Calculus Comes to the Rescue in Tylenol Poisoning

(Inside Science) -- When a patient is rushed into an emergency room with an overdose of Tylenol or another drug containing acetaminophen, the doctors have to make a fast life-or-death decision -- and a newly developed mathematical technique can now help them.

If the amount of the painkiller taken is relatively small, there is an antidote they can use that prevents the drug from damaging the liver. If the poisoning is more extensive the only way to save the patient's life is a liver transplant. If they wait too long to make the decision or make the wrong one, it will be too late for the transplant.

A team of mathematicians and physicians at the University of Utah have come up with a set of calculus equations which may take the guesswork out of that parlous decision. The computer program will tell the physician almost instantly what to do.

The research is to be published in the journal Hepatology.

Acetaminophen, the main ingredient in Tylenol and equivalent generic-brand pain killers, is found in such popular over-the-counter analgesics as Anacin, Pediacare and Triaminic, and in the more potent prescription drugs Tylenol with Codeine, Percocet and Vicodin. The margin of error in taking acetaminophen is unusually slim for a non-prescription drug.

The normal maximum dose of acetaminophen is 4 grams in a 24 hour period, which is contained in eight tablets of extra-strength Tylenol. Just slightly more than that, six grams (12 tablets), can cause liver damage in some people. And according to the researchers, 20 grams could destroy 70 percent of the liver and could cause death.

The drug is the leading cause of acute liver damage in the United States, producing as many as 56,000 hospital visits, with 26,000 admissions. Five hundred Americans die of acetaminophen poisoning a year, they wrote.

"About half the time that is intentional, someone trying to hurt themselves," said Paul Watkins, director for the Hamner-University of North Carolina Institute for Drug Safety, who was not part of the research team. They are either trying to commit suicide or making a dramatic call for help, he said.

It is so common that when an incoherent, jaundiced patient is brought into an emergency room doctors almost automatically test the blood for acetaminophen, Watkins said.

The liver is the most forgiving of organs, but without a functioning one, life is impossible, after it has been completely destroyed.

"A lot of damage is fixable," said Norman Sussman, one co-author of the study, now at the Baylor Medical College. But the physician has to act quickly.

The current method uses statistics and history, Sussman said. Called the King's College Criteria, it estimates who can be treated with the antidote and who needs a transplant based on the results of tests and statistics from past patients. It has become less useful as treatment has changed, he said.

"Care has gotten better," Sussman said. "What was lethal in the past is less lethal now. You have to keep changing the outdated statistical model."

The technique developed by Utah mathematician Fred Adler, Sussman, doctoral student Chris Remien and their team is "dynamic." Instead of being just statistical, it looks at what actually is happening to the liver of each patient.

They used the medical records of 53 patients at the Utah University hospital, taking the results of four common tests, and using eight differential equations produced a chart showing when the patient took the acetaminophen and how much. The equations predict who is likely to survive with treatment and who must have a transplant.

The study was what scientists call retrospective. The Utah researchers applied their model to data in the records of past patients to see what would have happened had they used the equations.

The equations would have predicted that 12 of the patients would die, including all eight that did die.  The method correctly predicted 39 of the 43 patients who survived.
The team is planning a one-year prospective study, using the new criteria to determine treatment strategy, of 50 patients admitted to Utah and Texas hospitals during the study period.

"I think this idea of moving to computer programming ... is really an important direction everybody is heading -- to use computer modeling to make treatment and prognostic decisions," said Watkins.