Don't treat the poison

Supportive care is key to improving survival.

Treat the patient, not the poison. If you remember just one thing about overdoses, that should be it, said Kennon Heard, MD, during a session on toxicology at the Society of Hospital Medicine's annual meeting in April.

“There are a lot of poisons but really only a few significant symptoms we have to worry about,” said Dr. Heard, an associate professor of medicine and emergency medicine and section chief of medical toxicology at the University of Colorado School of Medicine in Denver. “Probably the most important advice I've ever gotten is to treat the patient regardless of what [he] took.”

Kennon Heard MD Photo courtesy of the Society of Hospital Medicine
Kennon Heard, MD. Photo courtesy of the Society of Hospital Medicine.

Supportive care is key, he added: If you can keep the patient breathing and his blood pressure stable, he will usually be fine.

The symptoms to worry about in the first few hours of a patient's hospital visit are deactivation or activation of the central nervous system (CNS), with the latter manifesting as agitation or seizures; cardiac dysrhythmias; depressed cardiac function; and anion gap acidosis.

Reining in the CNS

Deactivation of the CNS is the most common clinical effect of drug overdose and the most common reason people die, Dr. Heard said. “And the reason patients die is because they lose their airway,” he added.

Common deactivating drugs include benzodiazepines, most seizure medications, opioids, muscle relaxants, ethanol, newer antidepressants, clonidine and antipsychotics, he said.

“Most of these drugs don't cause respiratory depression other than the opioids, so if you can manage the patient's airway, he is going to survive,” Dr. Heard said. “People die from things like hypoxia: If the airway closes off they hypoventilate, or they aspirate.”

Intubation is the usual solution to managing the airway, he said. It's best to use a short-acting sedative—like propofol or midazolam—to avoid prolonging the course of the intubation. Noninvasive ventilation isn't recommended, due to the risk of aspiration, he added.

The typical patient with an overdose of a CNS-activating drug has a high heart rate (for example, 160 beats/minute) and diaphoresis, and is nonfocal. It might be difficult to get his or her blood pressure reading, Dr. Heard said.

Often the cause is cocaine, amphetamines/methamphetamines, anticholinergics, withdrawal from sedatives or ethanol and, less commonly, phencyclidine, he added.

“The bottom line for all these drugs is they act by increasing the excitatory neurotransmitter in the brain. The treatment, therefore, is to increase the inhibitory neurotransmitter,” he said. “You want to turn off the brain.”

Benzodiazepines are the most common treatment, though propofol is sometimes used, as are—rarely—barbiturates. The goal is to get the patient sedated and let the drugs exit his system, Dr. Heard said.

“Once you get the patient under control and have taken care of the things that will kill him, you need to address other complications that may occur, like hypothermia or an end-organ injury. But first things first: Get him sedated,” he said.

Occasionally a patient shows up with an overdose that seems to be activating the CNS—perhaps because seizures are present—but that's not actually the case. This becomes obvious if the patient doesn't respond well to benzodiazepines, he said.

For example, isoniazid poisoning causes seizures by interacting with vitamin B6, which in turn leads to the depletion of gamma-amino butyric acid (GABA), the brain's main neuroinhibitory transmitter. The treatment is to replenish the brain with vitamin B6 (which is required to produce GABA).

Cicutoxin, or water hemlock, and anticholinergics are two other special-case drugs to consider as a cause if a patient isn't responding well to benzodiazepines, Dr. Heard noted.

“Cicutoxin will often respond to very large doses. For anticholinergics that are not responsive to benzodiazepines, we will often use physiostigmine to reverse the toxidrome,” he said. “And general anesthesia is an option for the truly refractory patient.”

Cardiac dysrhythmias

Three types of cardiac dysrhythmias most often accompany drug overdoses: narrow and fast, wide and fast, and slow, Dr. Heard said.

“These are important from a toxicological standpoint because the treatment for…these is different. And some of the treatments are actually different than what you would do with your standard ACLS [advanced cardiac life support] algorithms,” he said.

Narrow, fast arrhythmias in the setting of poisoning are almost always due to a drug affecting the brain, not the heart, so treatment should target the brain, Dr. Heard said. “Sedate the patient first, rather than trying to use something like beta-blockers,” he said.

In the case of a 23-year-old patient who showed up agitated and with a narrow fast arrhythmia after ingesting bath salts, the treatment was two milligrams of lorazepam and two liters of saline. Two hours later, the patient was in sinus rhythm and his left bundle-branch block had resolved, Dr. Heard said.

Wide and fast rhythms in the setting of an overdose are usually the result of a drug that blocks sodium channels, like tricyclic antidepressants, cocaine or local anesthetic. They are usually accompanied by hypotension and coma, and the treatment is sodium bicarbonate boluses, he said.

Slow rhythms are relatively uncommon, and “probably the scariest ones we see,” Dr. Heard said. The cause is usually a calcium-channel blocker or beta-blocker. Digoxin also can cause this kind of rhythm, but usually some ventricular ectopy is also present, he said.

“Drugs that go along with this rhythm will usually also depress myocardial function, so you will get markedly decreased blood pressure,” Dr. Heard said. “You will see decreased blood pressure with wide, fast rhythms, as well.”

In terms of treating the slow rhythms, the clinician should focus more on the blood pressure and less on the rate, he said.

“Most of the time treating the myocardial depression—with vaospressors or specific therapies such as high dose insulin for calcium channel blocker toxicity—is more important than restoring rate,” Dr. Heard said.


Usually in medicine, when one thinks about metabolic acidosis, one thinks about MUDPILES (methanol, uremia, diabetic ketoacidosis, propylene glycol, isoniazid, lactic acidosis, ethylene glycol, salicylates).

From a toxicological standpoint, however, “we talk about KULT—ketones, uremia, lactate and toxins,” Dr. Heard said. “Fortunately I don't have to talk about uremia because I'm not a nephrologist, so I'll concentrate on the other three.”

In illustrating the difference between alcoholic ketoacidosis and toxic alcohol ingestion, he gave the example of a 35-year-old woman who was a heavy alcoholic with no history of toxic alcohol ingestion. She presented at the hospital with headache, weakness and vomiting over a two-day period. She had a moderate decrease in bicarbonate (12 meq/L) and a high anion gap, no detectable ethanol, and an acceptable blood glucose, he said.

The patient also had “a bit of a bump” in creatinine for a young woman and ketones in her urine. Her arterial blood gases indicated “moderate metabolic acidosis with respiratory compensation,” he said.

Regardless of cause, he added, he would treat this patient based on symptoms: two liters of D5 saline and the “standard alcoholic vitamins,” then repeat labs in two hours. At that point, there are two potential clinical courses. If her bicarbonates increased, her anion gap decreased and her creatinine and pH improved, she has alcoholic ketoacidosis. However, if her bicarbonates and acidosis worsened and her anion gap widened, then she has toxic alcohol ingestion.

In the latter case, you would give the patient an alcohol dehydrogenase inhibitor and probably call a nephrologist, Dr. Heard said. “All of this is still going to happen faster than you can get back an ethylene glycol concentration from most labs,” he said, stressing the importance of treating symptoms before determining an exact cause.

“So if you're faced with someone and aren't sure which of these they are, start treatment and reevaluate frequently,” he said. “If they are getting worse, it's toxic alcohol and you begin therapy towards that. If they are getting better, it's alcoholic ketoacidosis, and you continue current therapy and they will continue to improve.”

Toxic lactic acidosis, meanwhile, comes in two forms. If the patient exhibits hypoperfusion, seizure, agitation and liver failure, then the cause is a toxin. “The patient is going to produce lactate as would someone who is septic,” Dr. Heard said. “[He] is most likely in shock and you should treat the shock state.”

If those symptoms aren't present, the toxic lactic acidosis is from mitochondrial poisoning, as from cyanide, carbon monoxide, metformin, antiretrovirals or aspirin. “You will direct therapy against those, so specific antidotes or hyperbaric oxygen, maybe dialysis for metformin,” he said.

If you get stuck on any of the treatments or diagnoses, however, remember that help is at hand, Dr. Heard said.

“Toxicology is one of the few areas where there is always someone to consult if you have questions and you aren't quite sure about something,” he said. “Call the poison center.”