking-impact syndrome. Pediatric Neurosurgery. 1996;24:292-298.
Duhaime and colleagues identified children at their institution younger than two years of age who were given the diagnosis of shaken-impact syndrome. The diagnosis was made if the child had clinical and/or radiological confirmation of a subdural and/or a subarachnoid bleed, lack of evidence for other medical diagnoses to account for the injuries and additional findings such as soft tissue injury, fractures or retinal hemorrhages.
The authors were able to contact the current caregivers of 14 children. A telephone questionnaire was developed and administered to the caregivers to assess current medical, functional, educational and social information about the children. The mean age of the children at follow-up was 9 years, with a range of 5.5-15.5 years.
Outcome at the time of follow-up was categorized in grades similar to those used in the Glasgow Outcome Score. One child died. Of the remaining 13 children, six were severely disabled; severe disability included blindness, seizures, hemiparesis, mental retardation or severe behavioral problems. Two children were moderately disabled; both required special education classes and one had a mild hemiparesis. Five children had a "good" outcome according to the authors, but three of these five children have repeated a grade in school or required a tutor, and two of the children had behavioral problems.
11. Matthews GP, Das A. Dense vitreous hemorrhages predict poor visual and neurological prognosis in infants with shaken baby syndrome. Journal of Pediatric Ophthalmology Strabismus. 1996;33:260-265.
Matthews and Das retrospectively examined the cases of five infants with confirmed shaken baby syndrome who also underwent pars plana vitrectomy to remove vitreous hemorrhages. All of the infants met the authors' criteria for shaken baby syndrome, which included evidence of intraocular hemorrhages and subdural or subarachnoid hemorrhages with interhemispheric blood and a "history of suspected child abuse". The infants ranged in age from 3 months to 8 months (mean 5.5 months) at the time of injury.
All of the infants had neurological examinations, but the interval to follow-up is not recorded. Two of the infants were reported to be "neurologically improved" at follow-up. The other three infants did not show neurological improvement. Of these three infants, one was left in a chronic vegetative state and died 5 months following injury. The other two infants had seizures, spasticity and hemiparesis.
The post-operative visual acuity was also measured for each patient. Two of the patients had fix and follow vision in both eyes. Of the remaining three patients, two patients had light perception in both eyes and one patient had light perception in one eye only.
12. Haviland J, Russel RI. Outcome after severe non-accidental head injury. Archives of Diseases in Children. 1997;77,:504-507.
The authors identified 15 children at their institution who were diagnosed with non-accidental head injury (NAI). The diagnosis was made if children had clinical, radiological and/or ophthalmologic evidence that suggested NAI, if the injuries were inconsistent with the history, and if police and social services were involved. The mean age of the children at the time of injury was 5.6 months.
The authors performed a review of the medical records and also contacted the children's caretaker to obtain information about the child's present neurological state and level of functioning. The authors categorized these findings into normal, mildly, moderately or severely handicapped. The follow-up interval ranged from 3 months to 3 years (mean 14.8 months) following injury.
At follow-up, one child was described as normal. Seven children were categorized as severely handicapped and were totally dependent for care. Four had a moderate handicap including severe hemiparesis, blindness and developmental delay. One patient was mildly handicapped with hemiparesis and seizures and two patients died.
They compared these 15 patients to 10 aged matched controls who were given the diagnosis of accidental head trauma. One child was categorized as severely handicapped; one child had a mild handicap. Seven patients were normal at outcome and one patient died. The outcomes of this control group were determined at hospital discharge.
13. Swenson J, Levitt C. Shaken baby syndrome diagnosis and prevention. Minn Med 1997;80:41-44.
Swenson and Levitt evaluated 38 children with the diagnosis of shaken baby syndrome. All of the children presented with neurological or cardiorespiratory symptoms. At the time of injury the children ranged in age from 15 days to 34 months, with a mean age at injury of 6.4 months.
The authors evaluated the children's outcomes; the criteria included neurological status, vision, physical disabilities and survival. Fourteen patients (37%) had no deficits, four patients (10%) had mild-moderate deficits (two had seizures and two had visual impairment), 6 patients (16%) had severe deficits (neurological impairments, visual abnormalities and cervical cord injury), and 12 patients (32%) died. For 2 patients, there was no information except that they had "good" outcomes. The authors do not state the time interval to outcome measurements.
14. Gilles EE, Nelson MD. Cerebral complications of nonaccidental head injury in childhood. Pediatric Neurology. 1998 ;19:119-128.
This retrospective study looked at 14 children with confirmed non-accidental head injury (NAHI). Confirmation of non-accidental head injury was made if there was a witnessed event, perpetrator confession, felony conviction, or in the absence of trauma history with physical evidence of head trauma. The children ranged in age at the time of injury from 53 days to 4.5 years with a mean of 12.2 months.
Outcome was determined from the Denver II Developmental Assessment. Neurological examination and visual examination were also recorded. Clinical records were collected retrospectively. Three patients died. Of the remaining patients, two were in a vegetative state, four had a severe developmental delay, three had a moderate developmental delay and two had a mild developmental delay. Four patients were left with a hemiparesis and five had a spastic quadriparesis. Ten of eleven patients who survived had some visual impairment including homonymous hemianopia and cortical blindness. Three children had seizures. The mean time to post-injury follow-up was 17 months following injury.
15. Mills M. Fundiscopic lesions associated with mortality in shaken baby syndrome. J AAPOS. 1998;2:67-71.
Mills examined 10 consecutive patients with the diagnosis of nonaccidental trauma and traumatic retinal hemorrhage. All patients had intracranial and intraoccular hemorrhage, and medical conditions causing bleeding were excluded, when medically indicated. In addition, the social and police criminal investigation concluded in each case that the history was not consistent with an accidental mechanism. The children ranged in age at the time of injury from 2-10 months with a mean of 5.2 months.
Of the 10 patients examined, 3 died during the initial hospitalization and 7 survived to discharge.
16. Ewing-Cobbs L, Kramer L, Prasad M, Canales DN, Louis PT, Fletcher JM, Vollero H, Landry SH, & Cheung K. Neuroimaging, physical, and developmental findings after inflicted and noninflicted traumatic brain injury in young children. Pediatrics. 1998;102:300-307.
Ewing-Cobbs et al. performed a prospective longitudinal study to analyze the neurobehavioral and developmental outcome findings in children with inflicted and noninflicted traumatic brain injury (TBI). They examined forty children 0-6 years of age hospitalized for moderate and severe TBI and who had no documented previous history of brain injury and no neurological or metabolic disorder. There were twenty children in the inflicted TBI group and twenty children in the noninflicted group. The two groups were comparable in terms of ethnicity, socioeconomic background, gestational ages, neonatal complications and severity of injury. The two groups were not comparable in terms of age. The inflicted TBI group was younger at the time of injury (mean=10.6 months) compared to the noninflicted TBI group (mean=35.6 months).
Determination whether an injury was inflicted or not was based on the assessment of the hospital and county child protection services. An algorithm similar to the one presented by Duhaime et al (1992) was also used to make this determination. Skeletal surveys and fundiscopic exams were performed on all children with suspected abusive injuries.
The outcomes of children suffering TBI was measured with multiple assessments including physical examination findings, the Glasgow Outcome Scale (adjusted for infants/children), and developmental measures. Developmental outcomes were measured with the Bayley Scales of Infant Development Mental and Motor Scales-Second Edition (for children 0-42 months at the time of assessment), the Stanford-Binet Intelligence Scale, Fourth Edition (for children aged 43-71 months at the time of assessment), and the McCarthy Scales of Children's Abilities (for children aged 43-71 months).
Developmental measures were completed at an average of 1.3 months after TBI, following resolution of posttraumatic amnesia. Cognitive testing scores were different between the two groups. More children in the inflicted TBI group scored in the mentally deficient range compared to children in the noninflicted TBI group (45% vs. 5%). Motor testing scores were similar for both groups; 25% of the children in the noninflicted TBI group and 25% of the children in the inflicted TBI group scored in the mentally deficient range. In the inflicted TBI group, 4 patients had a good recovery, 13 patients had a moderate disability and 3 patients had a severe disability. Moderate disability is defined as hemiparesis, borderline cognitive scores, requiring more than one rehabilitation service or requiring early childhood intervention. Severe disability is defined as a total dependence for daily care, severe motor or cognitive deficiency.
17. Ewing-Cobbs L, Prasad M, Kramer L, Landry S. Inflicted traumatic brain injury: relationship of developmental outcome to severity of injury. Pediatric Neurosurgery. 1999;31:251-258.
These authors performed another longitudinal prospective study to examine the neurobehavioral outcomes of children with inflicted traumatic brain injury compared to a comparison group of uninjured children. There were 28 children in each group all younger than 42 months and who had no documented previous history of brain injury and no neurological or metabolic disorder. The two groups were similar in age (9.28 months vs. 9.43 months). The two groups were also matched for ethnicity, socioeconomical status and neonatal course.
Determination whether an injury was inflicted or not was based on the assessment of the hospital and county child protection services. An algorithm similar to the one presented by Duhaime et al (1992) was also used to make this determination. Skeletal surveys and fundiscopic exams were performed on all children with suspected abusive injuries.
The initial evaluation was performed an average of 1.6 months after the injury (baseline), and follow-up was 3 months after the initial assessment. Outcome measures included the Bayley Scales of Infant Development-II, the Bayley Behavior Rating Scale and the Glasgow Outcome Scale.
At baseline, the inflicted TBI group scored significantly lower than the uninjured comparison group on both the mental and motor developmental indicies of the Bayley Scales of Infant Development-II.
At the 3 month follow-up, the inflicted TBI group continued to score below the comparison group in both the mental and motor indicies.
When evaluated with the Glasgow Outcomes Scale, most of the children in the inflicted TBI group had a moderate disability (17/28 children). Four children in this group had a severe disability. The scores were stable from baseline to the three month follow-up; one child improved from moderate disability to good recovery.
When compared to children similar in age and in socioeconomical status but without injury, children with inflicted traumatic brain injury scored lower on both mental and motor domains. The deficit in mental and motor outcomes for the children with inflicted TBI did not improve at the 3 month follow-up, giving these children flat recovery curves. The children with inflicted TBI also had limited recovery of the behavioral measures.
18. Barlow KM, Minns RA. The relationship between intracranial pressure and outcome in non-accidental head injury. Developmental Medicine Child Neurology. 1999;4:220-225.
Barlow and Minns evaluated 17 children with non-accidental head injury. The diagnosis of non-accidental head injury was made if the children had evidence of head injury plus at least two of the following: retinal hemorrhages, skeletal fractures, soft tissue injury or a history which was inconsistent with the injuries. The children ranged from 1-20 months of age at the time of injury with a mean of 5.1 months of age.
Outcomes were assessed with a six-point outcome scale (normal, mild deficit, moderate deficit, severe deficit, vegetative, dead). The scale evaluates outcome on the basis of neurological deficits, presence of seizures and functional outcome. The average length of follow-up was 33 months (range 3-122 months). At follow-up, two children had died, three had a severe deficit, three children had a moderate deficit and two children had a mild deficit. Seven children were classified as normal.
19. Kivlin JD, Simons KB, Lazoritz S, Ruttum MS. Shaken baby syndrome. Ophthalmology. 2000;107:1246-1254.
Kivlin and colleagues performed a retrospective case series of 123 children younger than three years of age who had subdural hematomas and a diagnosis of child abuse. The determination of child abuse (shaken baby syndrome) was made by a child advocacy physician and was based on the presence of subdural hematomas, characteristic bone injuries and the absence of a history of compatible accidental trauma.
Neurologic functioning was determined by review of the medical charts. Twenty-two of 87 survivors (25%) were documented as having a "good" neurologic outcome, defined as normal for age, at the last follow up visit,. Eight patients (9%) were mildly impaired at the last follow-up visit; impairments included attention deficit disorders or mild speech delay. Twenty-six patients (30%) were severely impaired, with deficits including hemiparesis, ataxia or severe developmental delay. Thirty-six patients (29%) died. There was no information available in the chart for 31 children (36%).
The authors also recorded the visual function and eye examinations of patients seen in follow-up by ophthalmologists, other specialists or pediatricians; the authors noted visual outcome for patients at their last outpatient visit (only 60% of the final evaluations were by an ophthalmologist). Mean time interval to follow up was 21 months (range 1 month-7 years). Forty-nine patients (72%) were reported to have "good" vision at their last follow-up visit. Seventeen patients (25%) had "poor" vision. Two additional patients reportedly had poor vision in one eye only. There was no follow-up information for 19 patients.
Cerebral visual impairment from cortical brain injury was the most common cause of bilateral visual impairment (15/19 patients). Poor visual outcome and poor neurologic outcome were highly correlated. Eye findings that were associated with death included initial lack of visual response, initial poor pupillary response and the presence of retinal hemorrhages.
20. Fung ELW, Sung RYT, Nelson EAS, Poon WS. Unexplained subdural hematoma in young children: Is it always child abuse? Pediatrics International 2002;44:37-42.
These authors examined the clinical details of four infants with subdural hematomas and who were also determined following multidisciplinary case conference to have suffered non-accidental injury. The mean age of the infants at the time of injury was 6 months.
The clinical records were reviewed, and the neurological status of the patients recorded. The authors do not state how long after injury the outcomes were measured. One patient had a mild deficit, described as a borderline developmental delay. One patient had moderate deficits with right lower limb weakness, developmental delay and a left convergent squint. Two patients suffered severe deficits with spastic quadriplegia (2 patients), epilepsy (1 patient), and retinal detachment (1 patient).
Visual Outcomes
1. Han DP, Wilkinson WS. Late ophthalmic manifestations of the shaken baby syndrome.
Journal of Pediatric Ophthalmology Strabismus. 1990;27:299-303.
Han and Wilkinson examined six children after a shaking episode to determine whether retinal abnormalities persisted. Shaken baby syndrome was defined by the authors as the findings of retinal, vitreous or subhyaloid hemorrhage along with intracranial injury without external signs of head trauma.
Follow-up examination consisted of visual acuity testing, pupillary testing for afferent defects, anterior segment examination and orthoptic evaluation for motility disturbances. Indirect and direct ophthalmoscopy was also performed.
The mean age at time of injury for the six patients was 9.1 months (range 3-20 months) and the mean age at follow up was 29.8 months (range 21-55 months). Three children had profound visual loss due to occipital lobe atrophy. These three children also had severe psychomotor retardation. Two children had moderate visual loss in one eye. These two children had previous retinal hemorrhages, macular folds and loss of retinal transparency. The last child's vision was described as central, steady and maintained; this patient also had a left hemiparesis.
2. Matthews GP, Das A. Dense vitreous hemorrhages predict poor visual and neurological prognosis in infants with shaken baby syndrome. Journal of Pediatric Ophthalmology Strabismus. 1996;33:260-265.
Matthews and Das retrospectively examined the cases of five infants with confirmed shaken baby syndrome who also underwent pars plana vitrectomy to remove vitreous hemorrhages. All of the infants met the authors' criteria for shaken baby syndrome, which included evidence of intraocular hemorrhages and subdural or subarachnoid hemorrhages with interhemispheric blood and a "history of suspected child abuse". The infants ranged in age from 3 months to 8 months (mean 5.5 months) at the time of injury.
All of the infants had neurological examinations, but the interval to follow-up is not recorded. Two of the infants were reported to be "neurologically improved" at follow-up. The other three infants did not show neurological improvement. Of these three infants, one was left in a chronic vegetative state and died 5 months following injury. The other two infants had seizures, spasticity and hemiparesis.
The post-operative visual acuity was also measured for each patient. Two of the patients had fix and follow vision in both eyes. Of the remaining three patients, two patients had light perception in both eyes and one patient had light perception in one eye only.
3. McCabe CF, Donahue SP. Prognostic indicators for vision and mortality in shaken baby syndrome. Arch Ophthalmology. 2000:118;373-377.
McCabe and Donahue reviewed the medical records of 30 children diagnosed with shaken baby syndrome. A diagnosis of shaken baby syndrome was made when bilateral retinal hemorrhages were seen and when the history was not consistent with the injuries, when other obvious signs of abuse were present or when there was a history of a previous suspicious episode of abuse. The mean age of children at the time of injury was 9.3 months.
Twenty-two patients survived. Twenty patients returned to the outpatient clinic for follow-up (range of time interval to follow-up: 1-36 months; mean 6.7 months). At each visit, the presence of retinal hemorrhages was noted. The ophthalmologists also examined patients for ocular motility, fix and follow vision and the presence of amblyopia.
After retinal hemorrhages resolved, 6 patients (27%) had poor vision in at least one eye. The poor vision was due to cortical visual impairment in two patients, optic atrophy in two children, retinal fibrosis and retinal scarring each in one patient. Twelve patients initially presented with fix and follow vision (40%). Of these twelve, ten retained fix and follow vision. Eighteen patients presented without fix and follow vision in at least one eye. Of these 18 patients, eight died. Seven of the remaining ten patients gained vision and three had no improvement.
There were 22 patients who had reactive pupils initially; all survived. All eight patients who initially had non-reactive pupils died.
4. Kivlin JD, Simons KB, Lazoritz S, Ruttum MS. Shaken baby syndrome. Ophthalmology. 2000;107:1246-1254.
Kivlin and colleagues performed a retrospective case series of 123 children younger than three years of age who had subdural hematomas and a diagnosis of child abuse. The determination of child abuse (shaken baby syndrome) was made by a child advocacy physician and was based on the presence of subdural hematomas, characteristic bone injuries and the absence of a history of compatible accidental trauma.
Neurologic functioning was determined by review of the medical charts. Twenty-two of 87 survivors (25%) were documented as having a "good" neurologic outcome, defined as normal for age, at the last follow up visit,. Eight patients (9%) were mildly impaired at the last follow-up visit; impairments included attention deficit disorders or mild speech delay. 26 patients (30%) were severely impaired, with deficits including hemiparesis, ataxia or severe developmental delay. Thirty-six patients (29%) died. There was no information available in the chart for 31 children (36%).
The authors also recorded the visual function and eye examinations of patients seen in follow-up by ophthalmologists, other specialists or pediatricians; the authors noted visual outcome for patients at their last outpatient visit (only 60% of the final evaluations were by an ophthalmologist). Mean time interval to follow up was 21 months (range 1 month-7 years). Forty-nine patients (72%) were reported to have "good" vision at their last follow-up visit. Seventeen patients (25%) had "poor" vision. Two additional patients reportedly had poor vision in one eye only. There was no follow-up information for 19 patients.
Cerebral visual impairment from cortical brain injury was the most common cause of bilateral visual impairment (15/19 patients). Poor visual outcome and poor neurologic outcome were highly correlated. Eye findings that were associated with death included initial lack of visual response, initial poor pupillary response and the presence of retinal hemorrhages.
