Why young children die from local anesthetic overdoses in dental offices

A parent brings a well-behaved child who weighs under 66 pounds into a general dental office and asks for 16 teeth to be treated in one visit. The drive is long, parking is expensive and missing two or four days of work is out of the question. Dr. Stanley Malamed, a dentist anesthesiologist, says agreeing to that request is the first hole in a chain that can end with the child in a seizure.

Malamed and Dr. Jason Brady, also a dentist anesthesiologist, laid out that chain in a recent episode of Health First Talks, the podcast brought to listeners by HealthFirst and hosted by Michelle Dawn Mooney. Their subject was local anesthetic systemic toxicity, or LAST, the term that has replaced the word overdose: the drug leaves the nerve it was injected beside, enters the bloodstream and reaches the brain, its target organ, and when the brain's blood level climbs high enough the overdose begins. The stakes belong to the general dentist who treats children: the drug produces the problem, Malamed says, but panic and mismanagement produce the death.

The protocol a clinician would reach for is itself in flux. The American Society of Regional Anesthesia and Pain Medicine (ASRA) published its current LAST treatment checklist on November 1, 2020, and says on the same page that the checklist is now under review. In a May 2025 feature, the dental trade publication Oral Health Group estimated LAST at 0.27 cases per 1,000 local anesthetic administrations and called it possibly the most lethal adverse event a local anesthetic can cause.

A parent's request, a plain drug and four quadrants at once

Malamed uses the Swiss cheese model of human error: hold three or four slices at random and the holes rarely line up, but on the occasion they do, the overdose passes straight through. Each step in his sequence is harmless alone. His authority on the sequence comes from the cases attorneys have brought him since he wrote a textbook on local anesthesia.

The study he cites is by Helen Lee and colleagues at the University of Washington, published in Paediatric Anaesthesia in June 2013 and archived by the National Library of Medicine, which summarized pediatric deaths associated with dental sedation and general anesthesia. Malamed's case count is his own, drawn from the legal cases attorneys have brought him since the textbook appeared in 1980.

The patient in Malamed's scenario is young, lightweight, well-behaved and in need of work on 16 of 20 teeth. The dentist is a recent graduate building a general practice rather than a pediatric specialist, under pressure from a parent to finish in one appointment. Every pediatric dentist Malamed has asked refuses to do 16 teeth in one visit.

The second hole is the drug. Three percent mepivacaine without a vasoconstrictor, an additive that narrows blood vessels at the injection site, is in Malamed's view a very good anesthetic for one quadrant, one of the four quarters dentists divide the mouth into. Beyond one quadrant in a young child, he says, epinephrine always goes in the cartridge.

The third hole is injecting all four quadrants at the start rather than numbing one side, working it, then injecting the next, which spreads the dose over an hour or two. The fourth is the dose itself: a full cartridge holds about 1.8 milliliters, and the general dentist gives the child the same full cartridge an adult would get. Malamed's comparison is a parent choosing between an adult aspirin and a pediatric one for a child with a fever.

Oral Health Group's May 2025 dosing table shows the gap in milligrams. It caps 2% lidocaine with epinephrine at 7 milligrams per kilogram and 500 milligrams in total, about 13 cartridges for a 70-kilogram adult, and it lowers the ceiling for 4% articaine in children to 5 milligrams per kilogram from 7. By arithmetic on that table, a child at the 66-pound line, roughly 30 kilograms, tops out near 210 milligrams of lidocaine, less than half the adult total.

The same feature gives the physiology behind the lower ceiling: muscle acts as a depot that soaks up circulating anesthetic and holds the blood level down, so a child with little muscle mass has less reserve against the same dose. Four quadrants of full cartridges can reach that ceiling before the first tooth is touched.

The American Academy of Pediatric Dentistry (AAPD), in its current best practice on local anesthesia for pediatric patients, lists the patient's medical history, developmental status, age and weight among the factors that govern anesthetic selection. It asks dentists to document the anesthetic chosen, the dose given, the injection type and the location.

Cartridge myopathy and the milligram math most offices skip

Brady calls the failure mode cartridge myopathy: the operator counts cartridges rather than the running total of milligrams across a long case, measured against the maximum recommended dose (MRD), the ceiling set for each drug by patient weight. A clinical effect that has worn off, he notes, does not mean the blood level has fallen, so a second round of injections stacks on the first. Body mass, liver dysfunction and heart failure all shift where a given patient sits on the curve.

By arithmetic on that rule, a child at the 66-pound line has about three cartridges of headroom. Malamed's scenario puts four full cartridges in that child at once, with no epinephrine to slow the drug's exit from the nerve. Brady's caution is that the rule is a green thumb and nothing more, and that a busy operator who trusts it can lose sight of the total.

The arithmetic is short: the percentage printed on the cartridge, multiplied by ten, gives milligrams per milliliter, so 4% articaine carries 40 milligrams in every milliliter and about 72 in a full cartridge. Brady put a multiple-choice question on screen to show how the labels mislead, and the option with the largest volume was far from the option with the most drug.

Mixing drugs compounds the problem. Brady's rule is that maximum doses add up across agents, so a patient who has received most of one drug's maximum has only the remainder available for any other, never a fresh allowance.

Brady preps a tray holding 90% of the patient's maximum recommended dose and nothing more. The dentist does not have to use all of it, but a distracted operator cannot reach past it. Where offices go wrong, he says, is keeping every cartridge within arm's reach and losing count in the middle of a busy case.

Warning signs are missing in many cases, which bends the model

The chain Malamed describes assumes the team notices the climb. Brady's early signs run from a metallic taste, ringing in the ears, dizziness and confusion to muscle twitching, before generalized seizures, respiratory depression and cardiac collapse, and sedation given on top of the local can mask all of them. His reading of the case literature says the prodromal signs, the warning symptoms that precede the main event, are frequently absent altogether.

Brady did not name the publication behind that reading on the episode. The pediatric case literature he points to does exist in peer-reviewed form: Regional Anesthesia & Pain Medicine published a review of recent pediatric LAST case reports by Archana Singaravelu Ramesh and Karen Boretsky online in June 2021, with a July 2021 infographic distilling it. Oral Health Group's May 2025 feature reports cardiac toxicity precedes nervous system signs in about 20% of atypical LAST cases and puts the usual onset at one to five minutes after injection. Brady's explanation is that the heart muscle is highly sensitive, and a high concentration diffusing into it destabilizes its electrical rhythm.

The mortality data carry a caveat of their own. Lee and colleagues wrote in June 2013 that few data existed to quantify pediatric morbidity and mortality from dental anesthesia, so their percentages are shares of the deaths they could find rather than a rate.

Brady's answer to the silent case is timing. A patient with no seizure history who seizes right after a local anesthetic is, in his words, a tell sign, and the diagnosis does not require every box on the symptom list to be checked.

Position, airway and a 911 call carry most patients through the seizure

Brady's first steps are blunt: stop injecting, stop the procedure, alert the team and activate emergency medical services (EMS), because a dental office cannot manage the full event. He warns that some teams have treated what looked like a simple seizure and missed the toxicity behind it.

The airway matters because of chemistry. An obstructed patient becomes hypoxic, short of oxygen, and hypercapnic, retaining carbon dioxide; the blood turns acidotic and cerebral blood flow rises, which carries more drug into the brain. Keep the patient oxygenated and ventilated, Brady says, and the brain's drug level falls while the seizure runs its course.

Malamed says a LAST seizure usually ends within about a minute if the airway stays open. His protocol for the general dentist is supine position, lying flat on the back, with head tilt, chin lift and protection from injury, since most patients keep breathing through the seizure with their blood pressure up. Brady adds a pulse oximeter, the fingertip clip that reads blood oxygen, and readiness with a non-rebreather oxygen mask or bag-valve-mask ventilation.

If the seizure does not stop, Brady says the next step is a benzodiazepine, a sedative class that suppresses seizures, given intravenously. Malamed is candid that most listeners cannot start an IV and do not stock the drug, so the realistic plan is positioning, airway, someone else dialing 911 and a clear handoff to paramedics. Brady wants that handoff to say the event followed a local anesthetic, since paramedics otherwise work through the H's and T's, the standard list of reversible causes of cardiac arrest.

ASRA's checklist formalizes the sequence. The society built its first checklist in 2010, replaced the 2012 version with a 2017 update published in Regional Anesthesia and Pain Medicine in February 2018, and had Joseph Neal, Erin Neal and Guy Weinberg revise it again in November 2020. ASRA posts it free for rescue kits, and Oral Health Group's May 2025 feature frames benzodiazepines as a step for dentists trained to give them.

Lipid emulsion pulls the drug out of the heart, and almost no dental office stocks it

If the event reaches the heart, Brady says intralipid, a 20% lipid emulsion given intravenously, is the definitive treatment and nothing else is. He also says very rarely does any dentist carry it, and dentist anesthesiologists themselves are few and far between. Time, in his words, is not the dentist's friend.

Oral Health Group dates lipid emulsion's status as the definitive LAST treatment to 2010. The mechanism, as the feature describes it, has two parts: the emulsion improves cardiac perfusion, conduction and contractility, meaning blood flow through the heart muscle, its electrical signaling and the force of its beat, and it sequesters the anesthetic away from its targets so the body can eliminate it. Sarah Ciechanowicz and Vinod Patil wrote in Anesthesiology Research and Practice in September 2011 that local anesthetic cardiac arrest resists standard resuscitation and that awareness of lipid rescue among practitioners was lacking.

The American College of Medical Toxicology (ACMT) is more guarded. In interim guidance posted on its site in June 2022, the college calls the human evidence highly anecdotal, says no standard of care requires using or withholding lipid therapy, and leaves the choice to the treating physician's judgment. It still calls lipid therapy a reasonable option when a highly lipid-soluble drug, one that dissolves readily in fat, causes serious hemodynamic instability, meaning blood pressure and circulation that will not hold, even before cardiac arrest.

ACMT's suggested regimen is a 1.5 milliliter per kilogram bolus, a single rapid dose, of 20% emulsion over two to three minutes, then an infusion at 0.25 milliliters per kilogram per minute with vital signs recorded at least every 15 minutes. The bolus may be repeated for patients with no cardiac output, and the college names Intralipid because that product was used in most published studies and case reports.

Standard advanced cardiac life support (ACLS) also changes. Oral Health Group's feature says clinicians should avoid lidocaine, beta-blockers, calcium channel blockers and vasopressin and cut any epinephrine to 1 microgram per kilogram to avoid arrhythmia. Brady gives the reason: too much epinephrine constricts the coronary arteries at the moment the team is trying to clear the anesthetic out of the heart.

A QR code, a calculator and an ASRA update due in 2027

Brady has built a downloadable cognitive aid, a reference designed to be read under stress, reachable from a QR code shown on the episode. It holds a crisis checklist and a local anesthesia calculator listing every local anesthetic sold in the United States with dosing in pounds and kilograms and a maximum cartridge count. A team planning 4% articaine alongside 2% lidocaine can read off each drug's share of the combined maximum before the patient is seated.

Malamed's final instruction is to calculate the MRD, stay well under it, and use as little anesthetic as produces good anesthesia. Oral Health Group's feature gives that instruction a name, the minimum effective dose: the lowest concentration and lowest volume that produce profound anesthesia for the length of the procedure. An overdose is uncommon, Malamed says, but the outcome when it happens is usually bad, and he calls local anesthetics the safest, most effective pain drugs in all of medicine on one condition, that they are used properly.

The next dated milestone is ASRA's revised LAST checklist, expected in early 2027. Brady's rule holds in the meantime: a maximum recommended dose is a guideline, never an aiming point.