Volume 3, Case 14
Loren G. Yamamoto, MD, MPH
Kapiolani Medical Center For Women And Children
University of Hawaiʻi John A. Burns School of Medicine
Paramedics are called to the home of a 3-year old male child because the child is noted to be poorly responsive. Upon arrival 11 minutes after the 911 called was received, the child is being carried by an adult outside the home toward the ambulance. Paramedics note the child to have agonal respirations. His EKG rhythm shows a bradycardia with multifocal PVC's. He is mask ventilated, then intubated at the scene. His rhythm improves to a sinus tachycardia. He is noted to have fixed and dilated pupils at the scene with a Glascow coma score of 3. He is transported to a rural emergency department.
Exam in the ED: VS T36.4 (axillary), P120, R 45 (bag ventilation), BP 80/60. Weight 10 kg. He shows no significant neurological response. His pupils are fixed and dilated. He is small for age and emaciated in appearance.
Initial laboratory studies:
He is given IV fluids and furosemide. He is then transferred to a children's hospital for further management.
Upon arrival, his management is assumed by a new set of physicians. His pupils are still fixed and dilated. No neurological response is noted. Retinal hemorrhages are noted on fundoscopy.
His laboratory tests are repeated:
A CT scan of the brain is performed.
View CT scan.
The image on the left is taken through the orbits. This cut is significant for hemorrhages noted over the surface of the retina. The image on the right is taken through the brain and the lateral ventricles. Although IV contrast has NOT been administered, the falx appears to be prominent. This white enhancement represents hemorrhage in the interhemispheric space. It is most prominent posteriorly. This represents a posterior interhemispheric subdural hematoma. There is evidence of cerebral edema and a slight midline shift.
This CT scan is pathognomonic of a shaken child injury (see Case 1 in Volume 1, Toxic Infant with a Full Fontanelle). The retinal hemorrhages are also highly indicative of child abuse.
Retinal hemorrhages are usually identified on fundoscopic examination. For medical legal reasons it may be best to have these substantiated by an ophthalmologist. In subtle cases, the retinal hemorrhages may not be seen on direct ophthalmoscopy, thus, an ophthalmologist is usually needed to perform indirect ophthalmoscopy. Most retinal hemorrhages are not visible on CT scan. Thus, CT scan is not useful to rule out the presence of retinal hemorrhages.
The finding of a prominent posterior falx on an unenhanced CT scan (no contrast) is indicative of a posterior interhemispheric subdural hematoma. In an infant, this classically is seen in the shaken infant syndrome. Blood also enters the subarachnoid space. Thus, if an LP is performed, it will most likely be grossly bloody.
The hypernatremia in this case is extreme and is what makes this case particularly interesting. One should be highly suspicious that this degree of hypernatremia may be long standing. Thus, a rapid correction of this hypernatremia may be clinically detrimental. It may be best to correct this slowly. Clinical hydration and neurological parameters should be followed closely to maintain a fluid balance most appropriate for the clinical situation. Rapid fluid and electrolyte shifts may result in cerebral edema. Fluid boluses may be required to correct hypovolemia and diuretics agents may be necessary to reverse cerebral edema; however, all such agents should be administered with extreme caution.
According to the child's father, they were at the beach four hours before calling the ambulance. The father states that the child was sitting in the water and began drinking salt water. The father told him to stop and he did so. The family went home and the child was playing and watching television. Three hours after this episode, while at home, he complained of some abdominal pain and vomited once. He then developed respiratory difficulty and rapidly worsened, prompting the 911 call.
This degree of hypernatremia is not possible from drinking sea water for a short period of time. Salt water near-drowning victims do not have this degree of hypernatremia. Whenever the history of events as described by the caretaker is not consistent with the clinical findings, child abuse should be suspected. In this case, the description of the child drinking some sea water at the beach in the afternoon, playing happily later that afternoon, then being found in a pre-arrest state a few minutes later by paramedics with extreme hypernatremia, retinal hemorrhages, and the posterior interhemispheric subdural hematoma noted on arrival at the hospital are an impossible sequence of events.
Deliberate poisoning of children by their caretakers is a recognized syndrome of child abuse. Sometimes, this is part of Munchausen Syndrome by proxy, but in most instances, it is just another form of inflicted harm on a child. Non-accidental salt poisoning is a common type of chronic poisoning administered to children by caretakers. Although many practitioners have not heard of this, it probably occurs more commonly than most believe. It may only present to medical attention if the poisoning results in severe hypernatremia.
Substantiated cases of salt poisoning are associated with severe hypernatremia, usually above 160 mEq/liter. It is sometimes in excess of 200 and is often in the 170 to 190 range. This finding is often found in association with other signs of physical abuse such as fractures, retinal hemorrhages, burns, failure to thrive, and emotional deprivation. In many of these instances, salt administration is used as a form of punishment.
The following case examples of children with extreme hypernatremia are most illustrative. In each instance, their parents' explanation for the hypernatremia was inconsistent with clinical findings.
a) Parents of one child said they had used salty foods to treat the child's sudden onset of voracious appetite and thirst. These behaviors were not observed while the child was hospitalized.
b) A 5-year old child was given spoonfuls of salt by his mother for enuresis. He was also noted to have burns on his feet and ankles.
c) A 6-year old child died of hyperkalemia and hypernatremia after eating food his stepfather had heavily seasoned with "light salt" (a salt substitute consisting of KCl and NaCl) as punishment to break the child's habit of adding too much salt to his food.
d) A 3-year old child was found with a cup of salt beside her crib.
These instances are not as benign as they sound. For the serum sodium to elevate this high, the child must be deprived of water and/or salt must be forcibly administered. In an attempt to re-create a serum sodium of 170 by mere salt administration as explained by the child's mother, investigators were able to administer only 20 grams of salt with great difficulty to the child which resulted in a maximum serum sodium of only 147. This implies that salt must be forcibly administered.
Some mothers were noted to put excessive amounts of salt in their infant's formula. Although two teaspoons of salt may not sound like much, this amount is capable of elevating one's sodium to 200 although one's kidneys would generally excrete as much sodium as possible to prevent this from happening. Two teaspoons of salt have a very strong taste and when added to formula, infants will reject it. Thus, only when conventional fluids and formula are withheld, would an infant be desperate enough to drink such salt-laden formula.
It is important to rule out organic causes of hypernatremia. Renal function should be ascertained, and normalization of the serum sodium under hospital or foster care with normal feedings should be documented. A urine sodium obtained while the child is hypernatremic should be obtained. Hyperaldosteronism and diabetes insipidus are associated with low urine sodiums suggesting inappropriate sodium retention, while salt poisoning is associated with highly elevated urine sodiums (the kidneys are attempting to correct the hypernatremia by excreting sodium). Hypernatremic dehydration secondary to gastroenteritis may mimic many of these findings, however, gastroenteritis and dehydration are usually associated with an elevated BUN, while salt poisoning does not result in as much azotemia. While hypernatremic dehydration generally results in only modest sodium elevations, salt poisoning is associated with extremely high degrees of sodium elevation.
The absence of polyuria by history makes diabetes insipidus less likely. However, the presence of polydypsia is often seen in salt poisoning in an attempt to compensate for the hypernatremia and/or fluid deprivation. Such children have been observed to lick water off windows and to drink from puddles, toilets, and fish tanks.
Although accidentally (or out of ignorance) administering undiluted formula concentrate to infants usually results in hypernatremia, this type of unintentional hypernatremia is usually not as extreme, and the other associated findings such as failure to thrive or inflicted injuries are not present.
Although initial interviews with parents guilty of salt poisoning their children did not reveal a willful attempt to harm the child, in repeated interviews months later, some parents confessed to wanting to kill their child. In a few instances when salt poisoned children were returned to their parents, the salt poisoning behavior recurred despite the parents' knowledge that this was harmful. This suggests that such perpetrators are severely disturbed, and these children should be placed in protected environments away from the perpetrator(s).
Although speculative, it is likely that there are lesser degrees of salt poisoning that result in only modest or transient hypernatremia, or hypernatremia that is difficult to distinguish from hypernatremic dehydration due to gastroenteritis. This may not be very harmful unless it leads to more severe salt poisoning. It is probably prudent to routinely question the caretakers of any child with even mild hypernatremia for the possibility of salt administration. If this inquiry is suspicious, or the child has any other high risk factors (failure to thrive, fractures, burns, inappropriate social behavior, developmental delays, etc.), frequent clinical and laboratory follow-up monitoring for signs of salt poisoning or other forms of child abuse and neglect would be in the child's best interest. Reporting a case to the local child protective service authorities would enable one to determine if any other suspicious events have ever been reported about the child.
In the case of our patient, one could speculate that this child was chronically salt poisoned. Because of extreme thirst, he began drinking sea water when he went to the beach that day. He also endured an acute or possibly chronic shaking episode(s) resulting in cerebral and retinal hemorrhaging. This type of injury results in axonal shearing and cerebral edema. If this was acute and severe enough, it may account for his vomiting and rapid demise. Rapid changes in serum osmolarity may have also contributed to the cerebral edema. The father's history of events could not possibly account for the child's clinical findings. This child was small for age. Failing to thrive at a mere 10 kg at age 3 years, he was not brought in for routine medical care. This family was previously known to the local child protective services, who had received reports in the past of suspected child abuse and neglect.
References