A patient is brought back from triage. He is uncooperative with your exam and it is difficult to obtain vital signs. His speech is mostly incoherent, and he is drooling. You notice with disgust that he is chewing on the bedrails. He is incontinent of urine and feces, but this is apparently his baseline. The good news: he is 2 years old, and this is all normal. The bad news: he was found with an open bottle of pills, and won’t confess to a thing.
As a pediatric emergency physician, toxicologist, and mother of 3, I’ve waged war on many fronts with much tinier humans. We pediatrician types love to discuss the differences between adults and our patients, who are “not little adults”. Pediatric toxicologists are no exception. But what’s the big deal, really? How often does it happen that the management of the poisoned child can’t be generalized from overall principles?
In truth, children do sometimes develop physiologic responses that are very similar to their adult counterparts. And often they can be treated in similar ways. It is not altogether inappropriate to treat a child based on size alone – it’s the principle on which length-based resuscitation (e.g., Broselow) tapes are based. But there are a lot of key pathophysiologic differences in the pediatric poisoned patient.
First, let’s define the patient population. A “pediatric” case report that details a 17-year-old overdose patient doesn’t add much to the understanding of poisoning in the young child. There are nuances all through development of course, and the smaller the child, the more pronounced the differences. As a general rule, a 12-year-old has about the same vital signs as an adult, and much of the same characteristics for the purposes of poisoning management. The following pertains primarily to the 6 and under crew, (about 0-40 kg) who comprise a majority of poisonings reported to Poison Centers every year (Mowry 2016 PMID: 29185815).
1) They ingest a higher dose per kilogram than an adult in similar circumstances. This is the obvious one. The outcomes in pediatric poisonings are generally better than in adults, because the preponderance of exposures in kids are exploratory. However, an exploratory ingestion of particular compounds is much more consequential in smaller patients. Thus the “one pill can kill” phenomenon, for things like calcium channel blockers and opioids (especially methadone). This explains why the same agents show up year after year as a leading cause of pediatric fatalities. This also applies to things like envenomations. A snake injecting venom doesn’t calculate a dose. Neither does a scorpion, which is why children become much sicker with North American bark scorpion (Centruroides) envenomations than adults.
2) They have a higher body surface area per kilogram, so toxins absorbed dermally are delivered in a higher dose (van den Anker 2011 PMID: 21882105). Think patch medications, like fentanyl or clonidine. They also have increased skin perfusion and hydration, so are more susceptible to dehydration and insensible losses. Consider for example a nerve agent or organophosphorus pesticide, some absorption of which will occur through the skin, delivering a higher dose. When that occurs, profuse sweating and overall secretory excess ensues. A young child will become hypovolemic from these losses faster.
3) They have a higher minute ventilation (tidal volume x RR) per kilogram. While the tidal volume is about the same as adults (approximately 7 ml/kg) – the respiratory rate is faster in young children (and canaries) delivering a higher dose of an airborne toxin. The most classic example of this is carbon monoxide poisoning. All of the occupants of a house (or coal mine) might become sickened from a CO exposure, but the smallest children are often most affected (chirp).
This is also a concern with nerve gas attacks and other agents of opportunity such as chlorine, phosgene, and others.
4) Their compensatory mechanisms are not the same (Calello 2014 PMID: 24275168).The good news is, they have healthy well-perfused myocardia which can tolerate tachycardia for days. It’s what allows us to give endless beta-agonists to a 5-year-old asthmatic with impunity. Children also maintain a higher adrenergic tone at baseline. This gives the illusion of well-compensated shock until cardiac output falls to a critical extent, and the bottom drops out. Speaking of cardiac output (a product of stroke volume and heart rate) – a young child can’t budge their stroke volume very much, so heart rate is all there is. Enter a toxin which causes bradycardia, such as calcium channel blockers or digoxin, and bad things happen quickly.
From a respiratory standpoint, a higher metabolic rate means kids desaturate much faster during periods of apnea or respiratory insufficiency. Reliance on the diaphragm over other accessory muscles leads to fatigue faster, which impairs their ability to maintain adequate oxygenation and ventilation after exposure to pulmonary toxins such as hydrocarbon aspiration. This can also influence a child’s ability to maintain hyperpnea in the face of acidosis or salicylate toxicity, and explains why children are more prone to acidemia in these situations.
A typical toddler does not have much in the way of glycogen stores or muscle mass. Fasting hypoglycemia is more likely to occur after exposures to toxins that alter glucose homeostasis, like alcohols and beta receptor antagonists. This is the story behind the child who can’t be woken up the morning after his parents have a party at the house: he drinks half a cocktail, goes to bed, and has a blood sugar of 32 in the morning.
5) Their pharmacokinetics are not mature. Developmental differences (often referred to as ontogeny) come into play in all phases of drug disposition (absorption, distribution, metabolism and elimination), although some are more pronounced than others. In the absorption phase, subtle changes in gastric absorption and transit time may influence the delivery of toxin from the GI tract into the circulation and target tissues. Children are made up of more total body water than adults (80-90% in an infant compared to 55-60% in an adult), and less fat, so hydrophilic drugs will have a larger volume of distribution in a watery, skinny kid (van den Anker 2011 PMID: 21882105).
The p-glycoprotein efflux transporter is responsible for getting drugs out of compartments, most notably the central nervous system.. It is underdeveloped until later childhood, and some compounds, like buprenorphine, utilize it. If you don’t have enough p-glycoprotein, buprenorphine won’t be transported out of the brain, and you’ll have more respiratory depression. This is exactly what happens in toddlers with buprenorphine exposures, and why there is more respiratory depression and more death in this group. There’s no way to efflux it out of the CNS once it’s in.
For metabolism, the differences abound. All of the enzyme systems we rely on for the metabolism of drugs (this means you, CYP2D6, 3A4, and the rest of the p450 gang) don’t fully develop until several years of age. One example of this is the theoretical “resistance” to acetaminophen hepatotoxicity. Sulfonation is brisker than conversion to NAPQI (a p450 process), even with high doses. Other examples include metabolism of opioids (2D6), midazolam and carbamazepine (3A4), phenytoin (2C9), caffeine/theophylline (1A2), codeine (2D6), benzodiazepines (2C19) and acetaminophen (2E1)….to name a few.
Lastly, elimination of renally excreted compounds is affected by maturity of the kidney. This is why residents in the NICU must constantly check serum concentrations of aminoglycosides: the GFR is a moving target. It’s sluggish at birth and increases to normal rates around 8 months of age or so.
6) Their hemoglobin is not the same. In the first 2-3 months of life (longer if there is a hemoglobinopathy), fetal hemoglobin persists and has increased affinity for oxygen. This can falsely elevate carboxyhemoglobin measurement on co-oximetry. There is also an age-related deficiency in methemoglobin reductase enzymes, making infants more prone to methemoglobinemia. This is why we hesitate to give nitrates in smoke inhalation to children and go straight to sodium thiosulfate or hydroxocobalamin. It’s also one of the reasons why infants under one year with gastroenteritis are prone to showing up a bit dusky (Pollack 1994 PMID: 8092592).
7) They have increased susceptibility to neurotoxins. There are windows of vulnerability throughout the prenatal and postnatal periods and beyond during which neurotoxins can create irreparable harm. Historically, this is best illustrated by the Minamata Bay disaster where methylmercury contaminated water led to children born with tragic malformations. After a Union Carbide pesticide plant leaked methylisocyanate gas into the town of Bhopal, India, causing thousands of immediate casualties, countless babies were born to exposed mothers with “twisted limbs, brain damage, (and) skeletal disorders” (Daily Mail 2014). By many accounts this continues today as a result of environmental contamination. The most discussed example of this is exposure in the US is to low levels of environmental lead, which can cause intellectual decrements and behavioral disorders. If the developing brain gets poisoned, it doesn’t develop normally. And that doesn’t reverse well, if at all.
8) There is a higher likelihood of medication error. A review of pediatric poisoning fatality data throughout the years has demonstrated a consistent pattern: serious medication errors happen in children, particularly medically complex children. A child in an ICU who gets one med via gastrostomy tube and another through a Broviac is an unfortunate setup to have the routes reversed one day and get enteral feeds IV. An infusion which can be given to an adult in a liter of D5W will cause hyponatremia in a 3-year-old (Sung 1997 PMID: 9282711).
Weight-based dosing is a necessity in pediatrics, but fraught with error. This is especially true with unfamiliar medications or in children whose weights are not typical for age. Even the way we document a child’s weight can contribute to the problem. Weight is most commonly recorded as kilograms, to facilitate mg/kg ordering. If instead it is measured in pounds – 2.2 times the kilogram weight – and the clinician doesn’t notice this detail and uses the weight for mg/kg dosing, there’s a 2.2-fold overdose. Units are important. Take for example the dosing of IV N-acetylcysteine: children whose dose is calculated in milliGRAMS/kg, but instead ordered in error as milliLITERS/kg, will receive a 10-fold overdose. This can and does happen (Dart 2012 PMID: 22271694).
9) Treatment can be tricky. Administering IV fluids quickly is hard through a 24 gauge IV in the hand, but obtaining a larger gauge is often technically difficult – which is why we sometimes opt for intraosseous (IO) access. Lavage is basically impossible because of tube size. Activated charcoal requires hostage negotiation skills (and a rain poncho). Performing hemodialysis on a young infant can be challenging and might require small-volume tubing, specialized priming solutions and close (closer) attention to fluid status. This is a job for a tertiary care center.
10) An inborn error may be lurking behind that poisoning diagnosis. Just when you think you’ve got ethylene glycol poisoning, BAM! It’s methylmalonic acidemia! (Shoemaker 1992 PMID: 1538288). Beware the anion gap. Buried beneath that MUDPILE is the rest of the story. If something doesn’t fit, or there are repeated episodes (or similar episodes in a sibling), suspect an inborn error. Or worse, Munchausen syndrome by proxy.
What’s the good news? (Is there any?) Fortunately, the majority of pediatric poisonings are small-volume, exploratory ingestions. They eat soap, not Seroquel. Mostly we observe and discharge, after whatever period of time seems safe depending on the characteristics of the exposure. And that’s a really good thing.