POST-MORTEM DRUG TEST ERRORS INCREASING

A technique for inferring how much of a drug a patient has taken may
be putting innocent people behind bars.

The problem seems to be that doctors are incorrectly applying the
method to corpses, in a bid to establish how much of a drug a deceased
person took, or was given, before their death. That error can result
in vastly inflated readings.

"There is no relationship between what you find in a living person and
what you find in a dead person," Bruce Goldberger, vice-president of
the American Academy of Forensic Sciences and editor of the Journal of
Analytical Toxicology told New Scientist.

To spread the word about the error, Steven Karch, a pathologist at the
San Francisco Medical Examiner's office set up an ad hoc committee of
senior forensic toxicologists and pathologists, including Goldberger,
at the American Academy of Forensic Sciences conference in Dallas,
Texas, in February. They plan to publish scientific papers in general
medical journals highlighting the issue.

Fudge Factor

In living patients, the dose someone has been given is calculated by
multiplying their body weight, the concentration of the drug in their
blood and a constant called the "apparent volume of distribution" (Vd).

The Vd is a fudge factor that averages out the distribution of a
particular drug between tissues in the body. The Vd of any particular
drug depends on how it interacts with an individual's cells.

Studies have suggested that the Vd for some drugs can vary fivefold or
more between different individuals, and the results also depend on
when the drug was administered. So the technique is approximate, at
best, even in the living.

But Derrick Pounder at the University of Dundee, UK, another member of
the committee, says the method fails completely in a dead body. "There
is an assumption on the part of some people that a corpse is a frozen
living person," he says. "But drug levels don't remain static after
death."

As cells die off they release the drug back into the bloodstream, so
the concentration can shoot up 10 times, says Pounder. He published a
paper in 1990 demonstrating this effect, but some expert witnesses are
still testifying in court that they can work out the dose a victim
took from post-mortem measurements.

Junk Science

Goldberger says the problem is that doctors without forensic
experience are applying the formula blindly. "People are acting
outside their field, it is junk science," he says.

Karch has testified in American, British and Australian court cases
and believes the mistake is becoming more widespread. He says around a
third of cases where the technique has been used to evaluate drug
levels involve fatal workplace accidents. Companies may try to shift
the blame onto the employee by claiming they were under the influence
of drugs when the accident happened.

But what if a conviction largely rested on this formula? In 2001, an
Arizona court found Brian Eftenoff guilty of murdering his wife, Judi.
The forensic expert for the prosecution claimed that high levels of
cocaine in Judi Eftenoff's blood indicated she died of an overdose
administered by him.

"I think that was a horrific miscarriage of justice," says Karch. In
his testimony for the defence, he argued that it was impossible to say
from post-mortem measurements whether she had even died of an
overdose, let alone at the hands of her husband.

We need to get this information out there, he says. "People are going
to jail."