Tuesday, September 24, 2019

Benign external hydrocephalus mistaken for child abuse

Wester K. Two Infant Boys Misdiagnosed as "Shaken Baby" and Their Twin Sisters: A Cautionary Tale. Pediatr Neurol. 2019 Aug;97:3-11.

BEH [benign external hydrocephalus] may, however, be associated with cognitive problems 319 or have considerably more severe consequences, including a predisposition for subdural hematomas (SDHs). This predisposition for SDH does not appear to be well-understood by many of the physicians who deal with child abuse. Moreover, the scientific evidence supporting a causal connection between nonaccidental trauma and SDH in infants is at best scant.

This article describes two pairs of dizygotic twins with nearly identical histories. Both twin pairs were born preterm just three weeks apart; each pair consisted of a boy and a girl, and most importantly, the boys, but not the girls, developed symptoms of increased intracranial pressure (ICP). All the children were diagnosed with wide subdural-blood-containing fluid collections, mainly over the frontal lobes, and medical expert witnesses told the court that all four children had been subjected to violent shaking. As a consequence one parent was sentenced to 1.5 years in jail and all four children were taken from their biological parents and raised in foster homes for more than three years until appeal courts decided in favor of the parents.

These children are by no means unique, but they may serve to illustrate some important points concerning BEH and “shaken baby syndrome" or abusive head trauma, including some striking epidemiologic similarities. In addition to documenting these four children, this contribution reviews the literature on the shaken baby syndrome or abusive head trauma with an emphasis on the possibility that some infants with BEH might be mistaken for shaken baby syndrome....


Computed tomography (CT) the next day showed fluid collections in widely enlarged subdural spaces, especially in the frontal region, with higher density than cerebrospinal fluid (CSF), but only a few, minor acute blood clots. The initial radiological description listed external hydrocephalus as a diagnosis; this was later abandoned, and the CT was finally described as showing a “large subdural hematoma,” SDH.

Child abuse was suspected. A total skeleton x-ray evaluation revealed no fractures; he had no bruises or subcutaneous hematomas. Ophthalmoscopy showed bilateral, extensive retinal hemorrhages (RH)...

Based on the radiological description of “a large subdural hematoma” and bilateral retinal hemorrhages (RH), the case was reported to the child protection team and the police, followed by a charge of child abuse against his father. The medical experts in the lower court (forensic medicine, ophthalmology, and pediatrics) testified in favor of child abuse, disregarding the perioperative finding of only clear fluid without visible blood. The father was sentenced to 1.5 years in jail but appealed the verdict. Both twins were taken from the parents and raised in foster homes for the next three years.

The boy had sustained severe brain injury; the follow-up MRI was described as follows: “ The restricted diffusion as seen here is a finding that in nearly all cases can be taken as a sign of severe lack of oxygen and an early permanent brain damage ” (the author's translation from Norwegian). The boy is permanently vegetative...

The appeal court

In the appeal court, the author was appointed expert witness in addition to those from the primary court, who still favored the abuse explanation. The father was found not guilty, as the court accepted BEH and associated complications as a more likely cause of the condition than physical abuse...

Boy 2

He had no bruises or other external indications of injury. On the third day, MRI and CT scans were performed. The CT scan showed “ a subdural haematoma/fluid collection with a definitely higher density than CSF, but lower density than cerebral cortex (hypodense/isodense) with hyperdense stripes in the frontal subdural space, probably representing fresh blood in the subdural space ” (the author's translation). The MRI scan was interpreted as showing the same, with the age of the hematomas estimated as two to three days to a few weeks. Twelve plain skeleton X-rays revealed no fractures.

Fundoscopy disclosed multiple RH in several layers of retina in one eye but only two minor peripheral bleedings in the other, the conclusion being “ findings compatible with retinal haemorrhages as seen in the shaken-baby syndrome, although it is atypical that the extent of bleedings is so different in the two eyes ” (the author's translation).

Child protection authorities and police were notified, and both infants were routinely placed in a foster home, where they remained for the next three years.

The appeal court

The police found it difficult to prove physical abuse and rested the case. The child protection authorities, however, insisted on keeping the children separated from their parents. A lower court had decided in favor of the parents, but the child protection authorities appealed that decision and delayed the transfer of the children back to their parents for another 16 months. The appeal trial was almost identical to the one described above, with almost the same set of expert witnesses, including the present author. The court decided in favor of the biological parents, and after more than three years, the children were reunited with their parents.

This report describes two pairs of three-month-old premature dizygotic twins who were separated from their biological parents for three years under the assumption that they had been subjected to vigorous shaking. This assumption was based solely on neuroimaging findings of extracerebral or subdural fluid collections described as containing chronic hematomas and small amounts of fresh blood. In addition, the boys, but not the girls, had dramatic symptom debuts and were found to have RH described as compatible with nonaccidental head injury caused by shaking. None of the twins had any sign of impact to the head or any extracranial finding indicating violence. Duhaime et al. stated in 1987 that impact is required to cause the findings of the triad.

In the author's opinion, all four infants exhibited extracerebral fluid collections compatible with external hydrocephalus, as defined by several authors. Most likely, their BEH condition had been complicated by spontaneous bleeding or oozing of blood products into the subdural space; BEH is known to predispose for spontaneous subdural bleedings in infants, and this predisposition can be a pitfall in the diagnosis of abusive head trauma.

Shaken baby syndrome and lack of medical evidence

These infants were diagnosed as having been shaken with the most serious consequences for their families. It therefore seems appropriate to analyze the quality of medical evidence behind the widely accepted notion that a triad consisting of SDH(s), RH, and encephalopathy can be used to prove a criminal act—shaking. As in other countries, in Norway guilt has to be proven beyond reasonable doubt. The solidity of the triad as proof of a criminal act must therefore not be doubted.

It is difficult to find scientific evidence above level 3 for a causal relationship between the triad and violent shaking, as also concluded after an extensive earlier review. 41 A more recent review, based on thousands of articles, concluded as follows: “ There is limited scientific evidence that the triad and therefore its components can be associated with traumatic shaking (low quality evidence). There is insufficient scientific evidence on which to assess the diagnostic accuracy of the triad in identifying traumatic shaking (very low quality evidence) ”. No study based on observed shaking could be identified. Only two studies in the literature were found to be based on confessed shaking; these confessions came during police custody or judicial investigations, weeks to months after the diagnosis. Confessions obtained under such circumstances are known to be encumbered with uncertainties.  These conclusions underline the importance of considering alternative etiologies for the triad and its findings.

Subdural hematomas will inevitably raise suspicion of child abuse; SDH may, however, also appear spontaneously in infants, often precipitated by one of two congenital conditions, namely, arachnoid cysts or BEH. Several mechanisms may explain why external hydrocephalus predisposes for SDH. To me as a neurosurgeon, the most likely is leakage of small amounts of blood from where the bridging veins enter the dura. It is a common intraoperative observation that even minor manipulations of normal bridging veins during a craniotomy may cause such leakage. In external hydrocephalus, these veins may ooze blood spontaneously just because they are stretched.

Subdural hematomas also occur in newborns, especially in preterm deliveries and twins; our twins were born four to five weeks preterm. Subdural hematomas occur more frequently after vaginal delivery than after a planned Caesarean section and even more frequently after emergency Caesarean sections and forceps or vacuum-assisted deliveries. These subdural blood collections may gradually develop into larger hematomas over time. There are factors in old hematomas that may induce neovascularization in the parietal hematoma membrane, and these pathologic vessels bleed easily. Other factors disturb normal coagulation in subdural blood collections. 

Male preponderance: Our group has recently found a male preponderance of 86% in a population-based epidemiologic study of BEH. Others have documnted a similarly high male preponderance, even as early as in 1944 by Ingraham and Matson.

Adamsbaum et al. and Vinchon et al. also reported male preponderance in their abuse cases, 76% and 64%, respectively. Pooled together, these two studies show a male preponderance of 73%. A male preponderance has also been demonstrated in most published series on shaken baby syndrome or abusive head trauma. The large number (157) of infants in the two studies above renders it unlikely that this male dominance is coincidental. Moreover, in a national register study comprising 306 infants with SDHs, we have recently demonstrated that even in this cohort, there was a clear overrepresentation of males...

To the present author, the twins' MRI and CT scans carry no resemblance to an acutely acquired traumatic hematoma. In this context, it is of interest that the Adamsbaum et al. 44 report included a CT scan (see their Fig 1) of an allegedly shaken infant; the scan appears to show the exact features of external hydrocephalus, as defined by Maytal et al.: slightly widened lateral ventricles, a large extracerebral space, and widening of the frontal interhemispheric fissure. Girard et al. provided a detailed discussion of external hydrocephalus as a differential diagnosis to abusive head trauma.
The only plausible explanation of the combination of an increased or rapidly growing HC, increased extracerebral fluid, and enlarged ventricles is that there is an increased ICP equally distributed within the intracranial compartments, in both the extracerebral space and the ventricles. A traumatic acute or chronic SDH would normally compress the brain, including the ventricles, and if unilateral, would cause asymmetry of the ventricles and a midline shift. These characteristic features were absent in our twins and surprisingly, also in other children published as victims of shaking...

 Bilateral, extensive bleeding in several retinal layers has been regarded as a key feature of abusive head trauma. However, RH may not be pathognomonic for abusive head traumas; they can also be seen in infants not related to abuse, e.g., in a large number of healthy newborns, in infants with “macrocephaly,” after “high-risk” deliveries, following acute life-threatening events, and after cardiopulmonary resuscitation. RHs have also been documented in premature infants; contrary to the rapid resolution of the bleeding one usually sees in most newborns, the bleeding in preterms tend to be long lasting.

As for the rest of the triad, there is no high-quality evidence that proves a causal relationship between retinal bleedings and violent shaking. The problem lies again in the lack of objective proofs of shaking...

My concern in this report is that many infants appear to be diagnosed as shaken babies without any clear signs of an inflicted trauma (impact). The scientifically weak documentation of a causal relationship between the triad and the criminal act of violent shaking without impact cannot be used as judicial evidence for child abuse, beyond reasonable doubt.

When suspecting the shaken baby syndrome or abusive head trauma, external hydrocephalus should always be ruled out as a possible alternative diagnosis, as this condition may have medical findings compatible with those claimed to prove violent shaking. Unfortunately, many physicians who deal with child abuse seem unfamiliar with the manifestations of benign external hydrocephalus in children.

No comments:

Post a Comment