|
Alfred A. Bove, MD, PhD
I noted in the national news in August
that a group of consumers is filing a class-action suit against
the maker of Claritin® because the public advertising
promised effects that were not substantiated by clinical trials.
Television advertising in particular
requires rapt attention
to understand what is being said and to realize what is not
being said.
The language in such commercials is carefully
designed to avoid outright mistruth, but most public viewers
are not aware of the need to carefully
dissect every word to determine what is being stated.
In the meantime, August also witnessed
the collective action of 11 medical journal editors to avoid
the same problem
in publication of clinical trial results in their journals.
The journal editors concluded that many
published clinical trials are presented in a way that will
cast a favorable light on the therapy being tested, and negative
data on the therapy may be withheld.
We rely on the results of drug trials
to determine the best method of treatment for patients with
cardiac diseases. Standards of care are often based on the
results of clinical trials; when a treatment guideline can
be supported by evidence from trials, the guideline is considered
to be a strong statement about the standard of care.
Our system of determining standards of
care, however, leaves a considerable amount of discretion
to the individual investigators who conduct the trials. In
many cases, the trial is designed by the sponsoring corporation,
and the investigator's role is to recruit patients, assist
in interpreting the results, and publishing them.
Many of today's trials are studies of
drug or device efficacy and are sponsored
wholly or mainly by the company that manufactures the drug
or the device. With this system, both the investigator
and the drug company have a strong incentive to reach positive
conclusions about the product from the clinical data.
The investigator does not benefit from
negative results because of the usual practice of peer review,
which gives clinical trials with negative results lower priority
for publication.
An investigator conducting a clinical
trial for a drug company also develops a better reputation
with the company if the clinical trial outcome supports the
goals of the drug or device company.
Clearly, a company is not interested
in negative results after it has spent millions of dollars
to support a clinical trial that is the culmination of a development
program that may have cost as much as $500 million to bring
the drug or the device to market.
Incentives to reach positive conclusions
in clinical trials encompass studies from all sources of support.
Federal grants for clinical trials are considered better spent
when a positive result is obtained, and investigators who
have difficulty publishing trial results are likely to have
trouble obtaining funding on federal grants.
Negative results may be blamed on poor
study design. The peer review process for grants depends heavily
on the peer review process for publication because the journal
publication record of an investigator is an important criterion
for approval of funding.
These incentives to succeed in research
can produce investigator
bias that encourages positive conclusions
by biasing interpretation of results in a favorable light
and by suppressing negative results that are not likely to
be published and are not likely to add prestige to the investigator's
reputation.
Despite these systematic problems with
conducting clinical research, there is no other system that
works as well. And, so, we return to the first paragraph of
this editorial and must act as consumers of medical information.
We
must read, look, and listen carefully to the investigators,
the companies, and the publications to be sure we understand
what is being said and what is not being said.
We must listen for over-enthusiasm by
the investigator and for marketing ploys by the companies,
and we must look for flawed data in studies that may appear
in well-respected journals.
The journal editors' efforts to improve
clinical trials reporting are to be applauded. There is a
need to maintain credibility, both for the journals and for
the authors. Peer review must be vigilant, and it must not
succumb to popular notions or industry campaigns.
We should teach our students and residents
as well as ourselves how to examine the message of a clinical
trial and understand what information is being provided and
to which patients it applies.
Because our practice guidelines are based
on clinical trials and the investigators who participate in
the trials are often those invited to create practice guidelines,
we run the risk of creating a credibility gap even in the
development of our guidelines.
It doesn't require much time watching
television to understand how language and imagery are manipulated
to provide strong but inaccurate messages about the commercial
products being displayed.
The incredibility of commercials is widely
accepted, to the point where the message is disregarded and
the entire exercise becomes only humorous entertainment. With
too much enthusiasm, disregard for data, and manipulation
of outcomes, we risk arriving at the same level of credibility
in clinical trial reporting as that expressed in a TV commercial.
I hope the efforts to improve clinical
trial reporting will be the beginning of a move toward more
balanced reporting of results so that we can tell a patient
that the therapy being advised is truly supported by well-conducted
clinical trials involving thousands of patients with the same
condition and that the outcome of their therapy should be
the same outcome as that described in the relevant clinical
trial.
American
College of Cardiology September 2001
| 