Transcript “Who stands behind the word?”

The pharmaceutical
industry’s influence on
medical publishing
Gavin Yamey MD
Deputy editor, wjm
(www.ewjm.com)
Assistant editor, BMJ (bmj.com)
A question
You read a guideline by the AHA
giving a “class I” recommendation for
using tPA in acute stroke
You then find out that most authors
had a financial relationship with
Genentech, who make tPA
Now what do you do?
A question (2)
You read an extremely convincing
RCT showing the huge benefit of
flutamide in treating advanced
pancreatic cancer
You then realize that the drug
manufacturer did the randomization,
gave statistical assistance, and did the
data analysis
Now what do you do?
A question (3)
You’re having a busy day, but you
agreed to see a drug rep who wants to
persuade you to use Neurontin for
conditions other than epilepsy
The rep offers you Neurontin pens,
pads, and a clock. She then says she’ll
pay for you to go to a symposium.
What do you do?
What I’d like to talk about
Industry influence over clinical trials
(funding, design, analysis, publication)
Reviews
Guidelines
Ghostwriting
Dissemination/promotion to physicians
Some solutions
Industry influence: a summary
 Industry designs and funds studies
likely to favor its products
 Companies analyze the data
 Ghostwriters are paid by industry to
add favorable spin
 If the data are not favorable, industry
suppresses or delays publication
$16 billion on drug promotion in 2000
Does this influence matter?
YES: Evidence shows that it
affects the outcome of clinical
trials, the conclusions of reviews,
and the prescribing practices of
physicians
Industry influence over
clinical trials
Funding
Design
Data analysis
Publication
Industry funding of trials
Analysis of 107 controlled trials
Did authors favor new or old drug?
Did authors have industry support or
not?
Trials funded by manufacturer of new
drug were significantly more likely to
favor new drug
J Gen Intern Med 1986;1:155-8
Industry funding of trials (2)
Only 5% of industry-sponsored
studies of cancer drugs reached
unfavorable conclusions about the
company's drugs; figure was 38%
in studies with nonprofit funding
JAMA 1999;282:1453-1457
Industry funding of trials (3)
Authors whose work supported
the safety of calcium-channel
blockers were more likely to be
funded by the drugs'
manufacturers than authors
whose work did not support the
safety of these medications
N Engl J Med 1998;338:101-106
Industry funding of trials (4)
At the end of 1998, 3 major
studies without industry support
found a higher risk of venous
thrombosis for 3rd generation
contraceptives.
Three sponsored studies did not.
BMJ 2000;320:381
Industry influence on trial
design (1)
Testing drug in healthy population,
rather than population that will receive
it (drug appears to have fewer side
effects)
Only 2.1% of subjects in trials of
NSAIDs were 65yrs+, even though these
drugs are more often used, and have a
higher incidence of SEs, in the elderly
CMAJ 1998;159:1373-1374
Industry influence over trial
design (2)
Using surrogate end points, and
only publishing favorable ones
Int J Technol Assess Health Care 1996;12:209-237
Industry influence over trial
design (3)
Testing new drug against insufficient
dose of old drug (new drug appears
more efficacious)
In one study, trials of NSAIDs always
found the sponsor’s drug to be equal or
better BUT in 48% of trials, the dose of
the sponsor’s drug was higher
Arch Intern Med 1994;154:157-163
Industry control over data
analysis
 Single author RCT of flutamide in
advanced pancreatic cancer
 Unexpectedly favorable, dramatic result
 BUT: randomization, statistical
assistance, and data analysis were all
done by industry
 Study was heavily criticized
BMJ 1998;316:1935-1938
Industry control over trial
publication
1987: Manufacturer of levothyroxine
(Synthroid) contracted with UC
researcher to compare it with existing
thyroid preparations
1990: No more effective
Sponsor refused to allow findings to be
published
JAMA 1997:277:1238-1243
Industry control over trial
publication (2)
The Immune Response Corporation
contracted with UCSF to perform an
RCT of an immune modulator to treat
AIDS
No effect
Company tried to suppress publication
West J Med 2001;175:225-226
Industry control over trial
publication (3)
1996: Olivieri and colleagues found that
deferiprone (used to treat thal major)
could worsen hepatic fibrosis
Apotex (trial sponsor) threatened legal
action if authors published the data
CMAJ 1998;159:955-957
Industry influence over reviews
106 reviews of passive smoking: does it
cause harm? 63% concluded harmful,
37% harmless
Multiple regression analysis controlling
for article quality, peer review status,
article topic, yr of publication: the only
factor associated with the conclusion
was whether the author was affiliated
to tobacco industry
JAMA 1998;279:1566-70
Industry influence over
guidelines
Survey of 192 authors of 44 clinical
practice guidelines: 87% of authors had
some form of interaction with the
pharmaceutical industry
BUT in published versions of the
guidelines, specific declarations about
the personal financial interactions of
authors with industry were made in only
2 cases
JAMA 2002;287:612-7
Guidelines: a cautionary tale
The AHA rated the thrombolytic agent
alteplase (tPA) as a class I (definitely
recommended) intervention for stroke
despite controversy about its safety and
efficacy
Why did it make this recommendation?
Guidelines: a cautionary tale (2)
Most of the AHA's stroke experts
had undisclosed ties to Genentech,
the manufacturers of alteplase
Genentech contributed over $11m to
the AHA in the decade before the
AHA recommendation on alteplase
Guidelines: a cautionary tale (3)
Following public scrutiny, the AHA
has been forced to withdraw
statements that alteplase for stroke
"saves lives"
Even a seemingly impartial non-profit
organization that issues professional
guidelines may have ties to the
manufacturers of recommended
interventions
Ghostwriting (the ghost-guest
syndrome)
 A professional medical writer ("the
ghost"): employed by industry, paid to
write, not named as author [“the session
musicians of the biomedical literature
world”]
A prestigious author ("the guest"): does
not analyze data, does not write the
manuscript, may or may not review the
manuscript
A rarity?
No
 In one study, 19% of original articles
surveyed had named authors who would
not meet ICMJE criteria for authorship
(JAMA 1998;280:222-224)
 11% had ghostwriters, who were not
named as authors
 Reviews, editorials, clinical guidelines:
higher prevalence?
The harms of ghostwriting
 Publishing articles that are over-
zealous about a product could
distort physician prescribing
Inaccurate or misleading
information conveyed to physicians
 Ghostwriting distorts the scientific
record
The Neurontin story (NY
Times, 15 May 2002)
Neurontin: FDA-approved for a narrow
use: seizure control in patients already on
one drug
No good evidence for any other indication
Whistleblower case: Warner-Lambert
hired 2 marketing firms to ghostwrite
articles claiming the benefits of
Neurontin for unapproved uses.
Company paid physicians $1000 to act as
“guest” authors
Ghostwriters speak out
"I agreed to do two reviews for a
supplement to appear under the names
of respected ‘authors.’ I was given an
outline, references, and a list of drugcompany approved phrases. I was asked
to sign an agreement stating that I would
not disclose anything about the project.
I was pressured to rework my drafts to
position the product more favorably."
Ghostwriters speak out (2)
"I was told exactly what the drug
company expected and given
explicit instructions about what to
play up and what to play down“
Drug reps, promotional
materials, samples, gifts
Contrary to the beliefs of most heath
care providers--samples, gifts, food,
and discussions with drug reps exert
significant influence on provider
behavior
Promotional materials and
presentations are often biased
www.nofreelunch.org
Drug samples
Industry gave out $7.2 billion worth of
free samples in 2000
Pharmaceutical companies'
"generosity" to provide drug samples
has a specific purpose: to change
physician behavior to write more
prescriptions for their particular drug.
Drug samples (2)
Physicians are more likely to
prescribe a drug if a sample is
available, even if they do not think
the drug is the best one for the
patient
What happens when the sample runs
out?
J Gen Int Med. 2000;15: 478-483.
Gifts: free trips to symposia
In one study in one hospital,
accepting such a gift was associated
with an increase in prescribing of 2
drugs manufactured by sponsor of
symposium
Prescribing patterns were
significantly different from national
patterns
Chest 1992 Jul;102(1):270-3
Contact with drug reps
Study of prescribing habits of 124
physicians after the introduction of
temazepam
Contact with the drug rep
regarding temazepam was the most
consistent predictor of favorable
reception.
Soc Sci Med 1988;26(12):1183-9
Contact with drug reps (2)
Requests by physicians to add a
new drug to the hospital
formulary are strongly
associated with physicians’
interactions with reps
JAMA 1994;272:355
Promotional material
Study of 106 statements made by
drug reps during 13 presentations
11% of statements were false, yet
physicians rarely spotted them
JAMA 1995;273:1296-8
The Neurontin Story (2) NY
Times, May 15 2002
Whistleblower case: Physicians
allowed pharmaceutical reps into
their examining rooms to meet with
patients, review medical charts, and
recommend which medicines to
prescribe
Called a “shadowing program” and it
involved hundreds of patients
Are you “drug company
dependent”? (CAGE)
Have you ever prescribed Celebrex?
Do you get Annoyed by people who
complain about drug lunches and free
gifts?
Is there a medication loGo on the pen
you're using right now?
Do you drink your morning Eye-opener
out of a Lipitor coffee mug?
www.nofreelunch.org
The 4 steps to recovery
 1. We admit we are powerless over
pharmaceutical paraphernalia-that our lives
have become unmanageable.
 2. We make a searching and fearless moral
inventory of ourselves and our desks.
 3. We are entirely ready to remove all these
defects of character, as well as pens, penlights,
and notepads.
 4. Having had a spiritual awakening as the
result of these steps, we try to carry this
message to others and to practice these
principles in all our affairs.
Industry influence:
some solutions
Contracts between industry and
researchers allowing freedom to publish
trials
Industry funds trials, but has no role at
all in design, implementation, analysis,
publication
Public funding of trials that matter
Reviews and guidelines: should they
exclude sponsored authors or at least set
a maximum level of industry support?
Disclosure, disclosure,
disclosure
“Better to light a candle than to curse the
darkness”
Yet competing interests (e.g. industry
support) are still not being declared
Recent study of 89 authors (75 articles):
69 responded, 45 had financial conflicts of
interest
But only 2 of the 70 articles had declarations
N Engl J Med 1998;338:101-5
Industry influence: some
solutions (2)
 Codes of conduct on gifts/relationships to drug
reps e.g. AMA Council on Ethical & Judicial
Affairs:
 “Any gifts accepted by physicians individually
should primarily entail a benefit to patients
and should not be of substantial value”
 “Subsidies from industry should not be
accepted directly or indirectly to pay for the
costs of travel, lodging, or personal expenses
of physicians who are attending conferences or
meetings”
Conclusions
Industry influences medical publishing at
all stages (funding, trial design, data
analysis, publication, promotion of
findings, ghostwriting)
Influence goes largely undisclosed
Industry influence arguably distorts the
scientific record and distorts clinical
medicine
Disclosure is a “partial panacea” (we have
a better chance of deciding for ourselves)
A matter of life and death
"The integrity of a body of literature
is itself our society's ultimate
temporal forum for negotiating life
and death, suffering and
wellness.....the medical well-being of
the society it serves is dependent on
the question of who stands behind the
word."
(Mark Gruber, anthropologist)