Your independent, respected source for information about medications and natural therapies.

Medication side effects are the #4 leading cause of death in the U.S. annually (JAMA 1998). Yet, few people receive adequate information when medication is prescribed. This website is dedicated to providing information to help you and your doctor make informed, intelligent choices about medications and natural alternatives to maximize the benefits and minimize the risks of treatment. Note: This website is free of drug company or government influence. Jay S. Cohen M.D.

Dr. Jay S. Cohen, M.D.

Bextra: Designed by Pfizer to Fail

An Open Letter to Pfizer Inc. on the Underlying Reasons That Bextra Failed, on Why Pfizer Made Bextra Unnecessarily Risky — and on How Pfizer Might Still Save Celebrex.

Dear Pfizer:

You weren’t happy when the FDA ordered the withdrawal of Bextra (valdecoxib), one of your top-selling COX-2 anti-inflammatory drugs, on April 7, but you should have expected it.  In fact, you should have expected it from the first day you marketed Bextra in 2001.  Even then it was obvious that Bextra was a high-risk drug.  I told the FDA this in November 2002 when, as the keynote speaker on the FDA’s Annual Science Day, I spoke for an hour about problems in drug safety with Celebrex, Prozac and other antidepressants, Premarin, Prempro — and Bextra — all of which have been subsequently linked with new toxicities.

I warned about Bextra many other times: in debate with the FDA’s top drug expert, Dr. Robert Temple, at the large conference of the Drug Information Association in June 2003, and during another debate with Dr. Temple at the American Society for Clinical Pharmacology and Therapeutics in March 2004.  In November 2003, I warned about Bextra while presenting at the American College for the Advancement of Medicine.  In September 2004, I warned about Bextra while presenting grand rounds at the Scripps Clinic Hospital and Medical Center in La Jolla.

I also focused on Bextra when speaking to drug company officials, including Pfizer officials, at the DTC conference in Boston in May 2003.  So you cannot say that you were not forewarned.  You might have taken steps then to improve Bextra’s safety, but you didn’t, and now it is too late.

Bextra: Unnecessarily Strong, Unnecessarily Risky

Bextra’s withdrawal was avoidable.  Bextra could have been a better, safer drug, but it was undone by the way you developed and marketed it.  When you considered your options for marketing Bextra, you decided, as many drug companies decide today, to beat your competition by introducing Bextra at a very powerful 10-mg dose that would produce high efficacy numbers.  These efficacy numbers would compare favorably to competitors’ drugs, giving your advertising executives and your thousands of drug sales reps a tool to influence doctors to prescribe Bextra.

The strategy worked.  It usually does, which is why so many drugs today are introduced at such strong doses — and why so many end up with black-box warnings or withdrawn.  But this wasn’t your concern when you introduced and intensively marketed Bextra.  High sales were your concern.  Doctors believed the “So Much Power, So Much Relief” ads extolling Bextra .1  Doctors didn’t realize that “so much power” also meant extra risks.  Most side effects are dose-related: the stronger the doses, the greater the risks and the more severe the reactions.  Thus, from the beginning, it was apparent that “powerful” Bextra was a high-risk drug.

One-Size-Fix-All Drugs Lead to Overmedication and Toxicities

It was high-risk for another reason: for most uses, Bextra was one-size-fits-all.  Again, your advertising extolled this as a virtue: “One powerful 10-mg dose provides 24-hour power.1”  Yes, one-size-fits-all drugs make it easy for doctors to write prescriptions quickly and to move on to the next patient.  But one-size-fits-all dosing assumes that all patients are alike — which even you know is ridiculous.  As you know well, a basic principle of medical  science is “individual variation,” the reality that people respond very differently to medications.  According to the American Medical Association: “In normal subjects, the extent of interindividual variability ranges from 4- to 40-fold depending on the particular drug and population.2

This is why medical students are taught the importance of individualizing drug treatment.  Their standard textbook, Goth’s Medical Pharmacology, states: “Many adverse reactions arise from the failure to tailor the dosage of drugs to widely different individual needs3.”  That’s why drug companies that are truly concerned about patients’ safety market medications in several sizes, so that doctors can prescribe the proper dose for the individual patient.  Other anti-inflammatory drugs such as naproxen (Naprosyn, Aleve) and ibuprofen (Motrin) come in four sizes.  Most patients do fine with over-the-counter naproxen or ibuprofen or with the lowest prescription doses.  In contrast, Bextra came in only one size, and that size was equivalent to the maximum doses of naproxen and ibuprofen.

In other words, when people received Bextra, they got the equivalent of the strongest doses of naproxen or ibuprofen.  And they got this very powerful dosage of Bextra no matter their age, size, state of health, use of other medications, or whether their pain was mild or severe.  With Bextra, the same 10-mg dosage would be used for Shaquille O’Neil and a 90-pound, 90 year-old woman taking 9 other drugs.  Such methods defy medical science and common sense, but that’s exactly how you designed Bextra to be used.  Is it any surprise that Bextra caused lethal skin and cardiovascular problems?  The only surprise is that it didn’t cause more harm.

Maybe you just made an error in judgment in selecting the powerful 10-mg, one-size-fits-all dosage for Bextra.  But it wasn’t an error, was it, but a calculated decision favoring marketing concerns over medical science.  How do I know?  Because you knew that Bextra, like ibuprofen and naproxen, worked very well at much lower dosages.  You had placebo-controlled, double-blind studies proving it.  Three of these studies were published, and they showed that Bextra in doses of just 2.5 and 5 mg/day — 75% and 50% lower doses than 10-mg Bextra — were significantly superior to placebo for osteoarthritis of the knee, and 5 mg was significantly effective for osteoarthritis of the hip.4-6

Defying FDA Guidelines

For decades, the FDA has advised the use of the lowest doses of anti-inflammatory drugs required by each patient.  Why?  Because we’ve long known that these drugs cause severe, often fatal gastric bleeding and kidney injury.  These side effects are also dose-related, so it is vital for drug companies to produce the lowest, safest, effective dosages of these drugs.  But you ignored this and preferred Bextra to be a super-strong, one-size-fits-all drug, and in doing so you placed millions of people at unnecessary risk.  Serious side effects were inevitable.

Your strategy with Bextra was not only medically improper, but also financially short-sighted.  Look at the drugs withdrawn recently.  Half were one-size-fits-all (Seldane, Hismanal, Redux, Pondimin, Lotronex).  When you produce a one-size-fits-all drug, if toxicities are discovered, your drug is history.  Compare this with Halcion (triazolam), for example.  This sleep remedy, still popular today, was quickly discovered to cause serious memory dysfunctions.  The manufacturer quickly marketed lower doses that  caused far fewer problems, and these lower dosages remained available after the highest dosage was withdrawn.

Another example: If the manufacturers of ibuprofen and naproxen had marketed these drugs as you marketed Bextra, producing ibuprofen and naproxen only at their maximum dosages (ibuprofen 800 mg four-times-a-day, naproxen 500 mg twice-daily), ibuprofen and naproxen would have caused so many serious side effects and deaths, they would have been withdrawn long ago.  Instead, ibuprofen and naproxen are used mainly at their lower doses, serious side effects are fewer, and the drugs have remained top-sellers for decades.

Saving Celebrex — and Protecting Patients

Perhaps this is a strategy you should consider with your one remaining COX-2 drug, Celebrex.  Like Bextra, you marketed Celebrex as a super-strong, mainly one-size-fits-all drug.  Now, Celebrex is on the ropes because of cardiovascular deaths.  The FDA has mandated large black-box warnings added to Celebrex product information.  If the FDA receives more reports of people harmed by Celebrex, the drug will be withdrawn too.

These problems with Celebrex are dose-related.  As with Bextra, you have studies showing that lower doses of Celebrex are safer and effective ,7-8 but you have ignored them.  Maybe it’s time to reconsider.  In fact, the time is long overdue for you to produce lower, safer doses of Celebrex for the protection of patients — and your own financial protection.

PS: Perhaps you could design all of your new drugs in multiple sizes that actually fit patients, and by doing so reassert yourself as a model for drug safety within your own industry.


Jay S. Cohen, M.D.
Associate Professor (voluntary)
Departments of Family and Preventive Medicine and of Psychiatry
University of California, San Diego

Author of:
Over Dose: The Case Against The Drug Companies (Tarcher/Putnam 2001).

What You Must Know About Statin Drugs and Their Natural Alternatives (Square One Publishers 2005).


1.  Bextra advertisement.  Pfizer Inc. 2004.
2.  American Medical Association.  AMA Drug Evaluations, Annual 1993.  Chicago: American Medical Association, 1993.
3.  Clark, WG,  Brater, DC, Johnson, AR.  Goth’s Medical Pharmacology.  12th Edition.  St. Louis: The C.V. Mosby Company, 1988.
4.  Kivitz, A, Eisen, G, Zhao, WW, Bevirt, T, Recker, DP.  Randomized placebo-controlled trial comparing efficacy and safety of valdecoxib with naproxen in patients with osteoarthritis.  Journal of Family Practice 2002;51(6):530-7.
5.  Makarowski, W, Zhao, WW, Bevirt, T, Recker, DP.  Efficacy and safety of the COX-2 specific inhibitor valdecoxib in the management of osteoarthritis of the hip: a randomized, double-blind, placebo-controlled comparison with naproxen.  Osteoarthritis & Cartilage 2002;10(4):290-6.
6.  Ormrod, D, Wellington, K, Wagstaff, AJ.  Valdecoxib.  Drugs 2002;62(14):2059-71.
7.  Bensen WG, Fiechtner JJ, McMillen JI, et al.  Treatment of osteoarthritis with celecoxib: a randomized controlled trial.  Mayo Clinic Proceedings 1999;74(11):1095?]105.
8.  Hubbard, R, Geis, GS, Woods, E, Yu, S, ZHao, W.  Efficacy, tolerability, and safety of celecoxib, a specific Cox-2 inhibitor, in osteoarthritis.  Arthritis and Rheumatism 1998;41(9 suppl):S196[abstract].


NOTE TO READERS: The purpose of this E-Letter is solely informational and educational. The information herein should not be considered to be a substitute for the direct medical advice of your doctor, nor is it meant to encourage the diagnosis or treatment of any illness, disease, or other medical problem by laypersons. If you are under a physician’s care for any condition, he or she can advise you whether the information in this E-Letter is suitable for you. Readers should not make any changes in drugs, doses, or any other aspects of their medical treatment unless specifically directed to do so by their own doctors.

If you have questions about your medications or medical care, Dr. Cohen is available for consultation at his office or by telephone.
If you find this article informative, please tell your friends, family members, colleagues, and doctors about and the free MedicationSense E-Newsletter.

Category: Articles and Reports