MEDICATIONS KILL 100,000 ANNUALLY, STUDY SAYS

Health: Adverse reactions to prescribed drugs are found to be far more
common than previously thought. But some question research methods.

Properly prescribed medications may kill more than 100,000 people a year,
taking more lives than diabetes or pneumonia, according to a new analysis
that suggests prescription medications cause more harm than previously
believed.

The study, appearing today in the Journal of the American Medical Assn.,
estimates that 76,000 to 137,000 people died in 1994 from such treatments.
That would make so-called adverse drug reactions between the sixth and
fourth leading cause of death in the United States.

Moreover, of the 33 million hospitalized patients in 1994, some 2.2 million
had a nonfatal reaction serious enough to require medical attention, the
researchers say.

Although some experts questioned the study's methods, the new estimates put
the problem in the most dramatic light yet.

The study ''puts into clear perspective that adverse drug reactions are a
major form of death and injury that can be prevented,'' said Dr. Sidney
Wolfe, director of the Public Citizen Health Research Group. He said the
injuries and deaths detailed in the study are nearly twice as high as
estimates recently done by his consumer group.

The findings should not encourage people to abandon vital medications, said
the study's leader, Dr. Bruce Pomeranz of the University of Toronto, who
said he was surprised by the death toll. ''What's needed is more awareness
of the potential problems with taking some drugs. Before you take a
medication you should know about its risk-benefit ratio,'' he said.

In addition, Pomeranz and other researchers say that the new findings
should spur hospitals and doctors to monitor patients more closely to
reduce potential toxic drug reactions.

The Pharmaceutical Research and Manufacturers of America, an drug industry
group, cautioned patients not to panic. ''Prescription medicines are safe
and effective when used properly,'' president Alan Holmer said in a
statement.

The analysis did not specify which drugs were most risky or what diseases
patients had when the adverse reactions occurred. But other studies have
found that the drug types causing the most serious medical problems in
hospitalized patients are painkillersfrom narcotics that halt breathing to
aspirin pills that induce stomach bleeding; antibiotics and antiviral
drugs, which can cause severe diarrhea, and cardiovascular and
anticoagulant drugs, which can cause a range of problems including internal
bleeding.

In addition, some people are allergic or sensitive to particular drugs,
while other bad reactions happen when two or more drugs are combined. For
instance, the widely prescribed allergy drug terfenidine, or Seldane, was
withdrawn from the market this year after reports that it triggered heart
rhythm problems in people also taking the commonly administered antibiotic
erythromycin.

Adverse drug reactions especially trouble the elderly, experts say, because
they are more likely to have multiple underlying health problems and also
tend to have a weakened liver and kidneys, the organs that break down and
eliminate medications.

Medical researchers believe that adverse drug reactions have been
underestimated for years. That is chiefly because hospitals and physicians
seldom report such reactions, dismiss them as unavoidable or mistake them
for disease symptoms.

The U.S. Health Care Financing Administration last fall proposed new
federal regulations requiring hospitals to step up drug-reaction
monitoring. But some physicians have criticized the proposed monitoring
system, which would involve periodic reviews of patient charts, as
intrusive and costly.

In their study, the Toronto researchers pooled and analyzed data from 39
U.S. studies on adverse drug reactions published between 1964 and 1996.
They looked at two groups: Patients who underwent an adverse drug reaction
while in the hospital that was at least serious enough to prolong their
stay, and also outpatients who had a drug reaction bad enough to
hospitalize them.

While other studies have looked at those two groups separately, this was
the first to combine them, leading to the ''extremely high'' prevalence of
drug reactions, as the researchers called it.

Between the upper and lower fatality estimates is the midpoint of 106,000
drug-induced deaths. That means that in 1994 - which the researchers chose
as a representative year - 0.32% of patients on a prescription drug, or
three out of every thousand, had a fatal reaction.

Their approach was ''conservative,'' the researchers said, in that they
focused only on correctly prescribed drugs. Their analysis did not consider
other sources of prescription drug problems, such as patient compliance
errors, intentional overdoses, narcotic abuse and accidental poisonings.

Nor did the researchers consider the presumably large number of people with
bad drug reactions who were not subsequently hospitalized.

''The truth is we missed a lot of people,'' Pomeranz says, including those
who ''died at home.''

Still, other researchers questioned aspects of the study because it is a
''meta-analysis,'' which involves statistically analyzing data pooled from
other studies, rather than studying real people. It is often difficult to
establish that a very sick person died from a drug reaction rather than an
underlying illness, said Dr. John Burke, a medical epidemiologist at LDS
Hospital in Salt Lake City, who has studied adverse drug reactions.

He urged caution in accepting the study as fact, but also praised it as a
''spur to action'' in attacking the problem of adverse drug reactions. At
LDS Hospital, which has one of the nation's few computer systems for
linking drug reactions and patient records, doctors have reduced adverse
events by 50%.

Dr. David Bates of the Brigham and Women's Hospital in Boston, who has
studied hospital reporting of adverse drug reactions, said ''even if the
true incidence of [adverse drug reactions] is somewhat lower'' than the
Toronto researchers say ''. . . it is still high, and much higher than
generally recognized.''

Wolfe, co-author of the book ''Worst Pills, Best Pills,'' said the analysis
underscored that the United States is ''an over-medicated society.''
Patients being prescribed a drug should ask if it is absolutely necessary,
he said, and should also be sure to tell doctors of any other drugs they
may already be taking.