FDA efficacy requirements may do more medical harm than good.BY DANIEL B. KLEIN, Santa Clara University and ALEXANDER TABARROK, George Mason University
than on-label. Thalidomide, for example, is an on-label treatment
from being sold in the United States unless
for leprosy but is used much more often in the treatment of some
forms of cancer and aspects of AIDS. In fact, most cancer and
AIDS patients are given drugs that are not fda-certified for the
drug’s safety and efficacy in a specified
prescribed use. In a large number of fields, a majority of patients
Fuse. The specified use is called the “on- are prescribed at least one drug off-label, and in some cases the
label” use and is what the fda officially
off-label prescription is the “gold-standard” treatment.
recognizes as the drug’s intended purpose.
Physicians prescribe off-label because medical knowledge
It often happens that, after the drug is permitted, physicians
advances at a faster rate than the fda. The U.S. Pharmacopeia Drug
and researchers discover that it has other uses. Physicians are
Information is a standard reference work that uses expert com-
perfectly free to prescribe a permitted drug for such “off-label”
mittees to compile and evaluate the dosing, indications, inter-
uses, and the fda plays no role in certifying those uses.
actions, pharmacology/pharmacokinetics, and side/adverse
Viagra, for example, was initially intended to treat angina, but
effects of drugs for both labeled and off-label uses. Researcher J.
when older men reported its unusual side-effect it became a
Howard Beales found that off-label uses that later came to be rec-
blockbuster treatment for erectile dysfunction. Since being
ognized by the fda appeared in the Pharmacopoeia on average 2.5
approved for that use, Viagra has also been found to be useful in
years before fda recognition. For drugs that are off-patent, many
the treatment of pulmonary hypertension. Strange as it may
uses that appear in the Pharmacopoeia will never receive official
seem, premature babies have been prescribed Viagra to help
them breathe, with apparently good success. Viagra’s off-label
Off-label prescribing also gives physicians and patients
uses have not been proven effective to the same degree as its on-
more options when standard treatments fail. Patients are het-
label uses. But the safety profile is good, and because doctors have
erogeneous, and what works for one person may not work for
few other treatment options for pulmonary hypertension and
another. If physicians were limited to labeled uses, they would
small trials suggest Viagra’s effectiveness, doctors are taking what
in many cases have no therapies to employ at all. Off-label pre-
appears to be the best course of action for their patients.
scribing is more common when standard treatment regimesdo not exist or fail.
O F F - L A B E L I S P E R V A S I V E A N D V I T A L
Finally, when all has failed — as it often does — it is not irra-
Off-label prescribing is very common in all areas of medicine, and
tional for patients to demand experimental treatments. Off-
it is not unusual for a drug to be prescribed off-label more often
label prescribing has few costs in this context and, in additionto providing hope, it may generate important new knowledge.
Daniel B. Klein is a professor of economics at Santa Clara University and director of
Santa Clara’s Civil Society Institute. He is also a scholar of the Ratio Institute of
Stockholm, Sweden. Klein can be contacted by e-mail at email@example.com.
H O W D O E S T H E P R A C T I C E O F O F F - L A B E L
Alexander Tabarrok is a professor of economics at George Mason University and
director of research for the Independent Institute in Oakland, Calif. He can be contacted
The research that goes into approving a drug typically provides
good information about how the drug works, at least in the test
Klein and Tabarrok are co-authors of FDAReview.org, an extensive Web site on FDA
population. Challenged by diseases without good treatments
or by patients for whom standard therapies have failed,
The off-label experience testifies to the fact that much knowl-
researchers and physicians may try new approaches that have
edge about efficacy and safety is produced outside the fda reg-
ulatory apparatus. The Pharmacopoeia’s recognition of off-label
Viagra, for example, is known to help dilate smooth-muscle
indications years ahead of the fda demonstrates that physicians
blood vessels by boosting the production of the signaling com-
and scientists have certified thousands of drug indications quite
pound nitric oxide. Pulmonary hypertension is a rare blood ves-
independently of the fda, even when those indications are not
sel disorder in which the lungs cannot distribute adequate oxy-
very closely related to the original indications. In addition to the
gen to the rest of the body. Few treatments exist for babies born
Pharmacopoeia, there are several other forms of professional cer-
with this condition but, knowing that lack of nitric oxide was one
tification, including the American Hospital Formulary Service DrugInformation, hmoformularies, and a wide
tion of nitric oxide a standard treatment.
array of specialist professional periodi-
sive and impractical for long-term use.
The drug-approval process excludes several
gra in the hope that it may provide a bet-
ter, longer-term solution. In a news story
about a case involving two babies, Dr.
tion” broadly. When a reputable medical
scribed Viagra: “Both babies were very ill
he is not issuing a formal certification the
might help people in serious need. “Cer-
tification” is as subtle and particularistic
as medicine itself. Indeed, one might see
the entire enterprise of medicine and the
health sciences as a system of certifica-
trials for those uses as on-label uses. This criticism misunder-
tions. In the final analysis, it is that system of certifications, not
stands the nature of much drug discovery. Several small-scale tri-
fda approval per se, that doctors actually rely on.
als for Viagra in infants have already been completed and largertrials are underway. The only reason the trials are occurring is
DOCTORS’ COMMENTS In a recent study, we engaged nearly 500
because physicians and patients with few other options tried Via-
doctors in an online survey about these matters. Some of their
gra and were encouraged by the results. Clinical practice often
remarks about the practice of off-label prescribing were espe-
precedes, rather than follows, clinical trials.
Each newly permitted drug projects a wide range of theo-
“Often, efficacy information is already available from
retically related and possibly effective off-label indications, and
the promise of each gradually diminishes the further (in termsof current medico-pharmacological understanding) such
“There is often data from Europe or in peer review jour-
prospective indications are from the on-label indications. Med-
nals. fda efficacy trials are important, but they are not
ical science explores possible related uses and, if they appear
the only measure (except legally in terms of company
marketing) of a product’s efficacy for a certain condition.”
“Off label use is very often based on valid smaller stud-
C E R T I F I C A T I O N W I T H O U T T H E F D A
ies concerning other than the index medical condition;
Physicians learn of off-label uses from medical research and expe-
those studies may not be large enough or the pharma-
rience conveyed by peer-reviewed publications, newsletters, lec-
ceutical company may not want to spend the [money] it
ture presentations, conferences, and conversations with trusted
colleagues. The new learning comes from many sources: uti-lization and outcome reviews, clinical and epidemiological stud-
“fda approval on efficacy lags behind peer-reviewed
ies, new theories advanced by scientists, new judgments made by
professional and scientific bodies, and new results reported by
“Almost all cancer chemotherapy is off-label. There is
pharmaceutical companies. As the enterprise of medical science
no way two or three drug companies can expend the
proceeds, the new learning flows back and forth between med-
effort to get a combination regimen approved.
ical researchers and practitioners, albeit in fits and starts.
Oncologists use the peer-reviewed literature to decide
off-label without the fda, so why do they not have more con-
therapy. Almost always decisions are based on random-
fidence in enjoying the wider range of new drugs that would
come forth if we dropped efficacy requirements on initialuses? Indeed, there seems to be a logical inconsistency in
“Plaquenil was developed and fda-approved as a malari-
favoring off-label prescribing but requiring proof of efficacy
al drug. Later it was found to relieve rheumatoid arthritis
for the drug’s initial use. Physicians are against fda-required
symptoms in the patients taking it for malaria. Studies
efficacy tests on new uses of old drugs, but many favor such
show that it worked and was efficacious but should we
tests on new uses of new drugs. Why the difference?
wait for the fda to prolong the relief of pain and suffer-
In our online survey, after asking the respondents’ opinion
ing for several years while the necessary drug compa-
on the two policy reform questions — adding efficacy require-
ny/fda studies are done or just use common sense?
ments for subsequent uses, and dropping efficacy requirements
Often there is no financial incentive for a drug company
on initial uses — we “challenged” those who were against
to pursue off-label indications for conditions that would-
restrictions on off-label prescribing but in favor of restrictions
n’t generate sufficient income to offset the cost of fda
on initial prescribing. Were their choices consistent?
approved trials. But university-based, double-blind, high-
Their responses fell into a number of basic arguments. Pon-
ly powered studies show benefits that outweigh risks.”
dering the responses better taught us that there are significant dif-ferences between on- and off-label uses, but on the whole we felt
O F F - L A B E L A S A N A T U R A L E X P E R I M E N T
that none of the responses successfully deflected the basic force
of the consistency argument for liberalization. Below, we briefly
In his 2000 article “Assessing the fda via the Anomaly of Off-
state the four central responses and provide our own brief rebut-
Patent Drug Prescribing,” Alexander Tabarrok argued that off-
tal. (Our 2004 paper takes up these issues in greater depth.)
label usage provides a “natural experiment.” In a sense, off-label
“Off-label uses are often related to the on-label
uses are regulated according to the pre-1962 rules, under which
use.” Many off-label uses are related to the on-label use,
the fda held new drugs only to safety requirements, whereas
but “related” does not guarantee effectiveness. The off-
on-label uses are regulated according to the post-1962 rules.
label uses actually practiced are a fraction of all possible
Thus, the same medical institutions — in the same country at
related uses, so we see medicine working to discover
the same time — are operating under dual systems of drug reg-
effective therapies and to weed out ineffective ones,
ulation. Off-label prescribing tells us something about how the
regardless of how related they may be. Moreover, many
world might look if the fda were restricted to safety-testing
off-label uses are not related to the on-label use. It is sur-
alone. If we evaluate off-label prescribing positively, then this
prising that Viagra works for pulmonary hypertension
provides some support for fda reform; a negative evaluation
as well as erectile dysfunction (surprising but not unex-
supports the current system and suggests fda control of off-
plainable). Finally, some uses come to be understood as
label prescribing might be in order.
related only because of demonstrated results of off-label
One piece of evidence in favor of off-label prescribing is the
usage. After all, “relatedness” is a function of the current
opinion of physicians. Do physicians clamor for the assurance
state of pharmo-medical understanding.
of the fda when it comes to off-label prescriptions? Or do theyregard as valuable the freedom to prescribe drugs that have not
“Efficacy requirements on initial uses are good, but
been fda-certified for the prescribed use?
because of patent life expiration it would be crazy
In an online survey, we asked nearly 500 doctors whether
to impose them on subsequent uses.” It is correct
the fda should hold off-label uses to proof-of-efficacy require-
that dwindling patent life helps explain why it would be
ments; the doctors responded with a resounding no. Fully 94
harmful to impose new requirements on subsequent
percent opposed the requirements, and many wrote strongly
uses. But this argument does not justify the initial effi-
worded objections that described the requirements as “clear-
cacy requirements. This response explains why the
ly naïve,” “stupid and unethical,” “dangerous,” “disastrous,” and
respondent rejects the consistently restrictive position,
claimed “medicine would grind to a halt.”
not why he rejects the consistently liberal position. “Initial efficacy requirements are desirable because they enhance knowledge of safety.” Efficacy require-
adding restrictions to off-label uses, but do they favor dropping
ments do enhance knowledge of safety, but if stronger
efficacy requirements on initial use? In our survey, we asked
safety testing is the goal, it ought to be pursued and jus-
doctors that question and 27 percent said yes, fda efficacy
standards should be made voluntary. The anti-liberalizers,however, outnumbered the liberalizers by slightly more than
“Without efficacy requirements, new useless drugs
two to one (58 percent said no, 15 percent were not sure). Most
would flood the market.” That might be true, but to
doctors supported the status quo, namely efficacy require-
what extent—for how long—would physicians prescribe
ments on initial uses but not subsequent uses.
and patients take inferior drugs instead of the drugs that
Support for the status quo among doctors is in one way
would really help them? The medical marketplace is far
curious. As explained, doctors get on quite well prescribing
from perfect, but patients and physicians are willing to
pay more for medicines that work and this demand pro-
Eliminate false steps and you eliminate progress. False steps can
vides an incentive to produce quality assurance.
be made less costly, however, if they are corrected quickly. Thus,it is valuable to compare error correction in the two cases of
E R R O R I S I N E R A D I C A B L E , B U T . . .
allowing a bad drug and withholding a good drug. In the first
Doctors who disagree with liberalization remind us of valid
case, the error can be corrected with experience, but it is much
points. For decades, women were prescribed hormone replace-
more difficult to gain experience about a non-approved good
ment therapy. But we now know from the Women’s Health Ini-
drug and impossible to do so if the drug is never researched and
tiative studies that, contrary to what was expected, such ther-
developed. Error correction tends to be rapid in decentralized
apy resulted in increased risk of heart attacks, strokes, blood
systems and rather slow in centralized, bureaucratic systems.
clots, breast cancer, and dementia for which the reduced risk
A more error-prone system is not necessarily worse than a less
of fractures and colorectal cancer do not appear justifiable.
error-prone system, if the errors are corrected more quickly in the
Hormone replacement therapy was largely prescribed off-label
former than in the latter. That point is familiar from the debate
(the on-label use is short-term prevention of hot flashes, sleep
over central planning. In capitalist systems, entrepreneurs make
difficulties, and other problems of menopause). It is clear, there-
a lot of mistakes, but error correction is swift. In centrally planned
fore, that off-label use can be harmful.
systems, error correction is very poor. It is now widely recognized
Of course, on-label use can also be harmful. Perhaps even
that progress is more rapid in the capitalist systems. The same may
more harmful than on- or off-label usage is the non-usage or
be true with respect to the regulation of pharmaceuticals.
suppression of would-be beneficial drugs because of excessivefda regulation and costs. Sam Peltzman and Steven Wiggins
have each estimated that increased fda regulation in the 1960s
Off-label usage provides a window onto how a less-regulat-
and 1970s reduced the number of new drugs by 60 percent.
ed drug certification system would operate. An extensive but
Even a few suppressed or long-delayed drugs could account for
little-studied system of drug certification exists outside the
control of the fda. The experience with off-label prescrib-
Rather than seeking a tally of mortality and morbidity, we
ing and the experience of pre-1962 America suggest that ini-
can perhaps understand some of those costs better by recog-
tial efficacy requirements may do more harm than good.
nizing that progress depends on discovery and discovery on
Dropping efficacy requirements is a proposal that should be
experimentation. Experimentation always involves false steps. R E A D I N G S
• “Assessing the FDA via the Anomaly of Off-Label Drug
• “Matching Prescription Drugs and Consumers,” by A. Masson
Prescribing,” by Alexander Tabarrok. The Independent Review, Vol.
and P. H. Rubin. New England Journal of Medicine, Vol. 313 (1985).
• “New Uses for Old Drugs,” by J. H. Beales III. In Competitive
“The Demand for and Supply of Assurance,” by Daniel B. Klein. Strategies in the Pharmaceutical Industry, edited by Robert B. Helms;
In Market Failure or Success: The New Debate, edited by T. Cowen and E.
Washington, D.C.: American Enterprise Institute, 1996.
Crampton. Cheltenham, UK: Edward Elgar, 2002. • “Do Off-Label Drug Practices Argue Against FDA Efficacy
• “Off-Label Drug Prescription: Filling the Regulatory Vacuum,”
Requirements? Testing an Argument by Structured
by William L. Christopher. Food and Drug Law Journal, Vol. 48
Conversations with Experts,” by Daniel B. Klein and Alexander
• “Off-Label Drug Use in Human Immunodeficiency Virus
• “Drug Discontinuations in the United Kingdom and the
Disease,” by C. L. Brosgart, T. Mitchell, E. Charlebois, R. Coleman,
United States, 1964 to 1983: Issues of Safety,” by O. M. Bakke, W.
S. Mehalkol, J. Young, and D. I. Abrams. Journal of Acquired Immune
M. Wardell, and L. Lasagna. Clinical Pharmacology and Therapeutics,
Deficiency Syndromes and Human Retrovirol, Vol.12, No. 1 (1996).
Vol. 35 (1984). • “Drug Safety Discontinuations in the United Kingdom, the
• “Persuasion or Information? The Economics of Prescription
United States, and Spain from 1974 through 1993: A Regulatory
Drugs Advertising,” by Keith B. Leffler. Journal of Law and
Perspective,” by O. M. Bakke, M. Manocchia, F. de Abajom, K. Economics, Vol. 24, No. 1 (1981).
Kaitin, and L. Lasagna. Clinical Pharmacology and Therapeutics, Vol.
• “The Price of Innovation: New Estimates of Drug
Development Costs,” by J. A. DiMasi, R. W. Hansen, and H. G.
• “An Evaluation of Consumer Protection Legislation: The 1962
Grabowski. Journal of Health Economics, 2002.
Drug Amendments,” by Sam Peltzman. Journal of Political Economy,Vol. 81, No. 5 (1973).
• “Product Quality Regulation and New Drug Introductions:
Some New Evidence from the 1970s,” by Steven N. Wiggins. www.FDAReview.org, by Daniel B. Klein and Alexander
Tabarrok. Oakland, Calif.: The Independent Institute, 2002. Review of Economics and Statistics, Vol. 63 (1981).
• “Limiting Physician Freedom to Prescribe a Drug for Any
• “Regulating the Prescribing of Human Drugs for Nonapproved
Purpose: The Need for FDA Regulation,” by S. A. Shapiro.
Uses under the Food, Drug, and Cosmetic Act,” by David Kessler. Northwestern University Law Review, Vol. 73, No. 5 (1979). Harvard Journal on Legislation, Vol. 15 (1978).
CURRICULUM VITAE August 2007 NAME: DATE OF BIRTH : June 17, 1941 BIRTHPLACE: PRESENT BUSINESS ADDRESS: The Center for Diabetes & Endocrine Care 1150 North 35th Avenue Suite 590 Hollywood, Florida 33021 (954)-963-7100 PREVIOUS BUSINESS ADDRESS: 3700 Washington Street Suite 305 Hollywood, Florida 33021 (954)-963-7100 PRE-PROFESSIONAL EDUCATIO