the article about overstating oxycontin deaths
ptpeet at nyc.rr.com
Tue Aug 3 14:50:46 EDT 2004
Sentinel overstated deaths caused solely by oxycodone
Posted August 1, 2004
An Orlando Sentinel series in October about the drug OxyContin used a key
statistic incorrectly and overstated the number of overdoses caused solely
by oxycodone, the active ingredient in OxyContin and other prescription
Citing data compiled by the Florida Department of Law Enforcement, the
Sentinel said oxycodone overdoses caused 573 deaths in 2001 and 2002.
However, a re-examination of FDLE data and autopsy reports showed that only
about a quarter of those deaths were caused solely by oxycodone. In roughly
three out of four cases, medical examiners concluded that at least one other
drug also contributed to the victims' deaths.
In addition, the series created the misleading impression that most
oxycodone overdoses resulted from patients' taking the drug to relieve pain
from medical conditions. But many of the victims were clearly drug abusers
using illegal drugs along with oxycodone.
According to the Sentinel's re-examination, blood samples in about 38
percent of the oxycodone-related deaths showed the presence of heroin,
cocaine, methamphetamine and/or marijuana. Many other victims also had
consumed one or more commonly abused prescription drugs, such as Xanax or
In February, the Sentinel published a story correcting factual errors about
two men featured in the series. The newspaper had labeled one of them, David
Rokisky, an "accidental addict" without doing background reporting that
would have shown he had a federal drug conviction. The other, the late Gerry
Cover, died from an overdose caused by a combination of drugs rather than
The Sentinel began re-examining the data this spring after Purdue Pharma,
the manufacturer of OxyContin, questioned the paper's methodology and
The reanalysis showed that oxycodone's role in drug-abuse deaths is complex,
and that the Sentinel had characterized it inaccurately.
The records do support the series' conclusion that oxycodone -- available in
OxyContin as a long-lasting, powerful narcotic -- is involved in a large
number of the state's overdose deaths. "The [law-enforcement] folks would
tell you it really is a problem; they're seeing a lot of oxycodone," said
Dr. Stephen Nelson, chairman of the state's Medical Examiners Commission.
But the reanalysis also showed that oxycodone is usually found in the
presence of other drugs.
The Sentinel's five-day series, which ran from Oct. 19 to Oct. 23, focused
on oxycodone addiction and the painkiller OxyContin. The drug was approved
in 1995 for treatment of persistent, moderate to severe pain. Designed to
provide relief for up to 12 hours, the time-release drug also caught the
attention of drug abusers, who seek the powerful high it can produce when
crushed and snorted or dissolved in liquid.
The Sentinel's reanalysis discovered that the annual drug reports published
by FDLE exaggerate the number of overdoses caused by a particular drug by
counting multiple times deaths involving more than one substance. For
example, someone who died from an overdose of oxycodone and cocaine is
counted as a death "caused" by oxycodone and as a death "caused" by cocaine.
State officials say these cases are more accurately described as
multiple-drug deaths. The drugs may have been present in such lethal levels
that each could have killed the person by itself. Or the victim may have
taken smaller amounts of different drugs that combined in a deadly cocktail.
In fact, the autopsies in these cases usually list the cause of death as
poly-drug intoxication, multiple-drug toxicity or combined-drug toxicity.
The Sentinel series failed to report that most overdoses involving oxycodone
fell into this category, although that information was readily available in
autopsy reports. The series also failed to adequately explain the state's
The state's annual reports counted a total of 573 deaths caused by oxycodone
in 2001 and 2002, which was the basis for the number used in the Sentinel's
series. In its reanalysis, the Sentinel examined records for 535 of those
overdose deaths. Complete documentation wasn't available for the others.
The causes of death in the 535 cases were verified by the autopsies or by
the medical examiners' offices that conducted them. The results showed 396
of the cases were actually multiple-drug deaths. The remaining 139 were
caused solely by oxycodone.
Of the 535 overdose victims, 202, or about 38 percent, had illicit drugs in
their systems. The vast majority of the 333 other cases had commonly abused
but legal drugs present -- most often alcohol, prescription painkillers or
anti-anxiety medications -- in addition to oxycodone.
Autopsy reports do not identify the presence of OxyContin specifically
because blood tests only indicate oxycodone, which is found in more than 40
painkillers. In the newspaper's series last year, however, additional
reporting linked 205 deaths during 2001 and 2002 directly to OxyContin.
A re-examination of 195 of those cases -- complete data were not available
for the others -- found that roughly two in three were actually
multiple-drug deaths. Of the 195 deaths, 134 were attributed to at least two
drugs; 61 solely to oxycodone.
In its criticism of the Sentinel's series, Purdue Pharma cited its own study
of 1,014 deaths in 23 states in which oxycodone was found in the body. The
findings were published in 2003 in the Journal of Analytical Toxicology
after peer review.
Researchers determined that 919 of the deaths were overdoses caused by drug
abuse. The others included people who didn't necessarily die from oxycodone,
though it was detected in their systems during autopsies. Dr. J. David
Haddox, Purdue Pharma's vice president of health policy, said they included
people with terminal illnesses, victims of car accidents, four murder
victims and others.
When focusing on the drug abusers, Purdue concluded that 889, or 97 percent,
were multiple-drug deaths
Researchers, using criteria for overdoses set by the federal government,
deemed any case "in which there was at least one other plausible
contributory drug" to be a multiple-drug death. Only drugs that would be
unlikely contributors -- such as antibiotics, birth-control pills and
acetaminophen -- did not trigger the multidrug classification.
That approach is questioned by some experts.
"It's not necessarily valid for them to say that all of these other drugs
contributed to the cause of death," said Dr. Jeffrey Bernstein, a toxicology
professor at the University of Miami School of Medicine and director of the
Florida Poison Information Center. "Maybe they did [contribute], and maybe
they didn't. But you can't conclude that from the simple fact that they're
Florida's medical examiners do not use the same approach that Purdue did
when classifying multiple-drug deaths. Not every case with more than one
drug is ruled a multidrug fatality. Instead, medical examiners consider the
amounts of the drugs, their synergistic effects, the deceased person's
medical history and other factors, Nelson said. There are no uniform
"Like all of medicine, it's very subjective based on the medical examiner's
training and experience," he said.
More information about the Ibogaine